Petitioner contests the Acting Superintendent of Insurance’s (hereinafter “Superintendent”) decision reducing certain of their instituted1 rate increases. By order of October 22, 1982, the Superintendent lowered petitioner’s rate increases with respect to property damage liability and collision coverage effective November 1, 1982.2 This decision was founded largely on the Superintendent’s holding that petitioner’s use of a five point exponential trend line,3 for purposes of predicting future revenue requirements and claim demands, was inadequately justified, and that the previously approved use of an eight point line was more appropriate.4 On that basis, petitioner’s rate increases were adjusted downward. See D.C.Code § 35-1704(c) (1981). Petitioner asserts that this decision was arbitrary, capricious and unsupported on the record. We agree and vacate the Superintendent’s order, remanding for further proceedings.
*815I
The procedural posture of this case is not in dispute. On September 3,1982, petitioner filed a rate increase with the Superintendent. See D.C.Code § 35-1704(a) (1981). The new rates were to be effective on all policies written on or after September 6, 1982. On September 13, 1982, the Superintendent notified appellant that a hearing would be held concerning the rate adjustments on September 23, 1982. See D.C. Code § 35-1704(c) (1981). At a hearing on that date, evidence was presented by petitioner through its actuarial expert outlining in detail the reasons for the rate increases and the methods by which they were determined. The Superintendent presented no direct evidence, but did extensively cross-examine petitioner’s expert. On October 22,1982, the Superintendent’s order, adjusting the rate increases downward, was issued. By order of October 29, 1982, this court stayed the order of the Superintendent pending our expedited review. Cf. Don’t Tear It Down, Inc. v. District of Columbia, 395 A.2d 388, 390 (D.C.1978). We note that it was necessary that the motions panel find that petitioner had a likelihood of success on the merits in order to grant the stay. Id. at 390.
II
At the outset, we are constrained to correct apparent misconceptions about the Superintendent’s statutory authority. See D.C.Code §§ 35-1701 through 1710 (1981). The District would have us confer almost unbridled discretionary authority upon the Superintendent to reject a stated rate increase. At the same time, it would establish an improper hurdle for the insurer seeking rate relief. We think neither of these results is contemplated by the statute.
The District’s attempt at a post-hoc rationalization of the Superintendent’s order evidences a fundamental misunderstanding of the statutory role of the Superintendent. We do not dispute that the insurer must bear the burden of demonstrating both the need and reasoning behind a rate increase request. See, e.g., Pack v. Royal Globe Ins. Companies, 224 Tenn. 452, 457 S.W.2d 19 (1970). Neither do we dispute the respect a court on review owes to the expertise of the Superintendent. See, e.g., Attorney General v. Commissioner of Insurance, 370 Mass. 791, 353 N.E.2d 745 (1976). We cannot, however, accept the unsubstantiated incongruity in the Superintendent’s order which allowed certain increases to pass untouched while actually reducing the premiums to be charged for property damage, when prior claim history supported an increase. Nor can we lose sight of the presumptions in favor of the increase established by the statutory scheme. See D.C.Code §§ 35-1703(f), -1704(c) (1981).
Comparisons with the regulation of utilities are unfounded. In the case of insurers, the law does not seek to restrain a conferred monopoly, but instead, to insure the financial soundness of the members of a highly competitive business. The Superintendent’s primary concern should be to assure that the insurer’s fiduciary obligation to the insured is fulfilled in the form of competitive rates which are not “excessive, inadequate or unfairly discriminatory.” D.C.Code § 35-1703(a). The Superintendent’s authority must be viewed in this light, contrary to any notion of absolute authority to compel business policy short of a threat to fiduciary obligations.
As noted above, D.C.Code § 35-1704(c) (1981) permits immediate implementation of the insurer’s new rates subject to subsequent review by the Superintendent. This, in effect, stamps the revised rates with the imprimatur of validity. It shifts to the Superintendent, at the very least, the responsibility to clearly articulate factually supported reasons for rejecting the increases. The Superintendent may not simply raise an insurmountable “burden of proof” hurdle, and then baldly state that petitioner has failed to surmount it.
Ill
In this case, the nub of the Superintendent’s order is his finding that the use of a *816five point data trend line was a material departure from previous, sanctioned5 practice, and that such a change was not adequately justified by petitioner. Without more, this is an insufficient reason to reject the rate increases.
Firstly, as was ably demonstrated by petitioner both in its brief and during argument, it cannot be fairly stated that the use of a five point trend line effects a material alteration from previous policy. Though the use of an eight point line had been approved in the most recent filing, ample evidence was presented that trend lines with various numbers of data points had been used in the past. The specific number of data points in any trend line will necessarily be a function of a variety of changing actuarial factors. See D.C.Code § 35-1703(b) (1981). It is thus error to require that petitioner justify the “change,” independent of its legitimate duty to factually support the rate increases. The Superintendent’s reliance, therefore, on petitioner’s failure to justify the use of the five point line is misplaced. It certainly is of no help in assessing the overall legitimacy of the rate increases.
Secondly, we are unable to discern why the superintendent rejected petitioner’s comprehensive testimony on the reasons for the use of the five point line in plotting future revenue requirements. The order is of little help. To paraphrase, the order merely states that (1) petitioner presented inadequate evidence to justify use of the five point trend line, (2) the five point line “fits” the data points no better than the eight point line, and (3) only a two point line would provide a perfect fit. Therefore, the eight point line will be used and rates adjusted accordingly.
This is a non-sequitur resulting from an erroneous notion about the burden of proof and persuasion as it relates to an insurer’s stated rate increases. It does not provide the record support necessary to uphold the Superintendent’s findings and conclusions of law. See Citizens Ass’n of Georgetown, Inc. v. District of Columbia Zoning Comm’n, 402 A.2d 36 (D.C.1979). See also Blue Cross of Kansas, Inc. v. Bell, 227 Kan. 426, 607 P.2d 498, 505 (1980). The entire hearing was devoted to petitioner’s testimony in support of the rate increases. The District’s cross-examination of petitioner’s expert did not negate the thrust of his testimony. In the face of such record evidence, it was not sufficient for the Superintendent to rule as he did. The Superintendent must articulate factually supported reasons for the adjustment or rejection of the rate produced by ordinary market influences.
Reversed and remanded.
. D.C.Code § 35-1703(f) (1981) provides that “[r]ates may become effective immediately upon filing or at such future time as the company or rating organization making them may specify.” Id. There is no lag time between the rate increase announcement and its implementation.
. On September 3, 1982, petitioner filed with the Superintendent notice and support materials for the following rate increases:
Coverage Percent Change
Bodily Injury No Change
Uninsured Motorists No Change
Property Damage + 8.2%
Collision + 15.0
Comprehensive + 14.3
Medical Payments No Change
Towing and Labor +25.0
All Coverages + 6.9%
By his October order, the Superintendent actually effected new property damage premiums which would be reduced 3.6% from that charged before the rate increase. He also reduced the collision premium increase to 9.2% from 15% as instituted.
. Trend lines are lines which join data points on a graph that represents prior revenue statistics. They are used as a statistical device to predict future revenue needs.
. The previous use of an eight point trend line had been approved in order No. 82-1 of February 18, 1982, by the Superintendent.
. See note 4, supra.