OPINION
¶ 1 Microsoft appeals the superior court’s award of $19,132,728 in attorneys’ fees to class counsel 1 for their representation of the class members in this consolidated antitrust action. That award is based on a settlement agreement with a “face value” of $104.6 million and a provision that Microsoft pay class counsel “reasonable attorneys’ fees ... calculated on a common fund basis.” Microsoft argues that the superior court erred in the calculation of the attorneys’ fee award by applying a 3.42 multiplier to an incorrect lodestar 2 figure.
¶ 2 While we have concerns with the amount of the fees awarded in this ease, those concerns are not sufficient for us to hold the superior court abused its discretion in selecting a 3.42 multiplier. We reach that conclusion given the deferential standard of review we must apply, the factual record, Microsoft’s own experts suggesting a multiplier in the range of 1.5 to 2.3, Microsoft agreeing to fees being determined under the common fund doctrine, and class counsel’s expert recommending a multiplier of 5.3. However, we do find the court erred in applying a multiplier to post-settlement fees and in calculating the lodestar figure. Accordingly, we affirm the court’s imposition of a 3.42 multiplier, reverse the court’s application of the multiplier to post-settlement fees and calculation of the base lodestar figure, and remand for further proceedings consistent with this decision. 3
Facts and Procedural History
¶ 3 The fee dispute at issue arose out of Arizona consolidated class actions,
4
in which the class alleged that Microsoft had violated Arizona law by asserting unlawful monopoly power and engaging in anti-competitive conduct. The Arizona class action was initiated after a federal judge made extensive findings of fact supporting the conclusion that Microsoft had engaged in improper use of monopoly power and violated federal antitrust laws.
See United States v. Microsoft,
¶ 4 Arizona class members used, and were encouraged to use, the ongoing litigation against Microsoft in other jurisdictions throughout the pretrial proceedings, including coordinated deposition discovery. Class members successfully moved the superior court to grant summary judgment that the findings of fact in the federal action have preclusive effect and demonstrate that Microsoft was liable for unlawful conduct. Accordingly, Microsoft refers to the Arizona class action as a “tag-along action.”
¶ 5 Notwithstanding Microsoft’s allegations that the Arizona class action was “tag-along” in nature, it took Microsoft and the class members over four years of litigation before they were able to settle the action. During that time, class counsel were successful in certifying the class and defending against dismissal based on federal precedent that indirect purchasers could not recover under federal antitrust law.
See Illinois Brick v. Illinois,
¶ 6 Ultimately, the parties reached an agreement to settle “All Cases” 6 against Microsoft (the “Settlement Agreement”). Pursuant to the Settlement Agreement, Microsoft agreed to make available to class members vouchers that ranged in value from $9—$15, which, if collected by every class member, would have a “face value” of $104.6 million. The Settlement Agreement also contained a cy pres 7 clause. Pursuant to that provision, if the total amount of vouchers claimed did not reach the face value amount, the difference between the face value amount and the total amount of issued vouchers claimed would be divided in half and distributed to eligible school districts. The cy pres amount would also include one half of the value of the coupons distributed but unused by March 2009.
¶7 The Settlement Agreement further provided that Microsoft agreed to pay reasonable attorneys’ fees “calculated on a common fund basis” for work performed by all counsel for plaintiffs in “All Cases in connection with this litigation.” The parties specified that:
(2) the amount of the attorneys’ fees will be determined upon the basis of the “common fund” doctrine rather than as a “prevailing party” or statutory fee, with each party being free to argue for what it believe[s] is a reasonable common fund fee; (3) plaintiffs are free to argue that the value of the “common fund” created by the litigation is any amount up to ... Face Value ... [A]nd Microsoft is free to argue that the “common fund” created by the litigation is any amount less than the Face Value or is the amount of the benefit conferred directly on class members____
¶ 8 Subsequently, class counsel filed a joint application for attorneys’ fees in the sum of $34.8 million, which is one third of the face value of the Settlement Agreement. Class counsel provided evidence of risks that were inherent in the class action, such as: the difficulty in prevailing in an antitrust action; Microsoft’s extensive resources allowed it to be an aggressive opponent; the indirect purchaser hurdle; the difficulty of certifying a class; and causation barriers. Class counsel suggested that the court cross-check the per *349 centage requested with a lodestar value of $6.6 million multiplied by 5.3. Class counsel pointed to persuasive, but not precedential, authorities to support their request of a 5.3 multiplier. The application for attorneys’ fees was supplemented by the declarations of three attorneys’ fees experts that all support the fee application and found the fee request to be reasonable under prevailing rates for attorneys’ fees.
¶ 9 Microsoft opposed the fee application and requested the court find an award of $7.1 million to be a reasonable request. Microsoft urged the court to consider the actual value, rather than the face value, of the Settlement Agreement when calculating the fees. Microsoft argued that, with the cy pres award, the actual value would be “in the range of $53.9 million.” Microsoft also argued that the lodestar method of calculations was more appropriate. Microsoft requested class counsel’s proposed lodestar be reduced by $1.7 million because class counsel rates were excessive and it included work that was not compensable. Microsoft also argued that “Arizona case law supports a multiplier between 1.5 and 2.3” and reasoned that, considering the risk of this case, a multiplier of 1.5 would be appropriate.
¶ 10 While Microsoft unmistakably argued in the superior court that no multiplier should be applied to the hours spent after the case had settled, it did not unequivocally contend no multiplier should be used at all. Instead, Microsoft first argued that a multiplier was appropriate, but should be no more than 1.5:
The law and record support a fee award to class counsel of $7.1 million, which would include a lodestar of $4.9 million and multiplier of 1.5 on time spent prior to settlement.
In this same vein, Microsoft analyzed the risk to class counsel in undertaking and litigating the case and concluded that based on risk, class counsel’s requested multiplier of 5.3 was “unwarranted” and there was no need to award a “significant multiplier.” In contrast to the appropriateness of a multiplier based on risks, Microsoft argued that no multiplier was appropriate based on counsel’s skill, the difficulty of the case and the results. Significantly, Microsoft concluded its argument by stating:
A 1.5 multiplier would provide more than enough enhancement for the risk class counsel faced of not being paid in this case. Were the Court to consider awarding an even higher multiplier, however, it should not be greater than 2.3. 8
¶ 11 In support of its opposition, Microsoft filed the declaration of Mitchell Polinsky, an attorneys’ fees expert, who recommended a fee of $12.2 million if the percentage-of-recovery method was used, or a fee in the range of $7.4—9.9 million if the lodestar method with a multiplier of 1.5 was used to calculate fees.
¶ 12 Microsoft also filed the declaration of another attorneys’ fees expert, Peter D. Baird, to support its opposition. Baird recommended a fee close to $12.2 million ($12.19 million to be exact) if the percentage-of-recovery method was used. As the actual value of the settlement is speculative, Baird opined that the “lodestar method may be ... the appropriate method for calculating attorneys’ fees.” Baird disagreed with the lodestar figure requested by class counsel because: (1) it included time litigating Bunker’s Glass; and (2) class counsel’s hourly rates were unreasonable and exceeded the prevailing market rate. Baird instead recommended recalculating the lodestar figure and enhancing it by a multiplier of 2.3. Baird’s final calculation of reasonable attorneys’ fees using an enhanced lodestar was $10,638,852.38.
¶ 13 Thus, the superior court was presented with Microsoft claiming a fee recovery in the range of $7.4-12.2 million including possible use of a 1.5-2.3 multiplier. In contrast, class counsel argued for an award of approximately $34.8 million based on a percentage of the settlement with a cross-check of enhancing a lodestar by a 5.3 multiplier.
*350 ¶ 14 After oral argument, the superior court found that attorneys’ fees could be awarded in a common fund class action either by a percentage of the fund or by the lodestar multiplier method. However, the court applied the lodestar method, as there was “no solid, true common fund.” The court further made the following findings to support its use of a multiplier: (1) the “enormous” economic and related risks to taking the case; (2) the precedent of Illinois Brick; (3) risk of loss throughout the litigation, such as the risks of surviving the dispositive pretrial motions filed by Microsoft, the appeals filed regarding those motions, and the class certification; (4) difficulty ascertaining damages, causation and amount; and (5) “[t]he fact that plaintiff was successful on the above risks.”
¶ 15 The court calculated the total reasonable attorneys’ fees at $19,132,728 by multiplying a lodestar figure of $5,594,365 by 3.42. The court’s calculation of the base lodestar was slightly different from the base lodestar requested by Arizona class counsel and significantly less than the lodestar requested by California class counsel. A comparison of the amounts requested and granted and a final allocation of the fees is as follows:
[[Image here]]
¶ 16 Microsoft timely appealed. We have jurisdiction pursuant to Arizona Revised Statutes (“A.R.S.”) § 12-2101(B) (2003).
Standard of Review
¶ 17 We review the superior court’s award of attorneys’ fees for abuse of discretion. Or
faly v. Tucson Symphony Soc’y,
Issues
¶ 18 Microsoft raises two issues on appeal: (1) Whether the superior court erred by enhancing the lodestar with a 3.42 multiplier; and (2) Whether the superior court erred in calculating the base lodestar figure.
1. Lodestar Enhancement
A. Authority to Use a Multiplier
¶ 19 In its opening brief, Microsoft contends the superior court erred in applying any multiplier to enhance the lodestar, arguing in part that this is a matter of contract interpretation. Alternatively, Microsoft argues that as a matter of law the superior court lacked authority to apply a multiplier and that a multiplier in this ease violates various policies dealing with the shifting of attorneys’ fees. We disagree with these arguments. The Settlement Agreement ex *351 pressly provides that the attorneys’ fees would be based on a common fund analysis. Such an analysis may include a multiplier and the Settlement Agreement does not prohibit a multiplier.
¶ 20 In the Settlement Agreement, the parties did not refer to any use or prohibition of a multiplier. Rather, they agreed to determine reasonable attorneys’ fees using the common fund doctrine. It is undisputed that there is no true common fund here.
¶ 21 As applied in its usual context, the common fund doctrine is an equitable exception to the general rule that each party bears his or her own fees and a superior court can only impose a fee award if authorized by contract or statute.
See Burke,
¶ 22 Under normal circumstances it would be error for the superior court to apply the common fund doctrine in this ease because the attorneys’ fees are not being spread among the class members. Instead, the fees are being shifted to Microsoft. However, Microsoft and class counsel agreed to determine the attorneys’ fees using the common fund doctrine.
¶ 23 The plain language of the Settlement Agreement reflects that the parties negotiated and agreed to apply a fee spreading doctrine in a fee-shifting situation. Accordingly, the court properly applied fee-spreading principles, even though the “primary element” of “spreading of attorneys fees among the benefited class is glaringly absent.”
Burke,
¶ 24 Alternatively, Microsoft relies on the policies against enhancement of a lodestar set forth in
City of Burlington v. Dague,
¶ 25 If the parties had agreed to compensate class counsel under the prevailing party
*352
or statutory fee-shifting doctrine, we might agree that those policies would preclude lodestar enhancement. However, the parties expressly agreed that fees would be determined using common fund principles, not by using any statute or prevailing party analysis. The policies against enhancement, applied in the specific context of determining fees under the statutory fee-shifting doctrine, are not applicable to fees awarded under the common fund doctrine.
Compare Dague,
¶ 26 Microsoft also suggests the superior court lacked authority to apply a multiplier because “Arizona appellate courts have never enhanced a lodestar with a multiplier when awarding fees under the common fund doctrine.” We note that there is some Arizona precedent to use a multiplier in the common fund context. In
Burke,
this Court relied on
WPPSS,
¶ 27 In its reply brief, Microsoft argues that the issue on appeal is whether the superior court erred in applying an enhanced lodestar under the common fund doctrine because “the properly calculated lodestar is the presumptively reasonable fee and multipliers are to be applied only in rare and exceptional cases.”
¶ 28 We first note that, in its opposition to the fee application, Microsoft failed to argue that the base lodestar figure was the only reasonable fee in this case. Rather, it argued below that a 1.5 multiplier would be appropriate and that even a 2.3 multiplier could be used. Supra, ¶¶ 10-12. Accordingly, even if we were to accept Microsoft’s contention without analysis, it appears Microsoft conceded that this case was appropriate for some kind of multiplier based on risk. Thus, the critical difference between Microsoft and class counsel is one of degree rather than kind — how much of a multiplier was appropriate and not whether a multiplier was appropriate at all.
¶ 29 Independently of Microsoft’s position below, we recognize that some courts have
*353
held that in common fund cases, the lodestar is presumptively reasonable and enhancement rarely and exceptionally applied.
E.g., Fischel v. Equitable Life Assurance Soc. of the United States,
¶ 30 Thus, in the common fund context, courts have authority and discretion to enhance a lodestar by use of a multiplier based both on risk, result and quality of services performed. This principle is consistent with the overall objective of awarding a reasonable amount of attorneys’ fees, to which these parties agreed.
Fischel,
¶ 31 Based on the risk of nonpayment in the typical common fund case, courts give great weight to that risk when determining enhancement. Generally, determination of whether a multiplier is called for is best left to the discretion of the superior court.
WPPSS,
(1) attorneys take a case with the expectation that they will receive a risk enhancement if they prevail, (2) them hourly rate does not reflect that risk, and (3) there is evidence that the ease was risky.
Fischel,
¶ 32 Accordingly, the superior court’s authority to apply a multiplier also applies to a situation where fee-shifting rather than fee-spreading applies, but the parties have agreed to use the common fund doctrine to determine the amount of fees.
B. Application of a Multiplier
¶ 33 The superior court is in the best position to evaluate the risks that attorneys assume when pursuing a case and whether the result is successful for the class.
WPPSS,
¶ 34 These findings are supported by the record and appropriate for enhancement.
*354
For example, the class in these actions faced a risk not present in either the federal antitrust action against Microsoft or the earlier California certified class action — whether indirect purchasers had standing to sue Microsoft under Arizona law.
11
Moreover, any indirect purchaser antitrust action suffers from the grave risk that unlawful prices either were not passed on to the indirect purchasers or cannot be determined.
Illinois Brick,
¶ 35 Moreover, in its opposition in the superior court, Microsoft raised the issue as to whether a multiplier was warranted based on factors other than risk, but contended that a multiplier was appropriate based on risk and that a 1.5 multiplier was more appropriate than a 5.3 multiplier. This request was supported by Microsoft’s own expert declarations that recommended enhancement with multipliers between 1.5 and 2.3. On that basis, we cannot find an abuse of discretion in applying a multiplier.
¶ 36 Microsoft also contends that even if a use of a multiplier was appropriate, the 3.42 multiplier used by the superior court to enhance the lodestar was erroneous.
12
We find no abuse of discretion in the superior court’s selection of the 3.42 multiplier. Use of a multiplier resulting in the total amount of fees is not a per se violation of public policy. An award calculated by an enhanced multiplier may be ipso facto unreasonable as a violation of public policy if the award is not comparable with reasonable awards using other methods, such as the percentage of the fund method.
See, generally, Ladewig,
*355
¶ 37 The actual value of the
cy pres
and the vouchers claimed by class members will not be certain until 2009. However, it is estimated that the class members will receive approximately $6.6 million in vouchers. Calculating the
cy pres
amount to be approximately $49 million, the actual benefit to the class and the
cy pres
is estimated to be somewhere around $55-58 million. Accordingly, the award of $19 million is approximately 33-35% of the estimated actual value to the class members. Awards of one third of a common fund are not per se unreasonable.
See, generally, Vizcaino v. Microsoft Corp.,
¶ 38 We restrain from substituting our judgment for that of the superior court when the multiplier requested is within a range that is advocated as reasonable by class counsel’s experts and is significantly close to the range advocated by Microsoft’s own experts. It would have been helpful if the superior court had explained how it came to the 3.42 figure, as it was neither advocated by the parties nor their experts.
14
See e.g., Rawlings,
¶ 39 In summary, while we might not have reached the same result as the superior court, we cannot find the court abused its discretion in its application of a 3.42 multiplier. The contract expressly provided for using the common fund doctrine to determine the award. The superior court had discretion in determining whether to use a multiplier in light of the contract, the law and the evidence presented. On the record before us, applying a 3.42 multiplier is not an abuse of discretion based on the risks of litigation and the success achieved.
¶40 We do, however, hold the superior court abused its discretion by applying a multiplier to post-settlement fees. A multiplier is not appropriate for time spent after settlement, when there is little risk of nonpayment.
See City of Detroit v. Grinnell Corp.,
*356 2. Lodestar Calculation
¶41 Microsoft argues that the superior court abused its discretion in its lodestar calculation because it did not adjust the rates for Arizona class counsel. Class counsel argues that Microsoft waived this point in oral argument in the court below when it made statements such as “we haven’t nit picked, particularly with respect to the Arizona counsel and the fees and the rates that they have requested.” While Microsoft preserved this argument by raising the issue in its opposition to Class Counsel’s fee application, on this record we cannot find the superior court abused its discretion on the rates calculated in the fee application.
¶ 42 The reasonableness of attorneys’ billing rates is based on the “prevailing market rate.”
Radish,
¶ 43 On appeal, Microsoft argues that the superior court abused its discretion by failing to adjust the billing rate of Arizona Class Counsel. 16 It points to the billing rates of specific attorneys Sehiffer, Jann and Wulkan as being significantly higher than the prevailing rate opined by their expert Peter Baird. 17 We agree that the rates charged by these attorneys are higher. However, we note that Sehiffer billed a total of 8.3 hours, Jann billed a total of 2.8 hours and Wulkan billed a total of .4 hours, whereas other attorneys billed over 600 hours. Accordingly, based on the record presented, we cannot find the superi- or court abused its discretion.
¶ 44 Additionally, Class Counsel presented the court with expert opinion that the hourly rates of Arizona Class Counsel were reasonable given “the prevailing hourly rates in the Phoenix ... communities.”
18
The superior court also adjusted the amounts requested by Arizona Class Counsel, albeit very slightly. Again, it would be preferable if the superior court had explained its adjustment.
Rawl-ings,
¶45 The superior court did err by awarding Class Counsel attorneys’ fees for the work they performed on the Bunker’s Glass litigation. Class Counsel argue that it is within the court’s “equitable power” to allow compensation for work done on other litigation. Wininger v. SI Mgmt. L.P., 301 F.3d 1115, 1121 (9th Cir.2002). However, as discussed previously, the attorneys’ fees awarded here were awarded by enforcing a Settlement Agreement, not solely under the equitable powers of the court. The Settlement Agreement does not allow for this type of recovery.
¶ 46 The Settlement Agreement specifies that Microsoft will pay attorneys’ fees for work performed by all counsel for plaintiffs in “All Cases in connection with this litigation.” “All Cases” is defined as Friedman v. Microsoft and Lucero v. Microsoft. Class counsel could have negotiated payment for work associated with Bunker’s Glass and included this into the agreement. It did not. Accordingly, we remand for proceedings to deduct class counsel’s award by the time billed for its work on Bunker’s Glass.
*357 Conclusion
¶47 The superior court acted within its discretion in awarding an enhanced lodestar. However, the court erred by applying the multiplier to post-settlement work and in calculating the lodestar by including fees for litigation in Bunker’s Glass that were not included in the Settlement Agreement. Accordingly, we affirm in part, reverse in part and remand for proceedings consistent with this opinion and the accompanying memorandum decision filed this date.
Notes
. Class counsel includes the Arizona firms of Shughart Thomson & Kilroy, Bonnett, Fairb-ourn, Friedman & Balint and the Irvine Law Firm. It also includes two California firms, Ler-ach Coughlin Stoia Geller Rudman & Robbins and Townsend and Townsend and Crew. The firms will be collectively referred to as class counsel, or Arizona class counsel or California class counsel.
. A lodestar figure is "the product of reasonable hours times a reasonable rate."
City of Burlington v. Vague, 505
U.S. 557, 559,
. In a separate memorandum decision, we affirm the superior court's order denying Townsend and Townsend and Crew's request that Microsoft be required to pay that firm additional money for Arizona class counsel’s use of that firm’s database.
. In Arizona, a number of actions were filed, including Friedman v. Microsoft and Lucero v. Microsoft. These class actions were consolidated into one class action by order.
. For more about the federal Microsoft antitrust action see
United States v. Microsoft,
. “All Cases” is defined in the settlement agreement as Friedman v. Microsoft and Lucero v. Microsoft.
. "Cy pres " is a derivative from French meaning "as near as.” Black’s Law Dictionary 415 (8th ed.2004). It is an "equitable doctrine under which a court reforms a written instrument with a gift to charity as closely to the donor's intention as possible, so that the gift does not fail.... It is also used to distribute unclaimed portions of a class-action judgment or settlement funds to a charity that will advance the interests of the class.” Id. In the context of a class action settlement agreement, when it is not feasible to distribute the class recovety or when there is a balance that remains after distribution, the court may direct "undistributed funds to be applied prospectively to the indirect benefit of the class.” 3 Alba Conte & Herbert Newberg, Newberg on Class Actions § 10.17 (4th ed. 2005) (“New-berg”). These funds are usually distributed to a third party for a specified purpose. Id.
. Thus, Microsoft began and concluded its arguments by stating that the risk of nonpayment did not justify the multiplier advanced by class counsel and that only a 1.5 multiplier was appropriate based on risk and certainly no multiplier above 2.3 was appropriate. Between that beginning and conclusion, Microsoft argued that other factors, such as skill, difficulty, and results of the litigation did not support a multiplier.
. Both parties cite to various unreported judicial decisions to support their arguments. We decline to rely on those decisions. See Ariz. R. Civ.App. Proc. 28(a)(2).
. None of the other cases relied upon by Microsoft urge a conclusion that enhancement is rarely applied in
common fund cases. See generally, Fischel, id.
(discussing the general proposition that enhancement is rare);
Kadish v. Ariz. State Land Dep’t,
. Thus, while Microsoft argues that the lodestar should not be enhanced because this is a "tag-along" or "coattail" action to the federal action against Microsoft, there were distinct differences between this action and the federal lawsuit.
. Microsoft does not contend the superior court should have used a percentage of recovery theory to award attorneys’ fees. When there is no true common fund created, a lodestar rather than a percentage of recovery is the means to determine the appropriate fee.
Weinberger v. Great N. Ne-koosa Corp.,
. The reason for such cross-checking is related to the history of attorneys’ fees awards in common fund cases. Initially, such awards were based upon a percentage of recovery. Use of the percentage method led to abuse and courts began to turn to use of a lodestar with possible enhancement by a multiplier based on the contingency or risk involved and the quality of work. Courts and commentators soon found problems with the lodestar method, including the burden placed on the courts, manipulation of hours and hourly rates and inconsistent decisions. This led to courts applying a reasonable percentage method, setting a benchmark percentage (usually between 20 percent to 30 percent of the settlement) or using lodestar as a cross-check. Pacold 68 U.Chi.L.Rev. at 1020-25; Vaughn R. Walker and Ben Horwich,
The Ethical Imperative of a Lodestar Cross-Check: Judicial Misgivings about “Reasonable Percentage” Fees in Common Fund Cases,
18 Geo.J.Legal Ethics 1453, 1455-64 (2005);
Ladewig,
. Microsoft takes issue with the superior court's failure to articulate why it chose to apply a multiplier at all. We find this issue to be waived by the nature of Microsoft’s opposition to the fee application. We will not require a trial court to explain an issue, whether a multiplier should be used at all, that was not in dispute before it. Moreover, for the reasons stated above, we believe the superior court adequately explained the reason it decided to apply a multiplier.
. For instance, the billing rate submitted by a named partner for California Class Counsel was $650 per hour and an estimated prevailing rate for the same level partner in Phoenix was $350 per hour.
. Those rates ranged from $115 to $400 per hour with the exception of three highly experienced attorneys whose rates were $460 per hour. Those three attorneys had between 32 and 39 years of experience each. Of the other Phoenix attorneys, only five had billing rates exceeding $350 per hour and each had no less than 23 years of experience.
. The billing rate requested by Sehiffer and Jann was $295 per hour while Wulkan’s requested rate was $225 per hour.
. The expert opinion provided by Class Counsel is not as analytical as the expert opinion provided by Microsoft. However, we cannot say that there was no evidence supporting the superior court’s decision on rates. Therefore, there is no abuse of discretion.