I. Summary of Facts and Holdings.392
II. Audited Data . ^ 05 CO
A. Standards of Judicial Review. ^ 05 CO
1. Excess of Statutory Authority . . . ÍD Oi CO
2. Unlawful Proceedings (Procedures) CO O ^
3. Arbitrary and Capricious Actions O (M ^
B. Summary . H W
III. North Carolina Reinsurance Facility . . . <M ^
A. Scope of Review. 03 ^
B. Material and Substantial Evidence . . CO ^
C. Statutory Scheme. CO ^
D. Acquisition and Service Costs. CO ^
E. Cap on Rate Increase . CO ^
F. Summary . CO ^
*387IV.Income on Invested Capital. rfS*. ^ O
A. Error of law . ^ H
B. Majority Rule . ^
C. Statutory Authority. ^ Ül
D. Summary . ^ Oi
V.Underwriting Profit . 00 ^ ^
A. Error of Law. O lO ^
B. Arbitrary and Capricious Actions T-Í lO ^
VI.Burden of Proof. CO lO ^
VII.Specificity of Commissioner’s Order lO LO
VIII.Adequacy of Notice. ÍO LO
IX.Bad Faith of Appellants. t> LO rtf
X.Other Holdings of Court of Appeals 00 LO
XI.Final Disposition . Oi lO Ttf
This opinion deals extensively with certain provisions of the North Carolina Administrative Procedure Act and the powers of State administrative agencies generally, as well as with our general insurance laws.
Numerous opinions of this Court cited in the body of this opinion contain a summary of the history and framework of North Carolina’s insurance laws, codified as Chapter 58 of the General Statutes. See especially In re Filing by Automobile Rate Administrative Office, 278 N.C. 302, 180 S.E. 2d 155 (1971). We therefore find it necessary to present only a limited summary here.
It has been long established that the insurance business is charged with a public interest, and that its, regulation is constitutional. German Alliance Insurance Co. v. Lewis, 233 U.S. 389, 34 S.Ct. 612, 58 L.Ed. 1011 (1914). Likewise, it has been long recognized that regulation of insurance is a function of the states rather than the federal government. Indeed, for many years no effort was made in any court proceedings to apply the Sherman Anti-Trust Act, 15 U.S.C. § 1 et seq., and other acts of Congress to insurance, on the grounds that insurance was not interstate *388commerce, and that Congress did not intend its acts to relate to insurance. However, in 1944, the Supreme Court of the United States held that insurance companies which conducted their activities across state lines were within the regulatory power of Congress under the Commerce Clause of the Federal Constitution, and that insurance was subject to the Sherman Anti-Trust Act. United States v. South-Eastern Underwriters Association, 322 U.S. 533, 64 S.Ct. 1162, 88 L.Ed. 1440 (1944).
Shortly thereafter, Congress enacted the McCarran-Ferguson Act of March 9, 1945, 59 Stat. 33, 15 U.S.C. §§ 1011-1015. The Act, as finally amended, provided, inter alia, that the business of insurance should be subject to the laws of the several states, and not to the acts of Congress (unless such acts relate specifically to insurance), except that the Sherman Act, and certain other acts should be applicable to the business of insurance after 30 June 1948 to the extent such business is not regulated by state law. 15 U.S.C. § 1012.
The North Carolina Legislature responded by enacting Chapter 381 of the 1945 Session Laws codified as G.S. § 58-248.1. The statute vested broad review powers in the Commissioner of Insurance to insure that insurance rates not be unreasonable, inadequate, unfairly discriminatory nor harmful to the public interest. Under the 1945 statute, the Commissioner could act “upon his own motion or upon petition of any aggrieved party.” Id. No periodic filings by the industry were required. However, the 1965 Legislature incorporated such a requirement into G.S. 58-248 by providing in pertinent part that
On or before July 1 of each calendar year the . . . Rate . . . Office shall submit to the Commissioner the data hereinabove referred to for bodily injury and property damage insurance on private passenger vehicles and a rate review based on such data. Such rate proposals shall be approved or disapproved by the Commissioner. . . . (Emphasis added.)
Both appellate courts in this State have had numerous occasions throughout the years to review proceedings before and orders by the Commissioner in ratemaking cases. During the years prior to 1977 the typical case on appeal involved the Commissioner’s disapproval of a rate filing. In most of those cases, this Court or the Court of Appeals found no legal basis for the Com*389missioner’s disapproval and upon remand the Commissioner would find yet another ground for disapproving a proposed rate increase. A stalemate was thus created by the statute’s “prior approval” requirement.
Seemingly in response, the 1977 Legislature enacted significant changes in our insurance laws. See 1977 N.C. Sess. Laws 1119, Ch. 828 (codified in various sections of Ch. 58, Cum. Supp. 1979). The new legislation effected major changes in three general areas of insurance regulation.
Insurance ratemaking was changed from a “prior approval” system to a “file and use” system. To promulgate new or revised rates, the insurer or rating organization is required only to file the rates and accompanying supportive data with the Commissioner prior to the effective date of the rates. The rates then take effect automatically and remain in effect until revised rates are filed. The Commissioner’s prior approval is not required for rates to take effect. See G.S. 58-124.20 (essential lines), G.S. 58-131.39 and G.S. 58-131.41 (nonessential lines).
The statutes also outline procedures by which the Commissioner may contest such rates after they are filed. He must hold a hearing. G.S. 58-124.21 (Cum. Supp. 1979) for essential lines of insurance and G.S. 58-131.42 for nonessential lines. If he finds that rates are not in compliance with statutory standards, G.S. 58-124.19 (essential lines) and G.S. 58-131.37 (nonessential lines), he may disapprove the rates and declare them ineffective. G.S. 58-124.21 (essential lines) and G.S. 58-131.42 (nonessential lines). His decision is subject to judicial review, G.S. 58424.22(a) (essential lines) and G.S. 58431.54(b) (nonessential lines), but the insurers may continue to use the rates pending such review if the purportedly excessive premiums are placed in an escrow account. G.S. 58424.22(b) (essential lines) and G.S. 58431.42(b) (nonessential lines).
In abandoning the prior approval system for the file and use system, North Carolina has joined the general trend of regulatory programs among the states. R. Keeton, Basic Text on Insurance Law § 8.4(b) (1971). Some states have even eliminated the filing requirement. Id.
For ratemaking purposes, the 1977 legislation divided insurance into two categories called essential and nonessential lines. Survey of Developments in North Carolina Law — Insurance, 56 N.C.L. Rev. 1084, 1085 (1978). Previously, the types of insurance subject to rate regulation had been divided into five categories, each regulated in a different manner — fire, casualty, miscellaneous lines, automobile liability and workers’ compensation. Rate regulation patterns for the two new categories are established based upon the mandatory or voluntary Rating Bureau membership. The new file and use system applies to both.
Nonessential lines of insurance, including certain fire and property insurance, casualty insurance and inland marine insurance are governed by Chapter 58, Article 13C. This statute establishes a system of voluntary rating bureau membership. It provides that insurance rates should not be “excessive, inadequate or unfairly discriminatory,” G.S. 58-131.34(1), and that the most effective way to achieve rates is through “reasonable price competition among insurers.” G.S. 58-131.34(3). Detailed provisions and factors to be considered are set out. Rating organizations, available to all insurers operating in the State, are authorized, G.S. 58-131.34(2), but insurers are not required to join a rating bureau and may use their own rates. G.S. 58-131.41. However, while recognizing that cooperation among insurers is desirable, the statute provides that regulation is necessary to prevent restraint of competition. G.S. 58-131.34(4).
Chapter 58, Article 12B governs the essential lines of insurance. These include certain residential fire and property insurance, automobile theft and physical damage insurance, automobile liability insurance and allied lines, and workers’ compensation and employers’ liability insurance. G.S. 58-124.17(1). The North Carolina Rate Bureau is established and all insurance companies writing any of the essential lines in North Carolina are required to be members. G.S. 58-124.17(1); G.S. 58-124.18. Hence, the major distinguishing factor in rate regulation between essential and nonessential lines is mandatory Bureau membership and man*391datory adherence to rules established by the Bureau. Moreover, no price competition is provided for among the companies in the essential lines, unlike the plan for nonessential lines which allows competition. G.S. 58-131.41.
G.S. 58-124.19 sets out the factors to be considered in establishing rates for essential lines. The basic standard for essential lines is the same as for nonessential lines — rates are not to be “excessive, inadequate or unfairly discriminatory.” G.S. 58-124.19(1). Risks may be classified for ratemaking purposes, but the classification plan for automobile insurance may not be based upon the age or sex of the persons insured. G.S. 58-124.19(4). Some of these and other factors to be considered in ratemaking are discussed in the body of this opinion.
The 1977 Legislature also made significant changes in the statutory scheme for dealing with high-risk insureds in motor vehicle insurance. G.S. 58-248.26 to .40. All insurance companies licensed to write motor vehicle insurance in North Carolina are required to participate in the North Carolina Reinsurance Facility, a statutory reinsurance pool for the high-risk driver of motor vehicles. G.S. 58-248.34(e).
The most far-reaching change in the operation of the Facility was the establishment of procedures to make the Facility self-sustaining. Under the new law, losses sustained by the Facility are to be recouped according to a statutory prescription. G.S. 58-248.34(e). A detailed discussion of statutes relating to the Facility is included in Section III. of this opinion.
* * * *
We note that all of our discussion in summary above involves only the 1977 insurance legislation. The 1979 Legislature also made significant changes in our insurance laws. We parenthetically mention some of these in our opinion. However, we issue the caution that, since all four insurance ratemaking decisions handed down today are based on pre-1979 legislation, reference should be made to the later changes for applicable rate filings.
Summary Of Facts And Holdings
On 29 November 1977 the North Carolina Rate Bureau, on behalf of its member companies and the North Carolina Reinsurance Facility, filed with the Commissioner of Insurance a proposed revised premium rate schedule for automobile insurance, including bodily injury and property damage liability, medical payments, and physicial damage insurance for non-fleet private passenger automobiles. The filing stated that calculations substantiated the need for a statewide average rate increase of 23.2%, but in accordance with the requirements of G.S. 58-124.26 the filing had been limited to an overall increase of 6%. The filing also proposed that rates for risks ceded to the North Carolina Reinsurance Facility be 10% higher than rates for risks voluntarily retained, and that ± 5% territorial rate differences be established.
The Commissioner gave notice of public hearing, contending that the filing failed to comply with statutory requirements in a number of respects. After the hearing, the Commissioner made extensive findings of fact and conclusions of law and disapproved the filing in its entirety. In his disapproval order, he allowed the Bureau 60 days to submit an amended filing consistent with his findings and conclusions and ordered that' the Bureau by its amended filing submit the exact data and information he had requested in the notice of public hearing.
The Rate Bureau appealed to the North Carolina Court of Appeals. That court, speaking through Arnold, Judge, affirmed in part and reversed in part.
We note the various holdings of the Court of Appeals and our response on review:
(1) The Court of Appeals held that the Commissioner may require that company data in this insurance ratemaking hearing be audited. We reverse. We hold that while such a requirement, as a general rule, does not exceed the Commissioner’s statutory authority, the Commissioner here failed to comply with lawful procedures and his actions were arbitrary and capricious.
(2) The Court of Appeals held that the proposed 10% rate differential for insureds ceded to the North Carolina Reinsurance *393Facility was unfairly discriminatory. We reverse. Applying the whole record test, we hold that there was insubstantial evidence in the record to support the Commissioner’s findings and conclusions of unfair discrimination.
(3) The Court of Appeals held that the Commissioner may require the consideration of income on invested capital in an insurance ratemaking case. We reverse. We hold that the Commissioner erred as a matter of law in concluding that the law of this jurisdiction allows consideration of income from invested capital in an insurance ratemaking case.
(4) The Court of Appeals held that the Commissioner’s implementation of a “capital asset pricing model” to calculate underwriting profit margins was erroneous. We affirm. We hold that the Commissioner’s attempted implementation of a “capital asset pricing model” to calculate underwriting profit margins was erroneous as a matter of law and was arbitrary and capricious.
(5) The Court of Appeals held that the enactment of G.S. 58-124.21 did not transfer the burden of proof in a ratemaking hearing to the Commissioner of Insurance. We affirm. We hold that the burden of proving the need and reasonableness of an insurance rate increase continues to rest with the Rate Bureau.
(6) The Court of Appeals held that the Commissioner did not fail to comply with the statutory requirement that in his order disapproving a filing he indicate “wherein and to what extent such filing is deemed to be improper.” G.S. 58424.21(a). We affirm, albeit for different reasons than those noted by the Court of Appeals.
(7) The Court of Appeals held that the Commissioner complied with the notice requirements of G.S. 58424.21(a). We reverse. We hold that the Commissioner failed to comply with the notice requirements of this statute because no notice was served upon appellants questioning the reliability of the data submitted.
(8) We hold that the Commissioner erroneously found and concluded that the appellants acted in bad faith.
(9) We leave undisturbed those portions of the Court of Appeals’ decision finding (a) that projections of territorial rate differences did not consider the new classification plan and that the *394alleged failure to consider the new classification plan resulted in excessive rates was not supported by the evidence, (b) that there was no evidence to support the Commissioner’s disapproval of deductible collision rates as being excessive, and (c) that the appeal by the Rate Bureau nullified the Commissioner’s order to submit an amended filing. These holdings were not brought before us on this appeal.
(10) While several portions of our decision are supportive of certain positions and apparent general goals of the Commissioner, the magnitude of the multiple legal errors in the proceedings before the Commissioner and in his order compel us to reverse the order, declare it null and void and order the filing approved. Moreover, we order that the escrowed premium funds representing this proposed rate increase be remitted to the member insurers pursuant to G.S. 58-124.21(b).
Other facts important to an understanding of our decision are noted below.
Appellants first contend that the Commissioner erred in finding and concluding that unaudited data in an insurance ratemak-ing hearing is unreliable and incredible. By this assignment of error, appellants compel our consideration of the several subsections of our judicial review statutes applicable to insurance ratemaking.
A. Standards of Judicial Review
G.S. 150A-43, a part of the North Carolina Administrative Procedure Act (NCAPA), provides in pertinent part that, “[a]ny person who is aggrieved by a final agency decision ... is entitled to judicial review of such decision under this Article, unless adequate procedure for judicial review is provided by some other statute, in which case the review shall be under such other statute.” (Emphasis added.) The Department of Insurance is an “agency” subject to the provisions of the NCAPA. G.S. 150A-2(1). The question, therefore, is whether “some other statute” provides “adequate procedure for judicial review” such that the NCAPA review statutes become inapplicable.
*395  In determining what is “adequate procedure for judicial review,” as those words appeared in our former statute, G.S. 143-307, this Court held that an adequate procedure for judicial review exists “only if the scope of review is equal to that under G.S. Chapter 143, Article 33, 143-306 et seq.” Jarrell v. Board of Adjustment, 258 N.C. 476, 480, 128 S.E. 2d 879, 883 (1963). Effective 1 February 1976, G.S. 143-307 was replaced by G.S. 150A-43. Law of March 24, 1975, 1975 N.C. Sess. Laws 44, Ch. 69, s. 4; Law of April 12, 1974, 1973 N.C. Sess. Laws 691, Ch. 1331, s. 2. We now hold that “adequate procedure for judicial review,” as those words appear in present G.S. 150A-43, exists only if the scope of review is equal to that under present Article 4 of G.S. Chapter 150A.
While it has been held that the scope of review provided by the NCAPA is substantially broader than that provided by other sections of G.S. Chapter 58 such that the NCAPA should control, Occidental Life Insurance Co. v. Ingram, 34 N.C. App. 619, 240 S.E. 2d 460 (1977), we find the applicable Chapter 58 provision for judicial review of the ratemaking cases to be practically identical to the NCAPA provisions. Compare G.S. 58-9.6(b) with G.S. 150A-51.
 There are of course subtle differences. For example, G.S. 150A-51 provides that an agency decision may be reversed or modified if the substantial rights of petitioners “may have been prejudiced.” (Emphasis added.) The comparable provision in G.S. 58-9.6 provides that such rights “have been prejudiced.” Id. 9.6(b). (Emphasis added.) For this reason, and in the interest of uniformity in judicial review of administrative decisions, see Daye, North Carolina’s New Administrative Procedure Act: An Interpretive Analysis, 53 N.C.L. Rev. 833, 899 (1975) (hereinafter Daye), we hold that G.S. 150A-51 is the controlling judicial review statute in insurance ratemaking cases. However, to the extent that G.S. 58-9.6(b) adds to the judicial review function as noted below and in light of the virtually identical thrust of the two statutes, we elect to proceed by applying the review standards of both G.S. 58-9.6 and G.S. 150A-51, where those standards may be construed as being consistent with each other. Both provide that the court may (1) affirm, or (2) reverse, (3) modify, or (4) remand the case for further proceedings. G.S. 58-9.6 also provides the court may declare the Commissioner’s order null and void if the *396substantial rights of the appellants “have been”1 prejudiced because the Commissioner’s findings, inferences, conclusions or decisions are:
(1) In violation of constitutional provisions, or
(2) In excess of statutory authority or jurisdiction of the Commissioner, or
(3) Made upon unlawful proceedings, or
(4) Affected by other errors of law, or
(5) Unsupported by material and substantial evidence in view of the entire record as submitted, or
(6) Arbitrary or capricious.
G.S. 58-9.6(b). See also G.S. 150A-51.
G.S. 58-9.6(b) also provides that “[s]o far as necessary to the decision and where presented, the court shall decide all relevant questions of law, interpret constitutional and statutory provisions, and determine the meaning and applicability of the terms of any action of the Commissioner.”
Here, appellants rely on G.S. 58-9.6(b)(2), (3) and (6) above, contending that the Commissioner has only such powers as are given him by statute and, absent specific statutory authority for the audited data requirement, the Commissioner is without authority to order this particular form of evidence.
1. Excess of Statutory Authority
We first address the question whether the Commissioner’s action was “in excess of statutory authority as contemplated by G.S. 58-9.6(b)(2) and G.S. 150A-5K2). Turning to the applicable statutory provisions, G.S. 58-9 sets out the general powers and duties of the Commissioner of Insurance and confers upon him the duty to
[s]ee that all laws of this State governing insurance companies ... or bureaus relating to the business of insurance are faithfully executed, and to that end he shall have power and authority to make rules and regulations, not inconsistent *397with law, to enforce, carry out and make effective the provisions of this Chapter, and to make such further rules and regulations not contrary to any provisions of this Chapter which will prevent practices injurious to the public by insurance companies. . . .
G.S. 58-124.19 sets out the standards and factors to be considered in ratemaking. It provides that “[r]ates shall not be excessive, inadequate or unfairly discriminatory.” At the time of the Commissioner’s order, G.S. 58-124.19(2) provided that:
Due consideration shall be given to past and prospective loss experience, within this State; to the hazards of conflagration and catastrophe; to a reasonable margin for underwriting profit and to contingencies; to dividends, savings or unabsorbed premium deposits allowed or returned by insurers to their policyholders, members or subscribers; to past and prospective expenses specially applicable to this State; and to all other relevant factors including judgment factors, deemed relevant, within this State. . . .
Our Legislature has generally addressed the question of data collection and availability in two other statutes. G.S. 58-124.18(d) provides:
The Commissioner of Insurance is hereby authorized to compel the production of all books, data, papers and records and any other data necessary to compile statistics for the purpose of determining the underwriting experience of lines of insurance referred to in this Article, and this information shall be available and for the use of the Bureau for the capitulation and promulgation of rates on lines of insurance as are subject to the rate-making authority of the Bureau.
G.S. 58424.20(c) provides that:
The Bureau shall maintain reasonable records, of the type and kind reasonably adapted to its method of operation, of the experience of its members and of the data, statistics or information collected or used by it in connection with the rates, rating plans, rating systems, underwriting rules, policy or bond forms, surveys or inspections made or used by it.
*398With this statutory background, we turn to the appellants’ contention that the Commissioner exceeded his statutory authority in ordering that data be audited. Appellants contend that none of the statutes require that data be audited and the Commissioner has no power to interpolate that requirement into the statutes. Moreover, appellants assert that G.S. 58-124.20 vests the authority to promulgate insurance rates in the Rate Bureau and G.S. 58-124.21 gives the Commissioner only a limited power of disapproval. The latter statute provides that: “If the Commissioner after hearing finds that the filing does not comply with the provisions of this Article, he may issue his order determining wherein and to what extent such filing is deemed to be improper. . . .” (Emphasis added.) Because Article 12B nowhere specifically states that data be audited, appellants argue, the Commissioner improperly rejected the filing in finding appellants failed to “comply with the provisions of [the] Article.”
Appellants rely on previous statements of this Court that the Commissioner has, in the regulation of insurance rates, only such authority as has been conferred upon him by statute. State ex rel. Commissioner of Insurance v. North Carolina Fire Insurance Rating Bureau, 292 N.C. 471, 234 S.E. 2d 720 (1977); State ex rel. Commissioner of Insurance v. North Carolina Automobile Rate Administrative Office, 287 N.C. 192, 214 S.E. 2d 98 (1975); In re North Carolina Fire Insurance Rating Bureau, 275 N.C. 15, 33, 165 S.E. 2d 207, 220 (1969).
In limited context, appellants correctly cite the established rule in this jurisdiction. In State ex rel. Commissioner of Insurance v. North Carolina Automobile Rate Administrative Office, supra, Justice Huskins, writing for the Court, stated:
While the Office of Commissioner of Insurance is created by Article III, sec. 7(1) of the North Carolina Constitution, sec. 7(2) of that Article says his duties shall be prescribed by law. Hence, the power and authority of the Commissioner emanate from the General Assembly and are limited by legislative prescription. The only power he has to fix rates is such power as the General Assembly has delegated to and vested in him.
*399The stated rule is in accord with well-established principles of administrative law. The powers and authority of administrative officers and agencies are derived from, defined and limited by constitution, statute, or other legislative enactment. 73 C.J.S., Public Administrative Bodies and Procedure § 49 (1951 and Cum. Supp. 1980) and cases cited therein. Thus, “[i]n fixing by law the premium rate, it is the legislative power of the State which is being exercised.” In re Filing by North Carolina Fire Insurance Rating Bureau, supra at 32, 165 S.E. 2d at 219. It is beyond question that the Legislature may so delegate this authority to an administrative officer provided it prescribes sufficiently clear standards to control his discretion. In re Filing by North Carolina Fire Insurance, supra; State ex rel. Utilities Commission v. North Carolina and Southern Bell Telephone and Telegraph Company, 239 N.C. 333, 80 S.E. 2d 133 (1954).
We note that appellants do not contend that the Legislature improperly delegated its authority to the Commissioner nor that it failed to prescribe sufficiently clear standards to control his discretion. They contend only that the Commissioner exceeded his existing statutory authority.
An issue as to the existence of power or authority in a particular administrative agency is one primarily of statutory construction. Joseph Burstyn, Inc. v. Wilson, 303 N.Y. 242, 101 N.E. 2d 665 (1951), rev’d on other grounds, 343 U.S. 495, 72 S.Ct. 777, 96 L.Ed. 1098 (1952).
In construing the laws creating and empowering administrative agencies, as in any area of law, the primary function of a court is to ensure that the purpose of the Legislature in enacting the law, sometimes referred to as legislative intent, is accomplished. In re Filing by the N.C. Fire Insurance Rating Bureau, supra; In re Dillingham, 257 N.C. 684, 127 S.E. 2d 584 (1962). The best indicia of that legislative purpose are “the language of the statute, the spirit of the act, and what the act seeks to accomplish.” Stevenson v. City of Durham, 281 N.C. 300, 303, 188 S.E. 2d 281, 283 (1972). In addition, a court may consider “circumstances surrounding [the statute’s] adoption which throw light upon the evil sought to be remedied.” State ex rel. N.C. Milk Commission v. National Food Stores, Inc., 270 N.C. 323, 332, 154 S.E. 2d 548, 555 (1967).
*400We should be guided by the rules of construction that statutes in pari materia, and all parts thereof, should be construed together and compared with each other. Redevelopment Commission v. Security National Bank of Greensboro, 252 N.C. 595, 114 S.E. 2d 688 (1960). Such statutes should be reconciled with each other when possible, and any irreconcilable ambiguity should be resolved so as to effectuate the true legislative intent. Duncan v. Carpenter, 233 N.C. 422, 64 S.E. 2d 410 (1951).
Applying the foregoing, we first note that neither G.S. 58424.18(d) nor G.S. 58424.20(c), the statutes dealing with data collection and availability, mentions a requirement that data be audited. However each section requires that certain data be collected, and the Commissioner is given a certain statutory flexibility in determining what and how that data is to be gathered. G.S. 58424.18(d) authorizes the Commissioner to require “any other data necessary to compile statistics.” Former G.S. 58-124.19(2), under which this proceeding took place, set out the factors to be considered in ratemaking, and also referred to “all other relevant factors including judgment factors, deemed relevant” in addition to the factors specially named. G.S. 58-9(1) provides that the Commissioner “shall have power and authority to make rules and regulations, not inconsistent with law . . . and to make such further rules and regulations not contrary to any provision of this Chapter which will prevent practices injurious to the public by insurance companies.”
 Viewing these statutes in pari materia, we think it without question that our Legislature intended for the Commissioner of Insurance to promulgate such reasonable rules and regulations as he deems necessary to discharge the functions of his office in seeing “that all laws of this State governing insurance companies ... or bureaus relating to the business of insurance are faithfully executed.” Thus the desire of the Commissioner that data submitted in a ratemaking case be audited is not, in our interpretation, in excess of the statutory powers so construed.
Our view is, we think, consistent with the weight of authority in other jurisdictions.
It is generally recognized that investigatory or inquisitorial powers, power to inspect, or to require the disclosure of information by means of accounts, records, reports, or statements are *401conferred on practically all administrative agencies. Indeed, such powers constitute functions which distinguish an administrative agency from a court. 1 Am. Jur. 2d, Administrative Law § 85. Administrative agencies often have the duty to inquire into the management of regulated businesses and in order to perform their functions efficiently it is essential that the agency have access to many facts, often not voluntarily supplied. State ex rel. Railroad and Warehouse Commission v. Mees, 235 Minn. 42, 49 N.W. 2d 386 (1951).
The fact that an asserted power is novel and unprecedented does not mean that it does not exist as a statutory power. United States v. Morton Salt Company, 338 U.S. 632, 70 S.Ct. 357, 94 L.Ed. 401 (1950).
The United States Supreme Court addressed this issue in the Permian Basin Area Rate Cases, 390 U.S. 747, 88 S.Ct. 1344, 20 L.Ed. 2d 312 (1968). There, the Federal Power Commission had, contrary to years of custom, set rates for a geographical area of natural gas producers instead of setting rates for individual companies within that geographical area. The Court held that the Federal Power Commission did not abuse or exceed its statutory authority in adopting the system of area price regulation, supplemented by a provision for moratorium upon certain price increases and for exceptions for smaller producers. In interpreting the provisions of the act creating the agency, the Court stated:
This Court has repeatedly held that the width of administrative authority must be measured in part by the purposes for which it was conferred; [citations omitted]. Surely the Commission’s broad responsibilities therefore demand a generous construction of its statutory authority.
Such a construction is consistent with the view of administrative rate making uniformly taken by this Court. The Court has said that the “legislative discretion implied in the rate making power necessarily extends to the entire legislative process, embracing the method used in reaching the legislative determination as well as that determination itself.” [Citations omitted.] It follows that rule-making agencies are not bound to the service of any single regulatory formula; they are permitted, unless their statutory authority otherwise plainly indicates, “to make the pragmatic ad*402justments which may be called for by particular circumstances. [Citations omitted.]”
Contrary to the reasoning in the Permian Basin cases, supra, that unless a statute forbids a practice, a ratemaking body should have authority to make pragmatic adjustments, appellants strenuously argue that had the Legislature intended the Commissioner to have the power to require audited data, it would have expressly given it to him. We think appellants expect too much of our Legislature and too little of our state administrative agencies.
One of the primary problems in the case before us, and in other cases involving the interpretation of an administrative agency’s power, results from the established law that legislative power may not be delegated to an administrative agency unless adequate standards are included in the delegating legislation. The Legislature can obviously not anticipate every problem which will arise before an administrative agency in the administration of an act. The legislative process would be completely frustrated if that body were required to appraise beforehand the myriad situations to which it wished a particular policy to be applied and to formulate specific rules for each situation. Clearly, then, we must expect the Legislature to legislate only so far as is reasonable and practical to do and we must leave to executive officers the authority to accomplish the legislative purpose, guided of course by proper standards. See, e.g., American Power and Light Company v. Securities and Exchange Commission, 329 U.S. 90, 67 S.Ct. 133, 91 L.Ed. 103 (1946). The modern tendency is to be more liberal in permitting grants of discretion to administrative agencies in order to ease the administration of laws as the complexity of economic and governmental conditions increases. The realities of modern legislation dealing with complex economic and social problems have led to judicial approval of broad standards for administrative action. Detailed standards are not required, especially in regulatory enactments under the police power. 1 Am. Jur. 2d, Administrative Law § 118 (1951).
North Carolina cases have long been consistent with this “modern tendency.” Pue v. Hood, 222 N.C. 310, 22 S.E. 2d 896 (1942), reviewed the action of the Commissioner of Banks in denying an application for a bank charter. There the Court stated,
*403It cannot be questioned that the Legislature would have the authority to investigate and decide this question before authorizing incorporation of a bank. But surely the Legislature cannot meet in session and determine the existence or nonexistence of this condition precedent which it has prescribed every time an application for a [bank] charter is received by the Secretary of State.
It may, instead, create an administrative investigatory, fact-finding agency to perform this function, administrative and not judicial in nature.
In State ex rel. North Carolina Utilities Commission v. Atlantic Coast Line Railroad Company, 224 N.C. 283, 29 S.E. 2d 912 (1944), this Court considered inter alia the question whether the Utilities Commission had authority to require certain utilities give 30 days’ written notice of rate increases. This Court held that under general authority to formulate regulations, an administrative agency of the State may prescribe by rule the procedure by which a right granted may be exercised.
The acts of administrative or executive officers are not to be set at nought by recourse to the courts. Nor are courts charged with the duty or vested with the authority to supervise administrative and executive agencies of our government. However, a court of competent jurisdiction may determine in a proper proceeding whether a public official has acted capriciously or arbitrarily or in bad faith or in disregard of the law. Pue v. Hood, Comr. of Banks, supra. And it may compel action in good faith in accord with the law. But when the jurisdiction of a court is properly invoked to review the action of a public official to determine whether he, in choosing one of two or more courses of action, abused his discretion, the court may not direct any particular course of action. It only decides whether the action of the public official was contrary to law or so patently in bad faith as to evidence arbitrary abuse of his right of choice. If the officer acted within the law and in good faith in the exercise of his *404best judgment, the court must decline to interfere even though it is convinced the official chose the wrong course of action. The right to err is one of the rights —and perhaps one of the weaknesses — of our democratic form of government. In any event, we operate under the philosophy of the separation of powers, and the courts were not created or vested with authority to act as supervisory agencies to control and direct the action of executive and administrative agencies or officials. So long as officers act in good faith and in accord with the law, the courts are powerless to act — and rightly so.
Id. at 407-08, 90 S.E. 2d at 702-03.
The Legislature has not set out in detail every incidental power belonging to and which may be exercised by the Commission. As a practical matter the Legislature could not foresee all the problems incidental to the effective carrying out of the duties and responsibilities of the Commission. Of necessity it provided for those matters in general terms. Where a course of action is reasonably necessary for the effective prosecution of the Commission’s obligation to supervise the construction, repair and maintenance of public highways, the power to take such action must be implied from the general authority given and the duty imposed. Mosteller v. Southern R. R. Company, 220 N.C. 275, 280, 17 S.E. 2d 133. “Administrative boards, commissions and officers have no common-law powers. Their powers are limited by the statutes creating them to those conferred expressly or/by necessary or fair implication. ... In determining whether a board or commission has a certain power, the authority given should be liberally construed in the light of the purposes for which it was created and that which is incidentally necessary to a full exposition of the legislative intent should be upheld as being germane to the law. In the construction of a grant of power, it is a general principle of law that where the end is required the appropriate means are given. . . . However, powers should not be extended by implication beyond what may be necessary for their just and reasonable execution.’’ 42 Am. Jur., Public Administrative Law, 26, pp. 316-318.
Appellants also argue that the Commissioner improperly found and concluded that unaudited data was unreliable. They assert there is a lack of sufficient evidence to support this finding and conclusion because only one witness, qualified at the hearing as an expert in accounting and financial reporting, testified that “unaudited reports cannot be relied upon.” This evidence was uncontested.
 In asserting their position, appellants correctly argue that the “whole record” test is applicable to judicial review of administrative decisions in North Carolina, citing In re Rogers, 297 N.C. 48, 253 S.E. 2d 912 (1979); Thompson v. Wake County Board of Education, 292 N.C. 406, 233 S.E. 2d 538 (1977). Moreover, both G.S. 58-9.6(b)(5) and G.S. 150A-5K5) put forth that test as a proper standard of judicial review of these proceedings. They argue that review of the record as a whole reveals insufficient evidence for the Commissioner’s finding that unaudited data is unreliable.
Unlike Thompson v. Wake County, supra, and In re Rogers, supra, where the Court was concerned with conflicting and contradictory evidence, the expert witness’s testimony here with respect to unaudited data was not contradicted. Indeed, the witness was not even cross-examined on this point.
Appellants further argue, however, that the whole record discloses “that the collection of insurance statistical data is an unbelievably complex process which has been painstakingly developed and meticulously documented;” and that the methods by which the statistics are collected and assembled are the same in North Carolina as in 47 other states. Appellants’ brief presents a lengthy explanation of how the statistical agents and the Rate Bureau compile and evaluate statistical data.
We are not concerned, however, with either the number of states who do things this way or the complexity of the data collection process. We are concerned with the amount of evidence in the record which supports the Commissioner’s order.
What appellants seem to be arguing is that we hold as error the Commissioner’s reliance on uncontested evidence presented to him. This we are unwilling to do.
*406  North Carolina is in accord with the well-established rule that it is for the administrative body, in an adjudicatory proceeding, to determine the weight and sufficiency of the evidence and the credibility of the witnesses, to draw inferences from the facts, and to appraise conflicting and circumstantial evidence. 73 C.J.S., Public Administrative Bodies and Procedure § 126. See, e.g., State ex rel. Commissioner of Insurance v. N.C. Automobile Rate Administrative Office, supra; State ex rel. Commissioner of Insurance v. N.C. Fire Insurance Rating Bureau, supra. The credibility of witnesses and the probative value of particular testimony are for the administrative body to determine, and it may accept or reject in whole or part the testimony of any witness. 73 C.J.S., supra at § 126. Hence, applying the whole record test to the issue of audited data, we find no error in the Commissioner’s election to accord the necessary weight and credibility to the testimony of the single uncontested expert witness testifying on auditing.
Finally, appellants’ reliance on previous decisions of this Court as authority for the position that the Commissioner exceeded his statutory authority in ordering audited data is misplaced. In each of the cases relied upon by the appellants, the Commissioner clearly exceeded his statutory authority in fixing premium rates in factual situations clearly distinguishable from that disclosed by this record. For example, in State ex rel. Commissioner of Insurance v. N.C. Automobile Rate Administrative Office, 292 N.C. 1, 231 S.E. 2d 867 (1977), the Commissioner ordered that private passenger automobile insurance rates be decreased by 23.8% for bodily injury and increased by 2.5% for property damage. This Court found that the statute applicable at that time allowed the Commissioner to (1) either approve all of the increase proposed by the rate office, (2) approve a part of the proposed increase or (3) disapprove the entire proposed increase. The statute did not authorize the Commissioner to order a reduction in then-existing rates. Therefore, he clearly exceeded his statutory authority when he ordered a reduction of a rate. In that same case, we note, Justice, now Chief Justice, Branch used language far more pertinent to the issue before us than that relied on by appellants: “The language of G.S. 58-248 does not restrict the Commissioner’s consideration to the statistical data furnished by the Rate Office and he may consider evidence from *407other sources if it is otherwise competent.” Id. at 18, 231 S.E. 2d at 876.
The Commissioner of Insurance is considered to be a specialist in the field of insurance and his projection of past experience and present conditions into the future is assumed to be correct and proper if supported by substantial evidence. Expert testimony, otherwise competent, that a trend upward or downward may reasonably be expected to continue into the future is evidence of “reasonable and related factors” which the Commissioner may consider in making his projections. The statute does not require that procedures and methods for trending loss experience for the future shall be frozen.
Id. at 21-22, 231 S.E. 2d at 878. (Emphasis in original.)
Therefore, the Court held that the Commissioner did not err when, rather than measuring automobile property damage insurance trends separately from paid claim costs and paid claim frequency as the automobile rate administrative office had done in its filing according to its usual methodology, he chose instead to apply trending factors to the composite of average paid claim costs and frequency or average loss cost per automobile.
Indeed, in many of our previous decisions on insurance, we have stressed the Commissioner’s statutory ability to compel special statistical data.
In In re N.C. Fire Insurance Rating Bureau, supra, Justice Lake said:
It is, of course, within the sound discretion of the Commissioner to require complex statistical exhibits to be made available to the adverse party prior to the hearing, to restrict or deny the use of newly developed statistical data sprung suddenly at the hearing by either party to the surprise of the other, and to grant such recess of the hearing as he may deem necessary to permit reasonable opportunity to study such data and to prepare evidence to refute it.
*408In State ex rel. Commissioner of Insurance v. North Carolina Automobile Rate Administrative Office, 293 N.C. 365, 239 S.E. 2d 48 (1977), this Court held, inter alia, that the fact that orders were based in part on calculations derived from operator license statistics maintained by the Department of Motor Vehicles and from the penalty point system was not a basis for disturbing the Commissioner’s orders. Justice Exum noted that the credibility of testimony is for the Commissioner to determine. He added, “There is nothing sacrosanct about so-called ‘insurance statistics.’ ” Id. at 384, 239 S.E. 2d at 60. And, “Insurance data compiled by the Rate Office, insofar as it is shown to be reliable and fairly compiled, is valuable and should be considered. The Commissioner may also consider evidence, otherwise competent, from other sources.” Id. at 384-85, 239 S.E. 2d at 60 (emphasis added).
The Commissioner’s statutory authority to require certain kinds of data submission is therefore unquestioned.
 In light of the foregoing, we hold that the Commissioner’s findings and conclusions that data submitted in an insurance rate-making case be audited were not “in excess of statutory authority” as contemplated by G.S. 58-9.6(b)(2) or G.S. 150A-5U2).
2. Unlawful Proceedings (Procedures)
 We next address the question whether the Commissioner’s action was “made upon unlawful proceedings” or “procedures” as contemplated by G.S. 58-9.6(b)(3) and G.S. 150A-5H3). We first note that, while the prohibition against agency action “in excess of statutory authority,” G.S. 58-9.6(b)(2) and G.S. 150A-5K2), and one “made upon unlawful procedure,” G.S. 58-9.6(b)(3), see also G.S. 150A-5H3), appear redundant, the distinction is significant indeed. The former refers to the general authority of an administrative agency to properly discharge its statutorily assigned responsibilities. The latter refers to the procedures employed by the agency in discharging its statutorily authorized acts. We have held above that the Commissioner had the general statutory authority to require audited data in this proceeding. We are now compelled to hold, however, that he did not follow lawful procedure in attempting to do so.
*409The rulemaking power of an administrative agency is restricted by law apart from the statute conferring power and an agency having authority to effectuate the policies of a particular statute may not effectuate such policies so singlemindedly that it wholly ignores other and equally important legislative objectives. 1 Am. Jur. 2d, Administrative Law § 72. See also Edgerton v. International Company, 89 So. 2d 488, 490 (Fla. 1956). This is especially true in the case of agencies which have both ac-cusatorial and judgmental powers. The potential for unfairness and abuse is obvious in a situation in which an administrative officer is vested with broad rulemaking powers, determining the admissibility and weight of evidence in hearings and making the final determination on the merits of an action, as is the Commissioner of Insurance in ratemaking cases. Indeed, one of the fundamental purposes in the creation of administrative procedure acts was to minimize the potential of unfairness in embodying in one person or agency these various functions. See generally 1 Am. Jur. 2d, Administrative Law § 78. Since an administrative agency is vested with powers both quasi-judicial and quasi-legislative, such procedural safeguards are essential.
Our Legislature, in providing that agency action is unauthorized if “made upon unlawful procedure” was clearly sensitive to the potential abuse mentioned above. “This provision authorizes a court to reverse or modify agency action that is not in accordance with the procedural requirements specified in the NCAPA; or with those required under another statute governing agency procedure.” Daye, supra at 914. We therefore turn to a consideration of lawful agency procedures in general and the North Carolina Administrative Procedure Act in particular.
Appellants argue, albeit briefly and without citation of authority, that the Commissioner converted a ratemaking case into a rulemaking hearing and thereby violated the terms of the North Carolina Administrative Procedure Act (NCAPA), G.S. 150A-1 et seq. The Commissioner’s response is equally terse: He argues that this proceeding is exempt from the NCAPA by virtue of certain of its provisions. A determination of the applicability of the NCAPA to this proceeding is therefore necessary to resolve the question whether the Commissioner acted “upon unlawful procedure” in finding and concluding that unaudited data presented in a ratemaking hearing is unreliable and incredible. We think *410that the NCAPA is applicable and that the Commissioner violated its rulemaking requirements.
G.S. 150A-9 provides in pertinent part:
It is the intent of this Article to establish basic minimum procedural requirements for the adoption, amendment, or repeal of administrative rules. Except for emergency rules . . ., the provisions . . . are applicable to the exercise of any rule-making authority conferred by any statute.No rule hereafter adopted is valid unless adopted in substantial compliance with this Article.
G.S. 150A-10 then defines “rule” to mean “each agency regulation, standard or statement of general applicability that implements or prescribes law or policy, or describes the organization, procedure, or practice requirements of any agency. The term includes the amendment or repeal of a prior rule. . . .”
The statute then lists six exclusions to the rule definition including the following two, interpretations of which are crucial to the issue before us:
“(4) Statements of policy or interpretations that are made in the decision of a contested case; . . .
(6) Interpretative rules and general statements of policy of the agency.”
The Commissioner argues that either of the quoted exclusions would relieve him of NCAPA requirements with respect to his determination that audited data is essential in a ratemaking hearing. G.S. 150A-2(2) does specifically provide that a ratemaking proceeding is a “contested case” within the meaning of the NCAPA. The primary question, therefore, revolves around the meaning of “interpretative” rules and “statements of policy.”
It becomes readily apparent from the statutory definition of “rule,” which includes six exceptions, that different types of rules were contemplated. This is crucial in the issue confronting us here for two reasons: (1) The distinction is important in determining the requirements that will be imposed in establishing the procedures used in adopting and promulgating the rule, and (2) the distinction between different types of rules is important in determining the validity and legal effect of a challenged rule.
*411  While the distinctions are sometimes blurred and rules often serve two or more purposes simultaneously, agency rules may be grouped into three general categories: procedural rules, interpretative rules, and legislative rules. 1 F. Cooper, State Administrative Law 173 (1965); Daye, supra at 851-53.
(1) Procedural rules are those which describe how the agency will discharge its assigned functions and the requirements others must follow in dealing with the agency. These are the fundamental rules of agency procedures and are essential to efficient agency operation. Generally these rules deal with such matters as forms, instructions and availability for public inspection of all agency rules and policy. See, e.g., G.S. 150A-1KD. Clearly, then, the requirement that data presented in a ratemaking hearing be audited is more than a procedural rule.
(2) Legislative rules are those established by an agency as a result of a delegation of legislative power to the agency. “Legislative rules fill the interstices of statutes. They go beyond mere interpretation of statutory language or application of such language and within statutory limits set down additional substantive requirements.” Daye, supra at 852-53.
(3) Interpretative rules have been defined as
those that interpret and apply the provisions of the statute under which the agency operates. No sanction attaches to the violation of an interpretative rule as such; the sanction attaches to the violation of the statute, which the rule merely interprets. Thus, for example, most of the regulations of the Internal Revenue Service are interpretative.
1 Cooper, supra at 174-75.
The crucial determination to be made here is whether the Commissioner’s conclusion that data be audited is a legislative or interpretative rule. This is so because interpretative rules and general policy statements of agencies are excluded from the NCAPA rulemaking provisions by G.S. 150A-KH6) and statements of policy or interpretations made in the decision of a contested case are excluded by G.S. 150A-10(4). On the other hand, substantive legislative rules are not excluded from the NCAPA, unless one of the other exclusions applies. We note that none of the remaining exclusions is applicable here.
*412The Commissioner contends that the auditing requirement is interpretative and therefore within the stated exclusions. However, we are not limited to the label placed on a rule by an agency, but must look instead to the substance of the rule in question. Lewis-Mota v. Secretary of Labor, 469 F. 2d 478 (2d Cir. 1972); Pharmaceutical Manufacturers Association v. Finch, 307 F. Supp. 858 (D. Del. 1970); Gibson Wine Company v. Snyder, 194 F. 2d 329 (D.C. Cir. 1952). As Professor Daye stated in his helpful article analyzing the NCAPA: “It should be emphasized that careful scrutiny of the substance of the rule in question is critical, since the interpretative-rule exclusion, if not confined to proper boundaries, could well subsume the rulemaking provisions.” Daye, supra at 853.2
 In applying the stated definitions to the record before us, we conclude that the Commissioner’s requirement of audited data amounts to a legislative rule and is therefore subject to the rulemaking provisions of the NCAPA. We are so persuaded because his new requirement clearly goes beyond a mere interpretation of the statute under which the agency he heads operates and sets up new substantive requirements. One has only to read the lengthy and learned briefs of appellants and amici curiae to know this is true. Furthermore, unlike an interpretative rule, this is certainly a rule with sanctions. Indeed, the Commissioner has dramatized the sanction for violation of his auditing rule: He has denied the requested rate increase for failure of appellants to comply with his newly established rule.
Put another way, the Commissioner’s enunciated rule was established as a result of a delegation of legislative power to his agency. G.S. 58-9(1) empowers the Commissioner to “make rules and regulations ... to enforce, carry out and make effective the provisions of this Chapter, and to make such further rules and regulations not contrary to any provision of this Chapter. . . . The Commissioner may likewise, from time to time, withdraw, *413modify or amend any such regulation.” Hence, the Commissioner’s rule here is clearly legislative in nature. It fills “the interstices of the statute,” and within the statutory limits, it sets down “additional substantive requirements.”
 Our holding that the Commissioner’s auditing requirement is tantamount to a legislative rule and therefore not excluded from the NCAPA is not, however, dispositive of the issue. The Commissioner correctly argues that a second mode by which administrative agencies can establish rules is through the case-by-case process of administrative adjudication. He relies primarily on the following language in the landmark case of Securities & Exchange Commission v. Chenery Corporation, 332 U.S. 194, 67 S.Ct. 1575, 91 L.Ed. 1995 (1947):
To hold that the Commission had no alternative in this proceeding but to approve the proposed transaction, while formulating any general rules it might desire for use in future cases of this nature, would be to stultify the administrative process. That we refuse to do.
* * *
There is thus a very definite place for case-by-case evolution of statutory standards. And the choice made between proceeding by general rule or by individual, ad hoc litigation is one that lies primarily in the informed discretion of the administrative agency.
The scope of our review of an administrative order wherein a new principle is announced and applied is no different from that which pertains to ordinary administrative action. The wisdom of the principle adopted is none of our concern [citations omitted]. Our duty is at an end when it becomes evident that the Commission’s action is based upon substantial evidence and is consistent with the authority granted by Congress. [Citations omitted.]
Id. at 207, 67 S.Ct. at 1582, 91 L.Ed. at 2004-05.
Clearly, the consequences of the choice between general rulemaking and ad hoc, case-by-case adjudication is of enormous significance. 1 Cooper, supra at 177-78. As noted by one com*414mentator, the “whole tenor” of APA procedures is different when establishing rules in the adjudication of contested cases, rather than following rulemaking procedures:
(1) The type of notice is different.
(2) The form of hearing is different.
(3) The mechanics of decision-making are different.
(4) The scope of judicial review is different.
(5) Most importantly, APA-established rules are normally prospective in operation, while decisions in adjudicatory matters are normally (like judicial decisions) retroactive. Id.
The discretion vested in administrative agencies in choosing between the two methods of establishing rules is not, however, unbridled. Indeed, the U.S. Supreme Court in Chenery provided qualifying guidelines in stating the quoted general rules. Ad hoc rulemaking in adjudication is necessary where:
problems may arise in a case which the administrative agency could not reasonably foresee, problems which must be solved despite the absence of a relevant general rule. Or the agency may not have had sufficient experience with a particular problem to warrant rigidifying its tentative judgment into a hard and fast rule.
Or the problem may be so specialized and varying in nature as to be impossible of capture within the boundaries of a general rule.
332 U.S. at 202-03, 67 S.Ct. at 1580, 91 L.Ed. at 2002.
Applying the foregoing to the record before us, we note: (1) the lack of unaudited data was not a problem unforeseen by the Commissioner, (2) “absence of a relevant general rule was not prohibitive of this ratemaking,” (3) here, the Commissioner had “sufficient experience” with the problem, and (4) certainly the problem of auditing is not so specialized and varying in nature as to be “impossible of capture within the boundaries of a general rule.” Indeed, with respect to the latter, one of the problems with the Commissioner’s sudden order to audit data was its vagueness, a problem which could have been avoided had the rule been promulgated in the orderly NCAPA process.
*415The Chenery Court also added:
Since the Commission, unlike a court, does have the ability to make new law prospectively through the exercise of its rule-making powers, it has less reason to rely upon ad hoc adjudication to formulate new standards of conduct within the framework of the Holding Company Act. The function of filling in the interstices of the Act should be performed, as much as possible, through this quasi-legislative promulgation of rules to be applied in the future.
332 U.S. at 202, 67 S.Ct. at 1580, 91 L.Ed. at 2002.
Decisions by the U.S. Supreme Court subsequent to Chenery have been less than helpful. For example, on the question whether an administrative agency can, through adjudication, overrule its prior clear rules when private parties have acted in reliance on the overruled decisions, NLRB v. Wyman-Gordon Company, 394 U.S. 759, 89 S.Ct. 1426, 22 L.Ed. 2d 709 (1969), goes in one direction while NLRB v. Bell Aerospace Company, 416 U.S. 267, 94 S.Ct. 1757, 40 L.Ed. 2d 134 (1974), goes in the opposite direction. Moreover, the Court has held that an agency, even when it had opened the way by first adopting an interpretative rule, could make law only through a legislative rule and not through ad hoc decisions based on the interpretative rule. Morton v. Ruiz, 415 U.S. 199, 94 S.Ct. 1055, 39 L.Ed. 2d 270 (1974). Yet, just two months later, in NLRB v. Bell Aerospace Company, supra, the Court unanimously held that the NLRB, even without first issuing an interpretative rule, could make new law in an adjudication. It has been stated that the Morton v. Ruiz decision was clearly excessive, though “in the right direction.” 2 K. Davis, Administrative Law Treatise § 7.27 at 140 (2d ed. 1979).3
We think the superior rule was stated by Professor Cooper some fifteen years ago and generally adopted by numerous court decisions since:
The general rule that should guide the agencies in making the choice between rule making and ad hoc adjudication might be formulated as follows: where an agency faces the *416alternative of proceeding by rule making or by adjudication, the process of rule making should be utilized except in cases where there is a danger that its utilization would frustrate the effective accomplishment of the agency’s functions. Where such danger exists, e.g., where the “agency may not have had sufficient experience with a particular problem to warrant rigidifying its tentative judgment into a hard and fast rule,” or where the problem is so “specialized and varying in nature as to be impossible of capture within the boundaries of a general rule,” the advantages to the agency of utilizing the ad hoc adjudication technique must be balanced against the possible deleterious public consequences resulting from the retroactive application of a new standard of general application to large numbers of parties who have had no opportunity to be heard as to what the standard should be. Unless the balance clearly preponderates in favor of the ad hoc adjudication method, the agency should utilize rule-making procedures.
The suggestion was well phrased in an A.B.A. committee report which recommended that:
Administrative agencies shall (1) as a fixed policy prefer and encourage rule making to reduce to the minimum the necessity for case-by-case administrative adjudications; and . . . (4) shall promptly formulate, incorporate and promulgate as a rule or statement of policy any and all general principles, not otherwise published as rules or specified in statutes, enumerated in any specific case decision.
More specifically, it has been well suggested that while the practice of working out policy piecemeal by ad hoc adjudication may be justified in the initial stages of administrative regulation of a new field, yet when time and experience have served to sharpen and focus the problems involved, then the agency should utilize rule-making procedures to lay down general rules for the future guidance of all parties affected.
1 Cooper, supra at 181-82 (footnotes omitted).
For decisions in accord with the stated rule, see generally, NLRB v. E. & B. Brewing Company, 276 F. 2d 594 (6th Cir. 1960), *417 cert. denied, 366 U.S. 908 (1961); NLRB v. Guy F. Atkinson Company, 195 F. 2d 141 (9th Cir. 1952); Gonzalez v. Freeman, 334 F. 2d 570 (D.C. Cir. 1964); Harnett v. Board of Zoning, Subdivision and Building Appeals, 350 F. Supp. 1159 (D.V.I. 1972).
We think the policy favoring rulemaking rather than ad hoc adjudication comports with the intent of our Legislature in enacting G.S. 150A-10. The exclusion of policy statements or interpretations “made in the decision of a contested case” included in G.S. 150A-10(4) clearly was not intended to embrace substantive rules with anticipated future applicability. This is so because of the difference between interpretative and legislative rules discussed above and because G.S. 150A-10(6) which excludes “interpretative rules and general statements of policy of the agency” would be unnecessary if G.S. 150A-10(4) were intended to apply to matters beyond the contested case in question. Professor Daye has correctly analyzed the exclusion:
[I]t would appear that if the agency, based on the result in a contested case, desired to promulgate a general rule to govern a matter in the future based on a given set of facts, the promulgation would constitute a rule subject to rulemak-ing requirements unless within another exclusion.
Daye, supra at 851, note 84.
The rationale for the rule we adopt has been stated as follows:
Rule-making provides the agency with a forum for soliciting the informed views of those affected in industry and labor before adopting a new policy. Giving the agency discretion to embark on the new course in an adjudication limits the views presented to those of the parties in the particular case.
. . . Chenery [supra] may allow adjudication as a vehicle for formulation of new agency policy. But the same license should not exist where the new policy revolutionizes long-established patterns of conduct. Where those affected have justifiably relied upon an agency-engendered belief in an established policy, the agency should not be permitted to change the policy except through rule-making. An agency decision branding as “unfair” the conduct always previously *418stamped “fair” should raise judicial hackles sufficiently to lead the court to refuse to follow Chenery and order the agency to engage in rule-making.
B. Schwartz, Administrative Law § 66 at 189-90 (1976).
 Applying the stated rule to the record before us, we first note that the Commissioner clearly intended for the auditing requirement contemplated in his order to apply both retroactively to the case at bar and prospectively to future filings. This is apparent from his finding of fact No. 32 which prescribes the minimum reasonable audit features “to be performed by ISO, NAII and the Bureau.” (Emphasis added.) Moreover, he rejected another automobile insurance rate filing on the same grounds only seven months after this filing. State ex rel. Commissioner of Insurance v. North Carolina Rate Bureau, 41 N.C. App. 327, 255 S.E. 2d 567 (1979), on appeal to this Court and decided today as No'. 86, and in other subsequent filings, see State ex rel. Commissioner of Insurance v. North Carolina Rate Bureau, 44 N.C. App. 191, 261 S.E. 2d 671 (1979), decided by this Court today as No. 54; State ex rel. Commissioner of Insurance v. North Carolina Rate Bureau, 44 N.C. App. 75, 259 S.E. 2d 926 (1979), decided by this Court today as No. 74.
Moreover, we find that in attempting to establish the auditing requirement the Commissioner’s following of normal NCAPA rulemaking requirements would have presented no “danger that . . . utilization [of the NCAPA] would frustrate the effective accomplishments of the agency’s functions.” In this connection, we note:
(1) This is not a situation where the Commissioner “may not have had sufficient experience with a particular problem” to warrant an NCAPA established rule. Indeed, the record discloses that the Commissioner intended to establish a “hard and fast rule.”
(2) The rules was not “so specialized and varying in nature as to be impossible of capture within the boundaries of a general rule.” Present regulations filed by the Commissioner pursuant to the NCAPA are easily adaptable to accomplish the Commissioner’s desired goal. See 11 NCAPA 10.301.
*419(3) In balancing the advantages of the Commissioner’s application of the ad hoc technique against the possible “deleterious public consequences resulting from the retroactive application of a new standard of general application to a large number of parties who have had no opportunity to be heard as to what the standard should be,” we think the scales tip in favor of appellants. The record discloses that the new requirement would be far reaching. It would instantly require a change in long-established procedure in this State and one utilized in practically every other state in the nation. No attempt has been made to determine the ultimate cost of the new requirement, an expense we suspect would ultimately be borne by rate payers in one way or another. No attempt was made to determine if the order was even capable of performance. For example, the audit requirement would obviously require an examination of “original source documents” of the many member groups reporting to the Rate Bureau. No attempt was made to determine where such records are kept by the national companies involved, whether the required information could possibly be retrieved within the time limits required by statute in rate filings or for what period of time such records are or should be maintained. These and other critical questions could properly be answered at a rulemaking hearing held pursuant to the NCAPA. We think it the only orderly and legally proper way to approach the promulgation of a rule so far reaching as that the Commissioner seeks to establish.
In summary, we hold that the practical operation of the Commissioner’s change of policy, when incorporated in the order now before us, is to work hardship upon appellants altogether out of proportion to the public ends to be accomplished. The inequity of such an impact of policy upon appellants presents a striking example of the very reason for the enactment of administrative procedure acts across the land. The Commissioner has ample ways of instituting, through the Legislature or pursuant to the NCAPA, rules he deems essential for the proper discharge of his duties.
We therefore hold that the Commissioner’s attempt to establish a rule requiring audited data in this ratemaking hearing was “made upon unlawful procedure” as contemplated by G.S. 58-9.6(b)(3) and G.S. 150A-5K3).
*4203. Arbitrary and Capricious Actions
 We next address the question whether the Commissioner’s action ordering audited data was “arbitrary and capricious” as contemplated by G.S. 58-9.6(b)(6) and G.S. 150A-51(6).
Agency decisions have been found arbitrary and capricious, inter alia, when such decisions are “whimsical” because they indicate a lack of fair and careful consideration; when they fail to indicate “any course of reasoning and the exercise of judgment,” Board of Education v. Phillips, 264 Ala. 603, 89 So. 2d 96 (1956), or when they impose or omit procedural requirements that result in manifest unfairness in the circumstances though within the letter of statutory requirements, 2 Cooper, supra at 761-69, note 8, and cases cited therein. “The ultimate purpose of rulemaking review is to insure ‘reasoned decisionmaking’ . . . .” Daye, supra at 922, citing Verkuil, Judicial Review of Informal Rulemaking, 60 Va. L. Rev. 185, 230 (1974).
We agree with appellants that the Commissioner’s order with respect to audited data is arbitrary and capricious for these reasons: The order is vague and uncertain in that (1) it does not establish the extent to which examination of “original source documents” is required, (2) it does not make clear whether the auditing must be performed by Certified Public Accountants, other accountants, or actuaries, (3) it does not specify the degree of precision and reliability required of “statistical sampling,” (4) it generally does not provide appellants with adequate guidelines for compliance with the general conclusion that data in a ratemak-ing hearing be audited, (5) it includes no determination by the Commissioner as to the possibility of performance of his new rule nor whether implementation of the rule would be economically feasible, (6) it includes no determination whether the statutory time limits could be complied with in face of the new rule, and (7) it includes no determination whether the “original source data” contemplated by the new rule is even available for the past years involved in this filing or whether such data, if available, is located in North Carolina or outside the State in the case of the several hundred companies writing insurance in this State.
In view of these omissions we hold the Commissioner’s order is grossly imprecise in attempting to enunciate a substantive rule *421involving sweeping ramifications and is therefore “arbitrary or capricious” as contemplated by G.S. 58-9.6(b)(6) and G.S. 150A-5K6).
The Court of Appeals affirmed the Commissioner’s conclusion that unaudited data is not a credible basis for justifying a proposed rate increase on the sole ground that there was “material and substantial evidence in view of the entire record as submitted” as contemplated by G.S. 58-9.6(b)(5) to support the Commissioner’s order. Had this been the only criterion for appellate review, the Court of Appeals’ decision would be correct as we have noted above. See Section I.A. above. That court erred, however, in failing to review the Commissioner’s order in light of subsections (3) and (6) of G.S. 58-9.6(b) and G.S. 150A-51, as we have done above. Therefore, and for the reasons stated above, this portion of the Court of Appeals’ decision is reversed.
We declare null and void all portions of the Commissioner’s order referring to the requirement of audited data, including but not limited to, findings of fact 22 through 35, conclusions of law 1 through 4 and usage of the phrase “purported to show” in all portions of the order wherein the phrase was obviously inserted to question otherwise undisputed and uncontradicted evidence but for the lack of formal audit.
Finally, we note that the issue presented here is not an isolated one. The proliferation of administrative agencies throughout the last several decades, both in federal and state governments, has created controversy and confusion over the question of proper legislative delegation of authority and the appropriateness of standards to guide effective agency action. The various treatises cited in this opinion are replete with citations to decisions from state and federal courts adopting practically every position imaginable. We agree generally that:
Power should be delegated [to an administrative body] where there is agreement that a task must be performed and it cannot be effectively performed by the legislature without the assistance of a delegate or without an expenditure of time so *422great as to lead to the neglect of equally important business. Delegation is most commonly indicated where the relations to be regulated are highly technical or where their regulation requires a course of continuous decision.
Where not only technical skill but continuous judgment is demanded the legislature is helpless. This is true of rate regulation which requires a vast number of individual determinations, a body of technical material, and an expert staff. Decisions must make a pattern, integrated yet flexible. [The legislature] could not frame a delegation which would settle all vital questions of policy.
Jaffe, Judicial Control of Administrative Action 35, 37 (1965).
The North Carolina General Assembly has effectively and properly delegated insurance ratemaking to the Rate Bureau with review by the Commissioner of Insurance. It can, and perhaps should, review the statutes with the view to providing clarity on such significant substantive matters as that presented here. In the meantime, it is incumbent on the Commissioner, in discharging the broad powers he possesses as head of a major State administrative agency, to follow the clear lawful procedures prescribed by our Legislature to guide all administrative agencies.
North Carolina Reinsurance Facility
In its filing letter of 29 November 1977 to the Commissioner, the Rate Bureau stated, “This filing proposes also that the rates for risks ceded to the North Carolina Reinsurance Facility be 10°/o higher than the proposed rates for risks voluntarily retained, subject to all applicable provisions of law.”
Based on several findings of fact, the Commissioner concluded, inter alia, as follows:
(20) That as the filing does not propose any fixed set of objective criteria for deciding which risks may be ceded to the Facility, the decision to cede a given risk is based entirely on the subjective judgment of the individual insurer.
*423(21) That the exercise of subjective judgment regarding cessions by insurers has resulted in a Facility population in which 475,704 (86.9%) of the ceded exposures have not caused a claim payment to be made, 389,111 (71.0%) have never been assessed SDIP points, and 341,273 (62.3%) have neither been assessed any SDIP points nor caused a claim payment to be made.
(22) That in view of the current composition of the Facility, a 10% increase in the Facility Rate [sic] above the rates for voluntary business would be excessive and unfairly discriminatory.
(23) That because acquisition and service costs are charged and accounted for as a percentage of premium, a Facility rate 10% higher than the proposed rates for insureds voluntarily retained will result in ceded risks paying disproportionately higher acquisition and service costs, and that the higher Facility rate is therefore excessive and unfairly discriminatory.
(24) That were the proposed 10% higher Facility rate approved, any increase in the percentage of total insureds ceded to the Facility would result in an overall rate level increase in excess of 6%, which is in contravention of G.S. 58-124.26.
The Court of Appeals agreed with the Commissioner, stating,
[T]here appears in the record material and substantial evidence to support the Commissioner’s finding of fact. The figures and percentages are drawn directly from the Bureau’s Exhibit # RB 33. And based upon the finding that there are no objective criteria for cession to the Facility, and the Commissioner’s finding that 62.3% of the insureds ceded to the Facility have neither assessed any SDIP points nor caused a claim payment to be made, we find that there is ample support for the Commissioner’s conclusion that a 10% rate increase for insureds in the Facility would be unfairly discriminatory.
41 N.C. App. at 320-21, 255 S.E. 2d at 564.
*424By their second assignment of error, appellants contend that the Court of Appeals erred in approving the findings and the conclusions of the Commissioner that the reinsurance rate differential is unfairly discriminatory.
This assignment of error presents to this Court the first substantial challenge to the North Carolina Motor Vehicle Reinsurance Facility created in 1973 by the General Assembly to replace the Assigned Risk Plan. Basically, the Facility represents a pool which insures risks which companies determine they do not want to individually insure. A review of the significant provisions of Article 25A, Chapter 58, N.C. General Statutes, will prove helpful to our disposition of this assignment of error.
G.S. 58-248.27 created the North Carolina Motor Vehicle Reinsurance Facility (“Facility”) in 1973 as a nonprofit entity to consist of all insurers licensed to write motor vehicle insurance in the State. All insurers within the State are required to be members of the Facility and to be bound by its rules of operation which are determined by the statute or promulgated by its board of governors. G.S. 58-248.31(a) is particularly significant. It provides that all insurers “as a prerequisite to the further engaging in this State in the writing of motor vehicle insurance . . . shall accept and insure any otherwise unacceptable applicant therefor who is an eligible risk if cession of the particular coverage and coverage limits applied for are permitted in the Facility." (Emphasis added.)
This statute also provides that all insurers “shall equitably share the results of such otherwise unacceptable business through the Facility” and that each company shall be bound by the acts of its agents in accordance with the provisions of the Article. Id.
G.S. 58-248.32(a) provides in part that no licensed agent of an insurer shall refuse to accept any application from an eligible risk for such insurance and to immediately bind the coverage applied for if cession of the particular coverage and limits are permitted in the Facility. The 1977 Legislature added a provision to this statute providing that agents shall write the coverage applied for at what the agent believes to be the appropriate rate level. G.S. 58-248.32(b).
*425G.S. 58-248.33 defines the Facility’s functions and administration. It first provides that the Facility shall assure the availability of motor vehicle insurance to any “eligible risk” and that the Facility shall accept all placements made in accordance with the Article. G.S. 58-248.33(a). It then sets forth the minimum coverage provisions for which the Facility shall provide reinsurance.
Subsections (d), (e), (f) and (g) of G.S. 58-248.33 spell out the composition, responsibilities and powers of the Facility’s board of governors.
The 1977 Legislature amended G.S. 58-248.33 by adding, inter alia, subsections (1) and (m). Former subsection (1), under which this proceeding occurred, provided in pertinent part:
The classifications, rules, rates, rating plans and policy forms used on motor vehicle insurance policies reinsured by the Facility may be made by the Facility by any licensed or statutory rating organization or bureau on its behalf and shall be filed with the Commissioner. The Commissioner may establish separate sub classifications within the Facility for clean risks as defined by the Commissioner. . . . Rates shall be neither excessive, inadequate nor unfairly discriminatory. ... If the Commissioner finds, after a hearing, that a rate is either excessive, inadequate or unfairly discriminatory, he shall issue an order specifying in what respect it is deficient and stating when, within a reasonable period thereafter, such rate shall be deemed no longer effective. Said order is subject to judicial review as set out in Article 2 of this Chapter. Pending judicial review of said order, the filed classification plan and the filed rates may be used, charged and collected in the same manner as set out in G.S. 58-131.42 of this Chapter. . . . All rates shall be on an actuarially sound basis and shall be calculated, insofar as is possible, to produce neither a profit nor a loss. . . . Rates shall not include any factor for underwriting profit on Facility business, but shall provide an allowance for contingencies. There shall be a strong presumption that the rates and premiums for the business of the Facility are neither unreasonable nor excessive.4
G.S. 58-248.34 sets forth the requirements for the Facility’s “plan of operation.”
G.S. 58-248.35 provides that, “Upon receipt by the company of a risk which it does not elect to retain, the company shall follow such procedures for ceding the risk as are established by the plan of operation.” (Emphasis added.)
The 1977 amendments applicable to this case, and incorporated in the statutory summary noted above, made several significant changes to the original 1973 legislation. The most significant was the establishment of the procedures designed to make the Facility self-sustaining. Prior to 1977, a high-risk insured whose coverage was ceded to the Facility paid the same amount for insurance as the high-risk insured whose coverage was not ceded. Law of March 6, 1945, Ch. 381, s. 2, 1945 N.C. Sess. Laws 461, formerly G.S. 58-248.2 (1975) (repealed 1977). Under the 1977 statute, losses sustained by the Facility are to be recouped “either through surcharging persons reinsured by the Facility or by equitable pro rata assessment of member companies.” G.S. 58-248.34(e). The member companies, in turn, are to recoup any such investment by surcharging policyholders. G.S. 58-248.34(f). This surcharge is to be assessed “on motor vehicle insurance policies issued by the member or through the Facility.” Id. Conversely, should the Facility realize any gain, any balance remaining after losses are covered is to be distributed to persons reinsured by the Facility. G.S. 58-248.34(e).
In commenting on the 1977 changes, the Legislative Research Commission’s report to the 1979 General Assembly stated:
Under the old law the participating company could not transfer more than 50% of their risks to the Facility, had to share Facility losses, and could not charge higher rates for automobile liability policies ceded to the Facility. House Bill 658 [1977 revision of insurance law] changed all of that by eliminating the 50% limitation on cessions, by permitting *427higher rates or surcharges to recover losses of the Facility, and by providing for distribution of Facility gains to policyholders reinsured by the Facility. The apparent intent behind the new provisions is to make the Facility self-sustaining, whereas under the old system the insurance industry in effect subsidized the Facility by absorbing its losses.
Legislative Research Commission, Report to the 1979 General Assembly of North Carolina 12-13 (1979).
That same report also stated:
House Bill 658 provided for a “clean risk” subclassification in the Facility (those drivers without points for the previous three years whose policies were ceded to the Facility), to be defined by the Commissioner. In his supplemental order of November 30, 1978, the Commissioner directed the Rate Bureau to submit a plan whereby no driver in the Facility would be surcharged more than a driver outside the Facility if they had the same number of driving record points or chargeable accidents. This was coupled with his October 30, 1978, order to eliminate the separate Facility rate in the classification plan submitted earlier, and was intended to compensate for any revenue shortfalls resulting from that elimination. Both orders have been appealed. It is arguable as to whether or not the Commissioner’s orders come within the letter or intent of the new provisions, but deference must be made to the courts for judgment on this matter. There is, however, implication in the language of G.S. 58-248.34(e) that the surcharge does not necessarily have to apply exclusively to drivers whose policies are ceded to the Facility.
Id. at 47-48.
A. Scope of Review
Turning to appellants’ contentions under this assignment that the Commissioner’s order disapproving a 10% surcharge on Facility policyholders should be voided, we first note that we have not been cited to any of the standards for judicial review of insurance actions in either G.S. 58-9.6 or G.S. 150A-51. We must therefore first determine the appropriate scope of review for this assignment of error.
*428The thrust of the arguments presented by both appellants and the Commissioner is concerned with the evidence presented at hearing on this issue. We think the appropriate subsection of our judicial review statutes thus invoked is that which calls for appropriate judicial action when the Commissioner’s findings and conclusions are “[unsupported by material and substantial evidence in view of the entire record as submitted,” G.S. 58-9.6(b) (5) or are “[unsupported by substantial evidence in view of the entire record as submitted,” G.S. 150A-5H5). However, because the Commissioner’s brief relies on certain cases from other jurisdictions involving constitutional determinations and because the phrase “unfairly discriminatory” carries constitutional implications, we first explain our decision not to consider this assignment of error on constitutional grounds.
 It is well established in this jurisdiction that the constitutionality of a statute will not be reviewed in the appellate court unless it was raised and passed upon in the proceedings below, City of Durham v. Manson, 285 N.C. 741, 208 S.E. 2d 662 (1974), usually by the trial court. “[W]e will not pass upon a constitutional question unless it affirmatively appears that such question was raised and passed upon in the court below.” State v. Dorsett & Yow, 272 N.C. 227, 229, 158 S.E. 2d 15, 17 (1967) (emphasis in the original). In State v. Cumber, 280 N.C. 127, 185 S.E. 2d 141 (1971), the constitutional question was not raised in the trial court but for the first time in the Court of Appeals. This Court held that it was not properly before the Court of Appeals nor this Court. We stated:
That belated constitutional question was injected for the first time on appeal to the Court of Appeals and therefore came too late. It was not properly before that court and is not now properly before us. “The attempt to smuggle in new questions is not approved. Irvine v. California, 347 U.S. 128, 129. Appellate courts will not ordinarily pass upon a constitutional question unless it affirmatively appears that such question was raised and passed upon in the trial court. State v. Jones, 242 N.C. 563, 564, 89 S.E. 2d 129. This is in accord with the decisions of the Supreme Court of the United States. Edelman v. California, 344 U.S. 357, 358.” State v. Grundler, 251 N.C. 177, 111 S.E. 2d 1 (1959). Accord, State v. Colson, 274 N.C. 295, 163 S.E. 2d 376 (1968).
Here, the Commissioner’s original order denying the Reinsurance Facility rate increase stated only that such rates are “unfairly discriminatory” presumably in the statutory sense. He never held that any of the statutes or actions were unconstitutional. In his brief, however, he does make vague assertions that it would be “constitutionally suspect” to interpret the statutes contrary to his findings and conclusions. He states, “The governing statutes should be construed so as to avoid serious doubts as to constitutionality.” Moreover, the Commissioner relies strongly on a recent holding of the Supreme Court of Michigan, Shavers v. Attorney General Kelley, 402 Mich. 554, 267 N.W. 2d 72 (1978), cert. denied, 442 U.S. 934 (1979). There, it was held that Michigan’s No-Fault Insurance Act was constitutional insofar as it provided benefits as a substitute for tort remedies it partially abolished. However, certain ratemaking mechanisms were constitutionally deficient in failing to provide due process. That court delineated several deficiencies of the Michigan statute, similar to deficiencies alleged here. However, the Michigan court unquestionably based its holding on constitutional due process considerations. Indeed, the Michigan action was a declaratory judgment action specifically brought to determine the constitutionality of the Michigan No-Fault Insurance Act. The constitutional question was the basis for the action from trial court to final appellate adjudication. This is completely unlike the case before us where the record discloses no constitutional question presented or passed on in the Commissioner’s original order.
Moreover, our judicial review statutes do not contemplate constitutional review in the present posture of the matter before us. G.S. 58-9.6(b)(l) provides essentially that the Commissioner’s findings and conclusions may be affirmed, reversed, modified, etc. if the substantial rights of the appellants have been prejudiced “in violation of constitutional provisions,” and G.S. 150A-5HU provides the same standard if the appellants’ rights “may have been prejudiced.” Here, appellants, the Rate Bureau and member companies make no assertion that their rights have been prejudiced because any of the Commissioner’s findings or conclusions were in violation of any constitutional provisions. This is only the belated *430argument of the Commissioner. Accordingly, we think no constitutional issues are before us.
B. Material and Substantial Evidence
We next address the question whether the Commissioner’s findings and conclusions with respect to this portion of his order were “unsupported by material and substantial evidence in view of the entire record as submitted” as contemplated by G.S. 58-9.6(b)(5) and G.S. 150A-5U5). (Emphasis added.)
We reiterate the rule stated in Section II. A. 1 of this opinion that it is for the administrative agency to determine the weight and sufficiency of the evidence and the credibility of the witnesses, to draw inferences from the facts, and to appraise conflicting and circumstantial evidence. 73 C.J.S., Public Administrative Bodies and Procedure, supra at § 126. It is not our function to substitute our judgment for that of the Commissioner when the evidence is conflicting. However, as also indicated above, when evidence is conflicting, the standard for judicial review of administrative decisions in North Carolina is that of the “whole record” test. Thompson v. Wake County Board of Education, supra; In re Rogers, supra. As Justice Exum stated in In re Rogers: “The ‘whole record’ test is not a tool of judicial intrusion; instead, it merely gives a reviewing court the capability to determine whether an administrative decision has a rational basis in the evidence. See Jaffe, Judicial Control of Administrative Action . . . 601 [(1965)]; Daye, supra at 920-921.” 297 N.C. at 65, 253 S.E. 2d at 922.
In Thompson v. Wake County, Justice Copeland clearly explained the “whole record” test:
This standard of judicial review is known as the “whole record” test and must be distinguished from both de novo review and the “any competent evidence” standard of review. Universal Camera Corp. v. N.L.R.B., 340 U.S. 474, 95 L.Ed. 456, 71 S.Ct. 456 (1951); Underwood v. Board of Alcoholic Control, 278 N.C. 623, 181 S.E. 2d 1 (1971); Hanft, Some Aspects of Evidence in Adjudication by Administrative Agencies in North Carolina, 49 N.C.L. Rev. 635, 668-74 (1971); Hanft, Administrative Law, 45 N.C.L. Rev. 816, 816-19 (1967). The “whole record” test does not allow the *431reviewing court to replace the Board’s judgment as between two reasonably conflicting views, even though the court could justifiably have reached a different result had the matter been before it de novo, Universal Camera Corp., supra. On the other hand, the “whole record” rule requires the court, in determining the substantiality of evidence supporting the Board’s decision, to take into account whatever in the record fairly detracts from the weight of the Board’s evidence. Under the whole evidence rule, the court may not consider the evidence which in and of itself justifies the Board’s result, without taking into account contradictory evidence or evidence from which conflicting inferences could be drawn. Universal Camera Corp., supra.
292 N.C. at 410, 233 S.E. 2d at 541.
 Applying the foregoing, we proceed to review the evidence to determine whether it is substantial in view of the entire record to support the Commissioner’s findings and conclusions. We first note that G.S. 58-248.33(1) which prohibits rates which are “excessive, inadequate or unfairly discriminatory” contains no definition of the latter phrase. That same statute provides, however, that “[ajl rates shall be on an actuarially sound basis.” Id. Moreover, it provides that “[t]here shall be a strong presumption that the rates and premiums for the business of the Facility are neither unreasonable nor excessive.” Id. While the Commissioner’s findings with respect to the numbers and percentages of ceded exposures which had not had assessed any SDIP points nor caused claim payments to be made are supported by the record, the Commissioner has made no findings with respect to the statutory standard “actuarially sound.” This is so even though the record is replete with evidence indicating that the proposed differential for the rate increase between ceded and voluntary business is actuarially justified. For example, the evidence for the Rate Bureau indicated there were only 1.42 claims per hundred cars involving bodily injury for voluntary risk policyholders in contrast with 2.95 accidents per hundred cars involving bodily injury for ceded risk policyholders. For property damage, the corresponding figures for voluntary risk were 5.70 accidents per hundred cars as compared to 10.21 for ceded risk.
The statistics revealed in the following chart indicated the pure premium or average loss per car was also significant:
*432Bodily Property Injury Damage
Voluntary Risks $22.17 $23.85
Ceded Risks 56.12 47.59
In other words, the claim frequency for Facility business for bodily injury was 108% higher than for a voluntary insured. For property damage, the claim frequency was 79% higher for a Facility insured than for a voluntary insured. On an average loss per car basis, the pure premium for a Facility insured for bodily injury liability was 153% higher and for property damage liability the Facility insured had a loss cost of 99.5% higher than a voluntary insured.
The Commissioner’s order also relies heavily on the percentage of insureds who had no SDIP points or claims assessed. The record indicates extensive questioning concerning what are referred to as “clean risks.” As we understand it from the record, parties to the hearing proceeded under the understanding that a “clean risk” was one who had neither SDIP points nor a claim assessed during the preceding three-year period.5 The Commissioner’s findings and conclusions concerning these statistics and the “current composition of the Facility” indicate that he based his conclusions primarily on what he considers unfair discrimination between a “clean risk” in the Facility and those in the voluntary market. Again, however, in following the statutory guide that the rates be “actuarially sound,” the statistics are significant. The following chart from the Rate Bureau’s exhibits indicates the claim frequency per hundred cars:
Bodily Property Injury Damage
Voluntary “Clean Risks” 1.30 5.46
“Clean Risks” in Facility 2.73 9.56
Increased Claim Frequency of Ceded Risks (110% higher) (75% higher)
*433Moreover, the comparable pure premium statistics (average loss per car) indicate:
Bodily Injury Property Damage
Voluntary “Clean Risks” $20.41 $22.52
“Clean Risks” in Facility 51.96 43.93
Increased Average Loss of Ceded Risks (155% higher) (95% higher)
We also note that, while as indicated above, the Commissioner’s alarming statistical finding that 62.3% of those in the Facility had neither SDIP points nor claims paid is supported by the evidence, the record indicates that his statistics were based on the experience for all carriers writing in North Carolina for the accident year ending June 30, 1976. Clearly, these statistics would not reveal the SDIP record or claims made for these insureds for the three-year period which we glean from the record to be contemplated by the definition of a “clean risk.” Apparently, the data relied upon by the Commissioner covers only the one-year period. We assume that data for the preceding accident year is that most commonly relied upon when a filing is made. However, it is obviously inconsistent to show concern about the composition of a Facility with respect to its high percentage of “clean risks” and not include statistics covering a three-year period; while the high percentage of insureds in the Facility might not have had any SDIP points assessed or claims made during the preceding year, this certainly does not mean that these insureds are “clean risks.” Some of them, perhaps many of them, might well not be “clean risks” if the three-year period were considered.
Therefore, on the basis of our review of the entire record, we are compelled to conclude that the Commissioner failed to consider material and substantial evidence concerning the actuarial soundness of the statistics and that the findings which the Commissioner made, while supported by the evidence, are legally irrelevant in light of the limitation to a one-year period. The *434evidence supporting the Commissioner’s findings and conclusions is, therefore, in our view, insubstantial.
C. Statutory Scheme
It is also helpful in addressing the question presented here to analyze the statutory scheme from the provisions set forth in the introductory section to this portion of our opinion. Insurance companies doing business in North Carolina are required to write policies for all qualified applicants with exceptions not pertinent here. A company which has written a policy it does not wish to retain has the absolute right to cede that policy to the Facility. The rates and classifications for the Facility risk are to be made by the Facility and there is a strong presumption that these rates are neither unreasonable nor excessive. Facility rates are to be on an actuarially sound basis and can produce neither a profit nor a loss.
 The Commissioner’s finding and conclusion therefore “that the filing does not propose any fixed set of objective criteria for deciding which risk may be ceded to the Facility” is simply not persuasive. Indeed, the setting of objective criteria by insurance companies would be legally unenforceable; G.S. 58-248.35 allows ceding merely upon the criterion that the company does not elect to retain the business. It is presumed the Legislature acted with reason and common sense and did not require an unjust and absurd result. King v. Baldwin, 276 N.C. 316, 172 S.E. 2d 12 (1970). We think the plain legislative intent is that Facility rates can be higher than those for the voluntary market if a higher Facility rate is actuarially indicated. See G.S. 58-248.32(b). We simply do not believe that the Legislature would require companies to insure all applicants, regardless of the risk they present, direct the Facility to accept unlimited policy cessions from insurance companies, direct the Facility to fix rates on an actuarially sound basis that will produce neither a profit nor a loss, and then provide that the Facility may not do precisely what it was directed to do because such actions would be “unfairly discriminatory.”
We are not inadvertent nor insensitive to the Commissioner’s concern that our statutory scheme allows insurance companies to cede any insured they elect not to retain without any criteria established by law. The answer is, first of all, that such a scheme *435is the obvious legislative intent and we perceive no constitutional attack, as indicated earlier, on the statute itself. The Legislature has obviously elected to leave the establishment of criteria for ceding to the individual companies, anticipating their awareness that the Facility is a non-profit, unincorporated legal entity, G.S. 58-248.27, and not allowed to make a profit. G.S. 58-248.33(1). Since the premium of an insured ceded to the Facility goes to the Facility and there is no profit to the company, the only way the company can possibly make an underwriting profit is by voluntarily retaining the risk. Such incentive, we think our Legislature reasoned, is sufficient to safeguard abuse of the ceding privilege and prevent “unfair discrimination.”
We further note that the 1979 amendments to G.S. 58-248.33(1) provide even more protection to clean risks ceded to the Facility. They can be charged rates no higher than clean risks voluntarily retained.
D. Acquisition and Service Costs
 We next turn to the Commissioner’s findings and conclusion that because acquisition and service costs are charged and accounted for as a percentage of the premium, a Facility rate 10% higher than the proposed rate for insureds voluntarily retained would result in ceded risks paying disproportionately higher acquisition and service costs. We simply find no evidence in the record to support the Commissioner’s bare assertion. The Rate Bureau’s Exhibit No. 1 states the acquisition cost figure for voluntary business to be 19.5%, plus 5% for profit on automobile liability insurance. The acquisition for Facility business, on the other hand, is stated to be 16.8%, with no figure for profit. Moreover, in the only testimony we find on this point, one witness testified that there is more general administration expense for the 12-point driver than for a driver with no points. We are therefore compelled to conclude that the Commissioner’s finding of fact No. 92 and conclusion of law No. 23 are unsupported by any material or substantial evidence.
E. Cap on Rate Increase
 We next turn to the Commissioner’s conclusion that any increase in the total number of insureds in the Facility would increase the overall rate level by more than 6% in contravention of *436G.S. 58-124.26. While this finding and conclusion is mathematically correct, it is erroneous as a matter of law. As noted above, the statutes permit insurers to cede any unwanted business to the Facility. Indeed, the previous limitation for cessions to the Facility at 50% without specific approval of the board of governors was repealed by the 1977 Legislature. At present, there is no limit to the number of insureds who may be ceded to the Facility. Consequently, as appellants correctly note, should the Commissioner approve the overall rate level increase of 6% and thereafter one single additional insured were ceded to the Facility, everything else being equal, an increase in the overall rate level above the 6% cap imposed by G.S. 58-124.26 would result. Under the Commissioner’s conclusion, all rate increases would be impossible to justify since there is no way to ascertain what the total future cessions to the Facility might be and thus whether, at some time in the year, additional cessions to the Facility might push the overall rate level above 6%.
Moreover, if an additional cession to the Facility resulted in an increase in “the general rate level” the same result would follow when other factors not directly related to the ratemaking process cause an increase or decrease in total premium collections. For example, premium variations are established by such factors as the number of SDIP points and territory in which a car is principally garaged. If we adopt the Commissioner's contention here, a general rate increase would occur anytime any insured in the State is convicted of a traffic violation or moved into a territory with a higher risk factor. Such was clearly not the legislative intent.
Construing the applicable statutes in pari materia and interpreting each in a way as would give meaning and effect to each provision and thus carry out the legislative intent, State ex rel. Commissioner of Insurance v. Automobile Rate Bureau Administrative Office, supra, we believe the Legislature did not intend a result impossible to obtain in practical terms, but instead intended that any overall rate increase should be limited to 6% given the same book of business as for the experience period.
The ratemaking process is premised on the underlying assumption that the book of business throughout the period for which rates are to be made will be the same as that which existed during the experience period.
*437Finally, we note again that the Commissioner’s statistical findings, to the extent stated, are supported by the evidence. However, the Rate Bureau’s exhibits tend to establish that while the statistical data in the filing support and justify an overall 11.8% increase in liability insurance rates for voluntary business, the corresponding data with respect to Facility insureds support and justify an overall 63.4% increase for insureds who have been ceded to the Facility. As explained earlier, the evidence clearly supports the assertion that the differential in rates between the voluntary market and the Facility are actuarially justified, the standard established by statute. Again, it is not our task to substitute our judgment for the Commissioner’s where evidence is conflicting. However, under the “whole record” test, which we are bound to apply, we do not merely consider the evidence which in and of itself justifies the Commissioner’s conclusion without taking into account contradictory evidence or evidence from which conflicting inferences could be drawn. Applying the whole record test, as explained above, we hold that the evidence to support the Commissioner’s findings and conclusions that the 10% rate differential was unfairly discriminatory was insubstantial in view of the entire record.
Before leaving this assignment of error, we deem it appropriate to note that until clear guidelines are established either by the Legislature or by the Commissioner, confusion will continue to abound over the phrase “unfair rate discrimination.” The phrase is not defined in our statutes nor, for that matter, in the model laws and is a source of continuing controversy. See S. Kim-ball and H. Denenberg, Insurance, Government and Social Policy 209-242 (1969).
Moreover, those in the profession and the industry are unable to agree when unfair price discrimination exists. As noted earlier in this opinion, for some it connotes constitutional consideration; others consider it in purely economic terms. “A widely accepted economic definition of price discrimination states that unfair price discrimination exists if, allowing for practical limitations, there are price differences that do not correspond to differences in cost or cost differences that are not reflected in price differences.” Kimball and Denenburg, supra at 210. It has been *438othewise stated that “rate equity exists where each insured’s net premium is exactly sufficient to defray the expenses of his expected losses (and loss adjustment expenses). Conversely, rate equity is absent where the premium dollar of some insureds must be used for the payment of losses suffered by other insureds in the same risk category. Applicants for insurance must be classified for purposes of premium computations in order that each applicant need carry only the expected cost of his own coverage. Procaccia & Shafton, Coinsurance Clauses and Rate Equity, Insurance L.J., February 1978 at 69 (No. 661).
We can understand the reluctance to define such a complex term as “unfair discrimination.” However, the vagueness now present will continue, in our opinion, to create severe operational problems for the persons charged with the application of the law. It is obvious that the insurance business, being essentially mutual in character, should provide for all policyholders to be treated fairly with respect to other policyholders. However, the formula for determining what that fairness is should be established by the policymaking body in lieu of reliance on case-by-case adjudications.
It should also be observed that the Commissioner does not contend that the rates charged to those in the Facility are excessive. His order deals only with unfair discrimination. It is apparent, therefore, that the Commissioner’s primary concern in this instance is with the composition of the Facility. Our Legislature, however, has determined that the makeup of the Facility should be determined by the insurers with the protective device that the insurers will not be allowed to make a profit on Facility business. The Commissioner’s recourse, therefore, is to request the Legislature to set the objective criteria for ceding insureds to the Facility which he desires. Moreover, we note that at the time of this proceeding the Commissioner was already authorized to establish separate subclassifications for clean risk in the Facility. The record discloses that he did not do so.
Finally, we think it worthwhile to note recent legislative action pertaining to the Facility.
The 1979 Legislative Research Commission Report to the General Assembly made several recommendations to the 1979 session. It expressly recommended that there should be a statutory *439definition of a “clean risk” subclassification within the Facility, in which subclassification the insureds would pay Facility rates but would not be subject to the Facility surcharge. The Commission stated that it would be unfair for motorists in the “clean risk” classification to subsidize other motorists in the Facility. The Commission also recommended that the Legislature consider the possibility of adding automobile physical damage (collision), theft, and comprehensive insurance coverages to the coverages provided by the Facility.
The 1979 Legislature responded to these recommendations with several changes in Article 25. G.S. 58-248.31 was amended by adding two subsections. These provided essentially that each company will provide the same type of service to ceded business that it provides for its voluntary market. Id. 58-248.31(b). The records of agents and brokers shall indicate that the business is ceded. Id. When an insurer cedes a policy to the Facility and the premium for that policy is higher than the insurer would normally charge for the policy if retained by the insurer, the policyholder shall be informed (1) that his policy is ceded, (2) that the coverages are written at the Facility rate, which rate differential must be specified, (3) what the reason or reasons are for the cession to the Facility, (4) that the specific reason or reasons for his cession to the Facility will be provided upon the written request of the policyholder to the insurer, and (5) that the policyholder may seek insurance through other insurers who may elect not to cede his policy. Id. Upon the written request of a person notified that his policy has been ceded to the Facility, the insurer ceding the policy must provide in writing to the insured the specific reason for the decision to cede. G.S. 58-248.31(e).
The 1979 Legislature also amended G.S. 58-248.33. Subsection (1) of that statute formerly provided that “the Commissioner may establish separate subclassifications within the Facility for clean risk as defined by the Commissioner.” That sentence was deleted and the following language inserted in lieu thereof:
The Board of Governors [of the Reinsurance Facility] shall establish a separate subclassification within the Facility for “clean risks” as herein defined. For the purpose of this Article, a “clean risk” shall be any owner of a motor vehicle classified as a private passenger non-fleet motor vehicle as *440defined under Article 13C of this Chapter if the owner and the principal operator and each licensed operator in the owner’s household have two years’ driving experience and if neither the owner nor any member of his household nor the principal operator had had any chargeable accident or any conviction for a moving traffic violation pursuant to the subclassification plan established by the provisions of G.S. 58-30.4, during the three-year period immediately preceding the date of application for motor vehicle insurance or the date of preparation for a renewal motor vehicle insurance policy.
That subsection was also significantly amended to provide, “However, the rates made by or on behalf of the Facility with respect to ‘clean risks, ’ as defined above, shall not exceed the rates charged ‘clean risks’ who are not reinsured in the Facility.” (Emphasis added.) Finally, the subsection was amended to provide that the “difference between the actual rate charged and the ac-tuarially sound and self-supporting rates for ‘clean risks’ rein-sured in the Facility may be recouped in similar manner as assessments pursuant to G.S. 58-248.34(f).”
Income On Invested Capital
The Commissioner concluded in his order that the proposed rate increase was “excessive to the extent that investment income is not properly taken into account in any of the rate level calculations contained in the filing.” (Emphasis added.) He also concluded that “it has long been recognized that investment income is an integral part of the return on any insurance transaction.”
Appellants contend that the Commissioner’s conclusions were erroneous to the extent that they contemplated a consideration of investment income on invested capital.
Appellants correctly note that the 1979 Legislature amended G.S. 58-124.19(2) to require consideration of investment income on unearned premium and loss reserves in reviewing rate filings. Because investment income on unearned premium and loss reserves were included in this filing and because the Legislature *441has clarified consideration of these items for the future, we are not concerned here with appellants’ investment income on unearned premium and loss reserves. The question before us relates solely to consideration of investment income on invested capital.
On this point, the Court of Appeals held:
[T]his Court has recently decided that investment income may be considered in evaluating the reasonableness of a filing. State ex rel. Comr. of Ins. v. N.C. Rate Bureau, 40 N.C. App. 85, 252 S.E. 2d 811 (1979). It is thus proper for the Commissioner to consider investment earnings on capital invested by insurers in reviewing the rate making formula.
41 N.C. App. at 318, 255 S.E. 2d at 562-63 (emphasis added).
We first note that the Court of Appeals misconstrued its earlier decision. A careful review of the earlier decision reveals that the question was whether the Commissioner might properly consider profits on investment income from unearned premium and loss reserves. The court there made no mention of the propriety of the consideration of income from invested capital and thus reliance on that decision was misplaced.
A. Error of Law
 We therefore turn to the propriety of the Commissioner’s determination that income on invested capital must be considered in a rate increase filing. In addressing this question, we think the applicable statutes of judicial review are G.S. 58-9.6(b)(4) and G.S. 150A-5K4), i.e., whether the Commissioner’s findings and conclusions in this respect were “[a]ffected by other error(s) of law.”
The statute enumerating the factors to be considered in ratemaking, G.S. 58-124.19, quoted supra, referred in 1977 “to a reasonable margin for underwriting profit and to contingencies” and “to dividends, savings or unabsorbed premium deposits allowed or returned by insurers to their policyholders. . . .” We note that the statute at that time made no mention of a consideration of investment income on either unearned premium and loss reserves or invested capital. We therefore review the pertinent case authority in this jurisdiction.
In In re N.C. Fire Insurance Rating Bureau, supra, the question whether investment income on capital should be properly *442considered in a rate hearing was not correctly before the Court. However, in discussing matters to be determined upon remand to the Commissioner in fixing appropriate premium rates, Justice Lake stated:
G.S. 58-131.2 imposes upon the Commissioner the duty of fixing such rates as will produce “a fair and reasonable profit” and no more. In the statutory plan for the regulation of insurance premium rates, there is nothing comparable to the procedure prescribed by G.S. 62-133 for the fixing of rates by public utility companies for their services. The statutes conferring authority upon the Commissioner of Insurance, and directing his use of it, do not use the term “fair return on fair value” of the property devoted to the insurance business in North Carolina. Here, the direction is to prescribe rates which will yield a “reasonable profit.” [Citation omitted.]
The question of consideration of income from invested capital was more directly addressed by Chief Justice Bobbitt in In re North Carolina Automobile Rate Administrative Office, 278 N.C. 302, 180 S.E. 2d 155 (1971), wherein it was said:
In the absence of a legislative formula or standards, the Commissioner has had no alternative but to look at the ratemaking procedures recognized in the industry and in other States. The words “pure cost” and “expense loading” as used, without explanation, in G.S. 58-248, facilitated this course. Thus, the Rate Office and the Commissioner adopted the industry view that the reasonableness of a profit to be allowed to a company writing automobile liability insurance was determinable on the basis of a percentage of the gross premium rather than on the basis of a rate of return on invested capital. Underlying this view is the fact that the required capital assets of a casualty insurance company are primarily reserves to guarantee its ability to discharge its liability rather than for use as working capital in the prosecution of its business. Such a company has no significant inventory of assets which are used and useful in the prosecution of its business. The primary function of such a company is to render a service. It is noted that the 5% of premium allowed for underwriting profit and contingencies in computing the *443rates proposed by the 1969 Filing is the same as that used in preceding filings and is the same as that generally approved in the industry.
Id. at 314, 180 S.E. 2d at 164.
In State ex rel. Commissioner of Insurance v. State ex rel. Attorney General, 16 N.C. App. 724, 193 S.E. 2d 432 (1972), Judge, now Chief Judge, Morris stated, “The statute . . . clearly requires the Commissioner to determine whether the rates charged are adequate to produce a fair and reasonable profit. This, it seems to us, refers to underwriting profit and does not include investment income.” Id. at 728-29, 193 S.E. 2d at 435 (emphasis in original).
The issue was more pointedly addressed in State ex rel. Commissioner of Insurance v. State ex rel. Attorney General, 19 N.C. App. 263, 198 S.E. 2d 575, cert, denied, 284 N.C. 252, 200 S.E. 2d 659 (1973). There, it was said:
This contention [that the Commissioner should have required the presentation of evidence relating to the amount of capital necessary to engage in workers’ compensation insurance business] has been rejected in several North Carolina cases. [Citations omitted.] Evidence of this type is commonly used in fixing utility rates. [Citations omitted.] It is much less relevant in determining insurance rates, because as the Court explained in the Automobile Rate Office case [278 N.C. 302, 180 S.E. 2d 155], an insurance company “has no significant inventory of assets which are used and useful in the prosecution of its business. The primary function of such a company is to render a service.” 278 N.C. at 315, 180 S.E. 2d at 164. Utility companies own large quantities of expensive equipment, which is necessary for them to provide their services. To purchase this equipment, large amounts of capital must be invested; and thus it is possible to determine utility rates by reference to the amount of capital invested in the company and the fair value of its property. Insurance companies, on the other hand, do not require so much costly equipment or so large a capital investment. The importance of the service they provide is not in proportion to the value of their property or the amount of their capital investment. For this reason the courts have determined that proper profit levels for in *444 surance companies may be more appropriately ascertained by taking a percentage of their premiums than by specifying a certain rate of return on their capital investment.
These and other decisions establish clearly that it has never been the law in this jurisdiction that income from invested capital is to be considered in an insurance ratemaking case.
B. The Majority Rule
We also find our view consistent with that prevailing in other jurisdictions. In 2 Couch, Insurance Law § 21:38 at 494 (Anderson ed. 1959) it is said:
In determining whether an insurer has made a reasonable profit, the amount of business done rather than its capital should be considered, and profits should be determined by subtracting losses and expenses from the total of premiums actually received, to the exclusion of profit on capital and surplus, and excess commissions paid to agents but considering interest on unearned premiums and related elements. (Emphases added.)
The law relating to the public service corporations contains features unlike anything in the rating act for insurance companies. The public service corporations are regulated in a way to insure them a reasonable return on their capital invested after defraying expenses. Often they are monopolies having no substantial competition. Insurance companies always have active competition among each other. The Rating Act was intended to remove the temptation to pool in violation of the anti-trust laws and to prevent ruinous competition. The statute contemplates that the rates shall be fixed with a view of the aggregate earnings and profits for the insurance business in the State. Each company may make as much money as it can. Some may make enormous profits, some may do a losing business, but the average profit, that is, the aggregate profit on the aggregate business, must be *445reasonable. That seems to be sufficient reason for taking the business done and not the capital invested, as a basis for measuring a reasonable profit.
Id. at 128, 285 S.W. at 68-69.
Returning to G.S. 58-124.19(2), we note again that it provides that due consideration shall be given “to a reasonable margin for underwriting profit and to contingencies.” This precise phraseology has been interpreted by several courts. Illustrative is Pennsylvania Insurance Department v. Philadelphia, 196 Pa. Super. 221, 173 A. 2d 811 (1961). There, the court stated:
The accepted meaning of “underwriting profit” is stated in Bullion v. Aetna Insurance Company, supra, 151 Ark. 519, 237 S.W. 716, 718 (1922), as follows: “We think the undisputed evidence shows that the term ‘underwriting profit’ has long had a definite, certain, and well-known meaning in insurance circles. A number of witnesses of highest authority in the insurance business testified that the term was understood alike by all insurance men, and that the word ‘underwriting’ refers to operations of the companies in accepting and carrying risks on the writing of insurance, and refers to that branch of the insurance business in contradistinction to the investment or banking end of the business, and that underwriting profit or loss is arrived at by deducting from earned premiums all incurred losses and incurred expenses.
C. Statutory Authority
As explained in subsections A. and B. above, we think the Commissioner, in finding and concluding that income on invested capital should be considered as a factor in insurance ratemaking, misconstrued the law in this jurisdiction and the prevailing law in other jurisdictions. Accordingly, the Commissioner’s findings and conclusions in this respect were “affected by other error(s) of law” as contemplated by G.S. 58-9.6(b)(4) and G.S. 150A-5K4) and those portions of his order must be vacated and set aside.
Our review of the cases and other authority cited above leads us to conclude that the issue presented here has historically resulted from confusion over the difference in income from unearned premiums and loss reserve funds and in income from invested capital. The industry’s position has been and is that the insurance business is divided into two separate and distinct branches, (1) the underwriting business and (2) the investment business. The industry argues that in establishing an appropriate rate for policyholders, only income from investments in the underwriting portion of the business should be considered. It argues that it is required to segregate certain of its assets in order to show its ability to pay claims from its policyholders and that it would be foreign to the arrangement to expect the corporate investors to donate the income from their investments. In other words, income from assets in excess of the assets required to conduct its underwriting business should not be pertinent in a ratemaking case.
The opposing view is that the business of insurance should be regarded as a whole. So far as the real owners of the business, the stockholders, are concerned, all income arising from the business, and all expenses and loss incident thereto, should be considered, and must of necessity be considered in determining whether or not there has been a profit, and the extent and amount thereof. Aetna Insurance Company v. Travis, 124 Kan. 350, 259 P. 1068 (1927), cert. denied, 276 U.S. 628 (1928).
North Carolina clearly appears to distinguish between the different categories of insurance company income. As stated earlier, prior decisions in this State have sustained the view that investment income from unearned premiums and loss reserve funds are appropriately considered in a ratemaking hearing. Indeed, the 1979 Legislature amended G.S. 58-124.19(2) to make this abundantly clear. Neither prior cases nor statutes, however, have permitted consideration of invested income from investment capital.
Other sections of Chapter 58 further argue against considering income on investment capital. G.S. 58-35 provides that companies “shall maintain unearned premium reserves equal to the *447unearned portions of the gross premiums charged on unexpired or unterminated risks and policies. . . It provides other details in this respect and authorizes the Commissioner to calculate a company’s unearned premium reserves “upon the monthly pro rata fractional basis, or, if necessary, on each respective risk from the date of the issuance of the policy” if he is unsatisfied with the company’s conformance with the statutory formula. G.S. 58-35.2 provides that “[i]n determining the financial condition of any casualty insurance or surety company . . . there shall be included in the habilites of such company loss reserves and loss expense reserves at least equal to the amounts required under . . . this section.” (Emphasis added.) That section also sets out an elaborate scheme for loss reserve requirements. Among these is the requirement that “[f]or all such liability policies written during the three years immediately preceding the date of determination, such reserves shall be the sum of the reserves for each such year, which shall be 60°/o of the earned premiums on liability policies written during such year.” G.S. 58-35.2(c)(2). (Emphasis added.)
G.S. 58-79.1 deals with investment requirements for fire, casualty, and miscellaneous lines. Subsection (a) requires certain types of investments on “minimum capital investment.” Subsection (b) addresses “reserve investments required.” It provides that after satisfying requirements for minimum capital investments, companies may invest funds as specified in subsection (c) of the statute “unless it shall at all times have and maintain cash and such reserve investments (including its minimum capital investments), . . . which, when valued, . . . shall be at least equal in amount to fifty per centum (50%) of the aggregate amount of its unearned premium and loss reserves as shown by its last sworn statement. . . .” Subsection (d), entitled “Residue and Surplus Fund Investment,” then provides that after satisfying minimum capital investments and reserve investments required in subsection (b), a company may invest its residue and surplus fund investments except as prohibited by the various divisions of that subsection.
We believe this statutory scheme indicates that our Legislature has differentiated between the income earned on capital and income earned on loss reserves and premium reserves of insurance companies operating in North Carolina. The strict requirements pertaining to unearned premium reserves and loss *448reserves indicate an appropriate legislative concern with insuring that policyholders will be paid when claims are filed. Beyond insuring that minimum level of capitalization, the Legislature has been unconcerned with income on capital investments. In construing the statute otherwise, the Commissioner erred.
The next question presented is one argued by the Commissioner in his brief. Unfortunately he failed to give any notice of appeal from the Court of Appeals’ determination adverse to his position, and appellants have consequently filed a motion to this Court to strike that portion of the Commissioner’s brief. In view of the importance of the question presented for future ratemaking hearings, however, we elect to address the question on its merits, depsite the procedural irregularity.
Among the factors to which “due consideration” is to be given in determining a proper insurance rate is “a reasonable margin for underwriting profit and . . . contingencies.” G.S. 58-124.19(2). In his order, the Commissioner in essence rejected the traditional five percent of gross premium allowed for underwriting profit and contingencies. In lieu thereof, the Commissioner adopted a complicated, lengthy and novel formula for determining underwriting profit allowance. Among his conclusions were the following:
18. That the determination of underwriting profit margins should be calculated in accord with contemporary concepts of risk and return as understood in financial theory, specifically the capital asset pricing model as testified to by expert witness Dr. William Bishop Fairley and detailed in the attached appendix[,] the use of which theory and methodology in automobile insurance ratemaking has been upheld by the Supreme Judicial Court of Massachusetts.
19. That, to determine the level of underwriting profit allowance which, if earned along with minimum reasonable investment results, would produce for the average carrier a rate of return on capital expressed, as a percentage of premium volume, equal to that achieved by a typical business *449of similar risk characteristics^] [F]ive sequential steps are involved: ....
Following the above, the Commissioner’s order sets out the detailed “five sequential steps,” comprising three pages of the record. We think the Court of Appeals correctly and succinctly presented the issue. Judge Arnold wrote:
William Fairley qualified as an expert in economics and statistics, and testified as to the proper method of calculating appropriate rates of return in the insurance industry. His theory, in essence, requires that a “target rate of return” to the insurance companies be established. This is done by considering the “systematic risk” in the industry, that is, the degree to which the variability in return on an investment in that industry moves with the stock market, and adding the necessary “reward” to encourage investors to hold those securities. (For example, a stock that went down twenty percent when the market went down ten would have a high systematic risk and would require a higher reward.) This target rate of return is then used to calculate the appropriate underwriting profit as follows:
Target Return = Underwriting + Investment + Investment Profit Return on Return on Cash Flow Capital
Underwriting = Target - Investment - Investment Return Return on Return on Cash Flow Capital
The Commissioner ordered that this “capital asset pricing model” be used to calculate underwriting profit margins.
41 N.C. App. at 318, 255 S.E. 2d at 562.
We first observe that we do not reject the Commissioner’s formula because it is either complicated, lengthy, or novel. As explained fully in Section II. of this opinion, we do not interpret prevailing law to require that administrative agencies be unimaginative in the discharge of their duties. Indeed, within the general guidelines of the statutes, administrative agencies are to work out the details in order to effectuate the general policy set forth by the applicable statutes. The expertise of the administrative agency is absolutely essential in dealing with *450technical and complicated matters such as that presented by this issue. However, for the reasons stated below, we agree with the Court of Appeals’ conclusion that this portion of the Commissioner’s order must be reversed.
A. Error of Law
We first address the question whether this portion of the Commissioner’s order was “affected by other error(s) of law” as contemplated by G.S. 58-9.6(b)(4) and G.S. 150A-5K4). We find this portion of the Commissioner’s order erroneous as a matter of law for the two reasons stated below.
 First, our holding in Section IV. of this opinion, above, forecloses use of the “capital asset pricing model.” The capital asset pricing formula, as noted in the Court of Appeals’ summary set out above, clearly contemplates consideration of income on invested capital. We held in Section III., above, that consideration of income from invested capital is not presently allowed by North Carolina law. Obviously, striking such an integral part of the formula causes it to fall in its entirety.
Secondly, the Commissioner’s requirement for the use of a hypothetical “risk free” rate of return would clearly violate the intent of our Legislature in authorizing insurance companies operating in North Carolina to invest in certain securities. G.S. 58-79.1 specifically requires casualty insurance companies to invest reserve funds in one or more of ten different categories of investments. Included are government bonds, municipal bonds, corporate bonds, preferred bonds, bankers’ acceptances, first mortgage bonds, ground rents and certain stocks and real estate.
We are not inadvertent to that portion of Dr. Fairley’s testimony in which he made it clear that, under the proposed formula, companies would not be required to actually invest in risk free U.S. treasury securities. Clearly, the proposed formula only contemplates that the rate of return would be computed on the hypothetical assumption that the companies did so invest their funds. In other words, implementation of the proposed formula would not preclude companies from investing pursuant to G.S. 58-79.1. It is, however, our function to interpret the legislative intent. It is inconceivable to us that our Legislature intended that insurance companies invest their funds in certain designated *451securities and then require that those companies’ underwriting profits shall be computed on the hypothetical assumption that they were invested in something else. Such an interpretation would, as appellants suggest, “make a mockery of the statute.”
B. Arbitrary and Capricious Actions
 There is yet another reason this portion of the Commissioner’s order must be set aside. We think it arbitrary and capricious as contemplated by G.S. 58-9.6(6) and G.S. 150A-5K6).
It is apparent from a review of the record and the Commissioner’s order that he based his new formula solely on the basis of the testimony of Dr. William B. Fairley, an employee of the Division of Insurance, State Rating Bureau of Massachusetts, and a decision of the Supreme Judicial Court of Massachusetts which generally affirmed this approach to determining underwriting profit. We, of course, are not bound by a decision of our counterparts in a sister state. Moreover, after reviewing the Massachusetts decision and Dr. Fairley’s testimony, we find that the Commissioner has simply copied a complicated equation of an experiment in another state without proceeding with the careful and deliberate manner that had been employed in that state. Hence, we are compelled to conclude that the Commissioner’s actions were arbitrary and capricious.
A review of Attorney General v. Commissioner of Insurance, 370 Mass. 791, 353 N.E. 2d 745 (1976), reveals that the court did indeed give general approval to the complicated formula adopted by the Commissioner in his order in this proceeding. There, however, we find these significant distinguishing factors: First, the court expressly stated that none of the parties in that hearing challenged the Commissioner’s general method of setting the profit allowance but that the figures used in applying the method provoked disagreement. In a footnote, the court stated:
The Bureau is careful in its brief to limit its acceptance of the Commissioner’s procedure. It claims that its acquiescence in the new procedure was the result of a “pragmatic” decision to allow a “full discussion of the method.” The Bureau does not attempt to defend the traditional method nor does it propose any alternative.
*452Also, the Massachusetts court had several criticisms of its commissioner’s reasoning. For example, it stated, “[T]he Commissioner’s approach seems suspect because it fails to confront and to consider all the elements of risk of an investment in such an insurer.” Id. at 817, 353 N.E. 2d at 762. The court went on to note, however, that it would not require a remand because “we feel [the] testimony [of expert witnesses], in its cumulative effect, provides adequate support for the Commissioner’s figure.” Id. at 819-20, 353 N.E. 2d at 764. Unlike the situation before us, three experts testified before the Massachusetts Commissioner.
Indeed, we take particular note of that court’s cautious approval of its commissioner’s action. The court stated,
[T]he judgmental estimate of the Commissioner of the proper adjustment is of the type that can be expected on the initial application of a novel methodology. Crude estimation could not be tolerated in normal circumstances, and we observe that the Commissioner called on future participants in the hearing process to present “more formal data and analysis.”
Id. at 821, 353 N.E. 2d at 764 (footnotes omitted).
In other words, while the Massachusetts court approved its commissioner’s adoption of the formula ordered by the Commissioner in this case, it expressly made clear that it was willing to do so only because its commissioner had satisfied the court that he would continue to refine the application of the “novel methodology.” In contrast, the Commissioner here has blindly adopted a novel approach with no such assurance.
That the Commissioner of Insurance of Massachusetts did not arbitrarily and capriciously attempt to inject this novel methodology into the ratemaking process in that state is made even clearer by reviewing the testimony of the witness Fairley. This witness’s testimony establishes clearly that the Massachusetts commissioner gave notice in an earlier decision that he was suggesting a new approach to rate-of-return regulation to take into account investment income and risk-free investments. All parties therefore had an opportunity to study and respond to it. Moreover, this witness was employed for the very purpose of helping to develop the novel approach. He testified with respect to the decision of the Massachusetts court discussed above:
*453Towards the end of their decision they noted that there were a number of unanswered questions and it was a novel procedure and that somewhat more refinement would be expected in its application in future years than in the beginning, but they accepted it and his numbers for that year. Well, now I came to work in the department just at that time and I began to work immediately on this issue in order to respond to the court’s suggestion that additional research was necessary.
The witness also testified that several economic experts had testified that the formula was “on the right track” but that they disagreed with certain aspects of it. Hence, work continues towards refining the formula in Massachusetts. Moreover, while testifying that, in his opinion, following the traditional approach in determining underwriting profit would probably result in excessive insurance rates in most states, Fairley stated, “Now I’m not — I have not made a study of North Carolina experience. I certainly cannot testify that the profit allowance, traditional allowance here in North Carolina is making rates excessive in this State.”
Other portions of this witness’s testimony are equally as revealing with respect to the non-refinement of the proposed method. The point is simply that the Commissioner of Insurance of North Carolina did nothing more, in adopting a complicated and novel formula for determining underwriting profit, than listen to one employee of an insurance department in a sister state which is refining the policy adopted and which was given only limited approval by the Supreme Court of Massachusetts. We think such an approach a clear example of an arbitrary and capricious action by an administrative agency as contemplated by our Legislature in establishing that criterion for judicial review.
Burden Of Proof
 Appellants next contend that the Court of Appeals erred in holding that there is no burden of proof on the Commissioner, in a proceeding of this nature, to disapprove a filing.
Appellants correctly note that prior to the 1977 changes in Chapter 58 of our General Statutes, prior approval by the Com*454missioner of Insurance was a condition precedent to any increase or decrease in insurance rates. Following the 1977 rewrite, the “prior approval” system was abandoned and a “file and use” system became effective. By this is simply meant that the Rate Bureau files proposed rate changes with the Commissioner of Insurance for coverages within the Bureau’s jurisdiction and those rates automatically go into effect unless they are disapproved by the Commissioner in accordance with specific statutory rules and procedures. G.S. 58-124.20(a); G.S. 58-124.21. Appellants argue that file and use rates are set by the Bureau and, nothing else appearing, are fully effective. If the Commissioner desires to challenge a rate, the burden is on him to take affirmative action. It necessarily follows, appellants contend, that when the Rate Bureau has made out a prima facie case, as here, the burden of proof shifts to the Commissioner to show by a preponderance of the evidence that the proposed rates are either excessive or unfairly discriminatory.
We think the Court of Appeals correctly rejected appellants’ contention. While the 1977 changes in our insurance laws were substantial, we discern not the slightest intent on the part of our Legislature to shift the burden of proof in insurance ratemaking hearings. We do not think our Legislature would have been silent had it intended such a radical change from past procedure. We also note that the original version of Chapter 828 of the 1977 Session Laws did contain a provision (Section 3[e]) which would have clearly placed the burden of proof on the Commissioner. This portion of the proposed legislation was deleted prior to final passage of our present statutes.
We are not inadvertent to our language in In re Rogers, supra, that “such a procedure [not requiring the burden of proof to shift to the administrative agency] would be in conflict with our usual civil practice on assignment of burden of proof. As a general rule in this jurisdiction, the party who substantially asserts the affirmative of an issue bears the burden of proof on it.” 297 N.C. at 59, 253 S.E. 2d at 919. While the general language in Rogers is supportive of appellants’ position here, we must agree with the Commissioner that had the Legislature intended such a drastic change in procedure, it would have said so. Under our present insurance laws, it is the clear intent of the Legislature that the proponent of a rate increase, the Rate *455Bureau, is to shoulder the burden of showing the reasonableness of the proposed increase. While the statutes place other burdens on the Commissioner, such as making findings of fact and conclusions of law, they leave no room for any other construction but that the underlying burden of proving the need and reasonableness of a rate increase rests upon the Rate Bureau.
We affirm the Court of Appeals’ holding that “[t]here is no burden upon the Commissioner to disprove the filing.”
Specificity Of Commissioner’s Order
Appellants next contend that the Court of Appeals erred in holding that the Commissioner complied with that portion of G.S. 58424.21(a) which requires that “[i]f the Commissioner after hearing finds that the filing does not comply with the provisions of this Article, he may issue his order determining wherein and to what extent such filing is deemed to be improper. . . .” (Emphasis added.) In so holding, the Court of Appeals simply noted that the Commissioner in his order had set out 99 findings of fact and 32 conclusions of law and that such were sufficient compliance with the statute.
Appellants argue that one of the purposes of G.S. 58-124.21 was to require the Commissioner’s order to show exactly how much a filing is affected by a proposed deficiency in order to allow for effective judicial review and to reduce the long-standing necessity for constant remand to the Commissioner. Appellants correctly note that the sheer number of purported findings and conclusions in a Commissioner’s order should not be determinative of the question whether the order complied with the statutory requirement.
Appellants also note that their evidence clearly supported a 23.2% increase, that only a 6% increase was allowed by statute and that an error or errors totaling in excess of $30 million in overall loss experience would have to have been made in the filing in order to justify the Commissioner’s rejection. Hence, appellants contend, the Commissioner has shown no justification for rejecting any of the proposed increases.
 It is obvious, as indicated by the extensive discussion in the first sections of this opinion, that the Commissioner rejected this *456proposed rate increase primarily on the basis of his finding that the data presented by the Rate Bureau was unreliable. Though we have held this basis of the Commissioner’s order to be improper, based on the record before us, we think the Commissioner’s order would be in compliance with G.S. 58-124.21 had the conclusion been justified that the data was indeed unreliable and had none of the other errors discussed in the preceding sections been committed. Hence, we find no merit in this assignment of error. However, appellants’ point is well taken that the present statute requires the Commissioner to be mathematically specific in rejecting proposed rate increases and future orders should specify “wherein and to what extent” the proposed filings are deemed improper.
Adequacy Of Notice
 Appellants next contend that the Commissioner failed to comply with that portion of G.S. 58-124.21(a) which provides: “At any time within 30 days from and after the date of any filing, the Commissioner may give written notice to the Bureau specifying in what respect and to what extent he contends such filing fails to comply with the requirements of this Article. . . .” (Emphasis added.) The Court of Appeals held that the Commissioner did comply with this portion of the statute.
The Court of Appeals based its holding on a finding that the filing did not indicate whether the data had been audited and that it could not assume that which is not supported by the record. Finding that the filing nowhere stated that the data was unaudited, the Court of Appeals held that the Commissioner complied with this portion of G.S. 58-124.21(a). The inference is that the Commissioner was surprised at the hearing to find the data was unaudited. He therefore had no opportunity to notify appellants of this deficit.
With this reasoning of the Court of Appeals, we disagree. It is perfectly clear from the record that the Commissioner knew the data was not audited. The filing made clear the extent to which the data was verified and that the verification methodology was consistent with that employed in previous years. The record is equally clear that previous filings had not required audited *457data. Since the Commissioner knew the data was not audited and his subsequent rejection of the filing was based primarily on that ground, both fundamental fairness and the quoted portion of G.S. 58424.21(a) mandated that he give notice of his dissatisfaction with the quality of the data in his notice of hearing. For this reason alone, and absent any other assignment of error, we would be bound to vacate and set aside the Commissioner’s order.
We wish to emphasize the narrow holding in this portion of our opinion. The Commissioner correctly notes that it was clearly not the intent of the Legislature to prevent the Commissioner from disapproving a filing if matters coming to his attention during the course of a hearing would compel such disapproval. Obviously, matters relating to credibility or other factors might arise during the course of a hearing for which the Commissioner could not have provided notice prior to the hearing. What we hold here, and all that we hold here, is that when the Commissioner knows prior to the giving of public notice “in what respect and to what extent he contends such filing fails to comply with the requirements of [the] Article,” then he must give the specifics in his notice of public hearing. Here, the Commissioner clearly failed to do this with respect to the reliability of the data.
V — I
Bad Faith Of Appellants
Appellants finally contend that the Commissioner improperly included in his order findings and conclusions that they were guilty of bad faith because of dilatory action with regard to the filing in several instances.
We do not find it necessary to enumerate the various data which the Commissioner found was not supplied by appellants. It involves such matters as an alleged failure by appellants to break down incurred losses into paid losses, cash reserves, and IBNR (incurred but not reported losses) and a failure to produce claim loss and frequency trend factors. Suffice it to say that we have carefully examined the record and believe that much of the data referred to by the Commissioner was difficult, if not impossible, to obtain in the short period of time between the notice of public hearing and the convening of the hearing. Moreover, in a number of instances, the Commissioner failed to inform the Bureau that it *458had not complied with his order to produce data to his satisfaction until the time of hearing.
We find no evidence of bad faith on the part of appellants and those findings and conclusions of the Commissioner’s order referring to bad faith on the part of appellants are vacated and set aside.
Other Holdings Of Court Of Appeals
To assist in understanding the result of our review of the Court of Appeals’ holdings, we here note the remaining portions of the Court of Appeals’ decision not argued by the parties on this appeal:
(1) The Bureau in its filing proposed that future premium rates vary within a range of plus or minus 5% according to the “territory” or geographical area of the State in which an insured is located. The Commissioner found that the projections of territorial rate differences did not take into consideration the new statutory classification plan and did not reflect reasonably anticipated territorial loss experience. He concluded that this made the proposed rate changes excessive and unfairly discriminatory. He also entered other findings and conclusions with respect to territorial rate differences. The Court of Appeals held that the evidence did not support the Commissioner’s findings and conclusions and these portions of his order were set aside.
(2) The Commissioner’s order included findings concerning collision insurance deductibles and a conclusion that the rate proposed for $25.00 deductible collision insurance was excessive in relation to the coverage provided. Again, the Court of Appeals found no evidence in the record to support the conclusion and those portions of the Commissioner’s order were set aside.
(3) In his order disapproving the filing, the Commissioner further ordered that the Bureau be allowed 60 days within which to file an amended filing consistent with his findings and conclusions. The Court of Appeals held that, since the Bureau excepted to and appealed from the Commissioner’s order, the appeal removed the matter from the Commissioner to that court and that part of the Commissioner’s order therefore became a nullity.
*459With respect to these three holdings, we do not disturb the Court of Appeals’ decision. None of the parties have raised these points on appeal to this Court.
In accordance with our discussion above of the various portions of the Court of Appeals’ opinion, the decision of that court is
Affirmed in part and reversed in part.
G.S. 58-9.6(b) provides that we may “affirm or reverse the decision of the Commissioner, declare the same null and void, or remand the case for further proceedings; or [we] may reverse or modify the decision if the substantial rights of the appellants have been prejudiced. . . .” G.S. 150A-51 provides that we “may affirm the decision of the agency or remand the case for further proceedings; or [we] may reverse or modify the decision if the substantial rights of the petitioners may have been prejudiced.” In cases involving narrow and specific error in the Commissioner’s order such as the Commissioner exceeding his authority or failing to set forth specific findings of fact, we would ordinarily remand the case to the Commissioner for further proceedings. Commissioner of Insurance v. Automobile Rate Administrative Office, 292 N.C. 1, 231 S.E. 2d 867 (1977); Commissioner of Insurance v. Automobile Rate Administrative Office, 287 N.C. 192, 214 S.E. 2d 98 (1975). Here, however, it is apparent that the multiple errors committed by the Commissioner in the proceedings and order before us are of such magnitude as to make remand futile. The order of the Commissioner dated 27 February 1978 is therefore
Reversed and declared null and void.
The former version of G.S. 58-124.22(b), under which this proceeding is governed, provided in pertinent part that:
Whenever a Bureau rate is held to be unfairly discriminatory or excessive and no longer effective by order of the Commissioner issued under G.S. 58-124.21, the members of *460the Bureau shall have the option to continue to use such rate for the interim period, pending judicial review of such order, provided each such member shall place in escrow account the purportedly unfairly discriminatory or excessive portion of the premium collected during such interim period and the court, upon a final determination, shall order the escrow funds to be distributed appropriately .... (Emphasis added.)
Accordingly, all escrowed premium funds representing this proposed rate increase shall be remitted to the member insurers forthwith.
It is so ordered.
Justice BROCK took no part in the consideration or decision of this case.