SYKES, Circuit Judge.

LaSalle Bank National Association ("LaSalle") appeals a district court order affirming the bankruptcy court's approval of a settlement of adversary litigation in the bankruptcy of Doctors Hospital of Hyde Park ("Doctors Hospital" or "the Hospital"). The settlement releases Dr. James Desnick ("Desnick") from adversary claims the Hospital brought against him, provides over $6 million in cash to the Hospital's bankruptcy estate, releases the Hospital from millions of dollars in claims against it, and ends a complex litigation. The Hospital, the bankruptcy trustee, Desnick, and the creditors' committee all agreed the settlement was in the best interest of the estate. LaSalle disagreed and objected. Following a lengthy hearing, the bankruptcy court held the settlement was in the best interest of the estate and approved it. The district court affirmed, and LaSalle has appealed. We affirm.

I. Background

Desnick was the owner and sole shareholder of Doctors Hospital and a number of other entities. One of his other companies, HPCH LLC ("HPCH"), owned the land on which the Hospital sat and collected monthly rent from it. Another of his companies, Medical Management of America, Inc. ("MMA"), managed the Hospital and received fees for these services. Desnick treated his companies like personal bank accounts, sometimes withdrawing money for himself, other times depositing money when his companies' coffers ran low.

Doctors Hospital guaranteed two loans that figure prominently in this litigation, though it enjoyed the proceeds of only one. In March 1997 MMA Funding (99% owned by Doctors Hospital) borrowed roughly $25 million from Daiwa Healthco (the "Daiwa loan"). Because the loan was* in practical effect, a loan to the Hospital, the Hospital secured the loan by pledging its receivables. In addition, Desnick personally guaranteed the loan. In August 1997 Nomura Asset Capital Corporation loaned HPCH $50 million (the "Nomura loan"). Although the Nomura loan went ostensibly to HPCH, it was secured by the Hospital's equipment and (like the Daiwa loan) by the Hospital's accounts receivable. The Hospital also executed a guaranty and surety-ship agreement in favor of Nomura. The Nomura loan proceeds did not go to the Hospital, however. Instead, the proceeds — some $48.5 million after administrative fees — were deposited into an account bearing the name of Desnick and his wife. Over time LaSalle Bank came to control the Nomura loan.

The Hospital filed for Chapter 11 bankruptcy protection in April 2000. Daiwa filed a claim against the Hospital to collect the outstanding portion of its loan, and Desnick personally paid the debt of about $9 million. The Hospital filed an adversary complaint against Desnick and numerous other defendants. Twelve of the other defendants were Desnick-controlled entities and four were former corporate officers or directors of the Hospital whom Desnick had effectively agreed to indemnify for their losses. The gist of the complaint was that Desnick and the other officers and directors caused the Hospital's bankruptcy through mismanagement and a series of fraudulent transactions — to the tune of about $34 million — which benefitted Desnick, his other companies, and Hospital management. The complaint asserted twenty-eight counts, including breach of fiduciary duties, conversion, violation of the Illinois Uniform Fraudulent Transfer Act, fraudulent transfers under the Bankruptcy Code, improper distributions to the shareholder, and equitable subordination of Desnick's claims against the Hospital (Desnick claimed the Hospital owed him roughly $16 million). The complaint also named LaSalle and sought to void the guaranty on the Nomura loan.

After two years of litigation, the Hospital moved the bankruptcy court for approval of a settlement agreement that had been reached by the parties (except LaSalle). Under the terms of the settlement, Desnick agreed to pay the Hospital roughly $6.1 million and also agreed to forfeit any subrogation rights he had to seek recovery of the $9 million he personally paid to Daiwa on behalf of the Hospital. He also agreed to withdraw any other claims he filed against the Hospital. Moreover, Desnick promised to use his best efforts to obtain dismissal or withdrawal of a $13 million claim against the Hospital filed by the Department of Health and Human Services ("DHHS") for Medicare/Medicaid reimbursements DHHS claimed were improperly paid to the Hospital. If he could not secure dismissal of the entire DHHS claim, Desnick agreed to pay 15% of the claim, up to $1.5 million. Finally, Desnick agreed to cooperate with the Hospital in its remaining claims against other defendants, including LaSalle, by (among other things) allowing the Hospital access to his expert. In exchange, the Hospital agreed to release from all claims Desnick, his companies, and the four individuals he agreed to indemnify.

LaSalle objected, and the bankruptcy court conducted a three-day evidentiary hearing on the Hospital's motion. The Hospital, Desnick, the bankruptcy trustee, and the creditors' committee all recommended that the settlement would be the best way to avoid protracted, expensive, complex, and uncertain litigation. The bankruptcy court concluded that the best-case scenario for the Hospital would be a $34 million victory and the worst case a $1.8 million victory. (LaSalle had argued that the high end was in excess of $80 million and the low end near $34 million; the bankruptcy judge thought that was unrealistic.) The court agreed with the Hospital and Desnick that the high-end $34 million victory was the least likely result because audited financials tended to support Desnick's position regarding the date of insolvency, which was critical to the bulk of the Hospital's claims. Given the range of litigation possibilities, the bankruptcy court concluded the settlement was reasonable because it would spare scarce resources, which might be wasted on protracted litigation that would likely yield mediocre results. LaSalle appealed to the district court, which affirmed.

II. Discussion

Bankruptcy courts may approve adversary litigation settlements that are in the best interests of the estate. In re Energy Co-op., Inc., 886 F.2d 921, 927-29 (7th Cir.1989); In re Am. Reserve Corp., 841 F.2d 159, 161 (7th Cir.1987). The linchpin of the "best interests of the estate" test is a comparison of the value of the settlement with the probable costs and benefits of litigating. In re Energy Coop., 886 F.2d at 927. Among the factors the court considers are the litigation's probability of success, complexity, expense, inconvenience, and delay, "including the possibility that disapproving the settlement will cause wasting of assets." In re Am. Reserve, 841 F.2d at 161. As part of this test, the value of the settlement must be reasonably equivalent to the value of the claims surrendered. This reasonable equivalence standard is met if the settlement falls within the reasonable range of possible litigation outcomes. In re Energy Co-op., 886 F.2d at 929; In re N.Y., New Haven Hartford R.Co., 632 F.2d 955, 960 (2d Cir.1980); see also Protective Comm. for Indep. Stockholders of TMT Trailer Ferry, Inc. v. Anderson, 390 U.S. 414, 424-25, 88 S.Ct. 1157, 20 L.Ed.2d 1 (1968); Depoister v. Mary M. Holloway Found., 36 F.3d 582, 586 (7th Cir.1994). Because litigation outcomes cannot be predicted with mathematical precision, only if a settlement falls below the low end of possible litigation outcomes will it fail the reasonable equivalence standard. In re Energy Co-op., 886 F.2d at 929.

The bankruptcy court's approval of the settlement is reviewed deferentially, for abuse of discretion. Depoister, 36 F.3d at 586. The bankruptcy court must independently evaluate the settlement, not simply accept the recommendation of the trustee. TMT Trailer Ferry, 390 U.S. at 424, 88 S.Ct. 1157; Depoister, 36 F.3d at 586-87; In re Am. Reserve, 841 F.2d at 162. If the decision demonstrates a command of the case, we will not engage in second-guessing; the bankruptcy court is in a better position "to consider the equities and reasonableness of a particular compromise." In re Am. Reserve, 841 F.2d at 162. Factual findings are reviewed for clear error; legal conclusions are reviewed de novo. FED. R. BANKR. P. 8013; In re Crosswhite, 148 F.3d 879, 881 (7th Cir.1998).

LaSalle complains of several errors, most having to do with the value of the settlement to the estate and whether it falls within the reasonable range of litigation outcomes. First, LaSalle maintains the bankruptcy court clearly erred in determining the value of the settlement to the estate. LaSalle argues that some of the claims Desnick gave up are worthless and that the settlement achieves practically nothing for unsecured creditors. Second, LaSalle contends the bankruptcy court miscalculated the range of litigation outcomes. According to LaSalle, both the low and high ends of possible outcomes are much higher than the bankruptcy judge thought, and the settlement does not fall within the range of LaSalle's higher projected outcomes. Finally, LaSalle argues the bankruptcy court did not make an independent evaluation of the settlement, but instead rubber-stamped the misguided recommendations of the trustee and the creditors' committee.

A. Value of the Settlement

The bankruptcy court held that the settlement had significant value to the estate. The Hospital released Desnick, twelve Desnick-controlled entities, and four former officers and directors of the Hospital from all claims against them. In return, Desnick paid the estate some $6.1 million cash. In addition, Desnick agreed to waive his claim against the estate for the $9 million or so he paid on the Daiwa loan. He also agreed to use his best efforts to help resolve the $13 million Medicare/Medicaid claim favorably to the estate. The settlement also made Desnick's financial expert available to the estate; he had detailed knowledge of the Hospital's books dating as far back as 1992, which could help the estate in its claim against LaSalle. Finally, the settlement narrowed the scope of litigation and allowed the Hospital to focus its resources on winning the few remaining claims rather than battling on all fronts.

LaSalle takes issue with several of the bankruptcy court's conclusions. First, LaSalle argues that Desnick's forfeiture of his claim for the Daiwa loan is worthless. LaSalle insists that when Desnick paid the Daiwa loan, the loan was extinguished and Desnick obtained no subrogation rights. LaSalle cites no law to support this argument, and the position is not defensible. When a guarantor pays a debt, he is subrogated to the rights of a creditor against the corporation on whose behalf he paid the debt. Weissman v. Weener, 12 F.3d 84, 87 (7th Cir.1993); Mid-State Fertilizer Co. v. Exch. Nat'l Bank of Chi, 877 F.2d 1333, 1336 (7th Cir.1989); see also Mid. Serv. Cas. Ins. Co. v. Elizabeth State Bank, 265 F.3d 601, 626 (7th Cir.2001) ("[S]ubrogation can arise simply from the fact of payment."). LaSalle suggests that any subrogation claim would be subject to equitable subordination based on Desnick's conduct, but it does not develop the argument. In any event, the point is immaterial because the bankruptcy court did not assume the value of the released claim to be worth $9 million to the estate. Rather, the court assumed that, if nothing else, the released claim saved the estate the expense of litigating to determine whether the claim was valid. That is not error.

LaSalle also argues Desnick's promise to use his best efforts to defeat the Medicare/Medicaid claim is of no value. By LaSalle's reckoning, Desnick already had a duty to obtain the release of the claim because of his position as director and sole shareholder of the Hospital. Here again LaSalle does not cite any legal authority for its position. Corporate directors do not have continuing fiduciary duties once they resign. See, e.g., Stand-age v. Planned Inv. Corp., 160 Ariz. 287, 772 P.2d 1140, 1144 (Ariz. 1988) (noting there is no more fiduciary duty when director or officer resigns); Microbiological Research Corp. v. Muna, 625 P.2d 690, 695 (Utah 1981) ("When a corporate officer ceases to act as such, because of his resignation or removal, the fiduciary relationship ceases."); Renpak, Inc. v. Oppenheimer, 104 So.2d 642, 644 (Fla.Dist.Ct.App. 1958); WILLIAM MEADE FLETCHER, 3 FLETCHER CYC. CORP. § 860 (2002). Desnick resigned as director of the Hospital in November 2000, so he had no continuing duty to right the ship in 2004. Moreover, Desnick agreed to pay for the filing of a report with DHHS regarding the Medicare/Medicaid claim — a report that could cost $500,000 — and he had no duty to do so personally, even in a sole shareholder situation (at least not without veil piercing, which LaSalle has not pressed). Cf. Kelly v. Fahrney, 145 Ill.App. 80 (1908) (noting in the context of a closely held corporation that "there is no duty on the part [of shareholders] to use individual pecuniary means to assist the corporation in its money difficulties or by use of such means shield it from financial destruction"). Finally, Desnick agreed to pay 15% of the claim, up to $1.5 million of his own money, should he not secure its release. So even assuming he had some sort of duty to help secure the release of the Medicare/Medicaid claim, Desnick sweetened the deal by offering to pay for the filing of the DHHS report and a portion of any remaining claims. Desnick's promise was hardly illusory.

B. Range of Litigation Outcomes

The date the Hospital became insolvent was critical to determining the range of possible litigation outcomes. Most of the Hospital's claims against Desnick centered on allegations of breach of fiduciary duty, fraudulent transfer, and conversion. Everyone agrees there are no viable claims for any period during which the Hospital was solvent; that is because as long as a corporation is solvent, directors typically owe fiduciary duties only to shareholders. Beach v. Miller, 130 Ill. 162, 22 N.E. 464, 466 (Ill. 1889); Paul H. Schwendener, Inc. v. Jupiter Elec. Co., 358 Ill.App.3d 65, 293 Ill.Dec. 893, 829 N.E.2d 818, 828 (Ill.App.Ct. 2005). Because Desnick was the sole shareholder of the Hospital, he can hardly have defrauded himself or breached a fiduciary duty to himself. See cf. In re Tufts Elecs., Inc., 746 F.2d 915, 917 (1st Cir.1984) (holding that corporate opportunity doctrine does not apply against sole shareholder who cannot defraud or conceal information from himself). If the Hospital was insolvent, however, Desnick's fiduciary duties extend to the Hospital's creditors. Paul H. Schwendener, 293 Ill.Dec. 893, 829 N.E.2d at 828. Likewise, the Hospital's fraudulent transfer and conversion claims depend on insolvency. See 740 III. COMP. STAT. 160/6 (1996); Nostalgia Network, Inc. v. Lockwood, 315 F.3d 717, 719 (7th Cir.2002) ("When a person transfers money or other property to another person without receiving anything in return, and the transferor is insolvent (or made insolvent by the transfer), the transfer is voidable. . . ."); Dannen v. Scafidi 75 Ill.App.3d 10, 30 Ill.Dec. 899, 393 N.E.2d 1246, 1251 (Ill.App.Ct. 1979) (no liability for converting corporate funds if all shareholders ratify the act and creditors are not impaired).

The Hospital maintains that it became insolvent in January 1997. Between January and August 28, 1997 (the date of the Nomura loan), some $23 million in transfers moved from the Hospital to Desnick. After August 1997, no more than about $10 or $11 million in transfers occurred. The bankruptcy court, sensibly enough, determined that the upper limit of possible litigation outcomes was about $34 million, assuming the Hospital could prove insolvency as early as January 1997 and would prevail on all of its claims.

LaSalle contends the bankruptcy judge's upper limit does not account for all of the Hospital's claims. According to LaSalle, the bankruptcy court should have included a claim for the $48 million Nomura loan because Desnick had pledged the Hospital's assets to secure it. That claim would have raised the upper limit of possible outcomes to nearly $82 million, and LaSalle contends the Hospital had a viable alter ego or "LBO-like" (leveraged buyout) theory that would permit recovery of the amount of the Nomura loan. The Hospital's lawyer disagreed. He testified that he entertained this theory, but ultimately considered it too far-fetched.

The bankruptcy court did not clearly err by leaving the Nomura funds out of the calculation. The Court noted that the adversary complaint did not actually make a claim against Desnick for the $48 million Nomura loan. The Hospital had only secured the loan; it never had a right to the loan proceeds, which were given to HPCH. Nor had the proceeds ever passed through the Hospital's bank accounts. The Hospital wanted only to be off the security hook; it did not have a claim to recover the money from Desnick.

From the estate's perspective, trying to get the Nomura money from Desnick on a novel legal theory would have been more trouble than it was worth. To do so would also run the risk of undercutting its claim against LaSalle to void the Nomura guaranty; it would essentially confirm LaSalle's claim against the estate for the Nomura funds. By disavowing any recovery against Desnick on the Nomura loan and seeking only to void the guaranty, the Hospital looked to knock out $48 million (or more) in claims. Perhaps the estate could have pressed the legal theories necessary to win the Nomura funds, but it made the strategic decision not to. The bankruptcy court's exclusion of the Nomura loan amount was not clearly erroneous.

The bankruptcy court determined the low end of possible outcomes by figuring the result if Desnick's defenses were established* which necessarily included an assessment of whether those defenses were any good. Desnick claimed the Hospital was not insolvent until September 1998. He also claimed to have given back to the Hospital some of the funds he transferred out. If his theories and proofs held up, Desnick maintained he could be liable for no more than $1.8 million. The bankruptcy court thus placed the low end of possible outcomes at $1.8 million.

LaSalle maintains the district court erred in this determination as well. At the low end, LaSalle argues, the Hospital could have recovered at least the $34 million because Desnick's defenses are not viable. We disagree. The parties do not dispute that insolvency is more difficult to prove further back in time, and the Hospital's position on insolvency is further back than Desnick's. LaSalle has effectively conceded that Desnick's claims have some merit — it asserted in the bankruptcy court that the Hospital was solvent until at least August 28, 1997 (the date of the Nomura loan). By taking that position LaSalle concedes some $23 million of the Hospital's claims against Desnick are out, leaving claims for only $10 to $11 million (the money transferred after the date of the Nomura loan). LaSalle's argument also makes no allowance for Desnick's transfers back to the Hospital. Moreover, the bankruptcy court concluded Desnick's litigation position was not only arguably defensible, but supported by audited financial statements. LaSalle's argument rings hollow in the face of its own concession on insolvency dates and its failure to account for money Desnick already repaid. It was not clear error to conclude that the lowest possible outcome for the Hospital was a recovery of $1.8 million.

As we have already noted, so long as the settlement falls within the reasonable range of litigation possibilities, it meets the reasonable equivalency standard. In re Energy Co-op., 886 F.2d at 929. Litigation outcomes are uncertain, not reducible to mathematical formulas. All kinds of variables can lead to all kinds of results. The sum certain of $6.1 million, plus the release of claims by Desnick against the Hospital, even if small compared to the best possible litigation outcome, reasonably represents a positive outcome for the estate in light of the risks, specifically the possibility that the Hospital would be found solvent up until late August 1997. This settlement falls squarely within the reasonable range of litigation outcomes.

As a last resort, LaSalle argues that the proponents of the settlement did not actually prove the range of possible outcomes because they did not present final insolvency analyses. Instead, Desnick and the Hospital presented only estimates and preliminary solvency evaluations. LaSalle argues the bankruptcy court could not have made an informed decision on the limits of litigation. But estimates are acceptable in this context; no one would ever settle a case if the claim and amount of recovery could be established with 100% certainty. By settling the parties have concluded that a definite amount today is more valuable than the risk and expense of trying the case in the future. Parties assess settlement offers by taking stock of the information they have, even if preliminary, and considering whether the possibility the case will improve is worth the risk of losing the settlement offer. In this case, the bankruptcy court considered the preliminary insolvency reports and noted that the audited financials lent support to Desnick's position. That is enough to show the settlement was good for the estate — it was more than three times higher than the worst-case scenario for the estate, to say nothing of the value of the claims Desnick surrendered and the money saved by foregoing additional litigation.

LaSalle's position, if accepted, would encourage waste. Preliminary evidence suggested the viability of some of Desnick's positions. Suppose the bankruptcy court had refused to approve the settlement until the parties brought in their final reports on insolvency. It is unrealistic to think that positions would have changed significantly (Desnick's expert was not likely to come back to court and say that after further research he realized the Hospital was insolvent in January 1997). The parties would have spent more time and money in litigation, and Desnick may have become less willing to settle for $6.1 million because his expert report and the audited financials supported his case. The Hospital would have spent more to gain less. And if the case should go to trial, all bets would be off; one side stands to lose nearly everything. The point of a settlement is that everyone stops gambling and walks away from the table with some winnings (or at least fewer losses). Estimates and preliminary solvency reports were sufficient to support the bankruptcy court's conclusion that this settlement was within the reasonable range of litigation outcomes.

C. Value of the Settlement v. Costs and Benefits of Litigation

At bottom, the best-interests-of-the-estate test is a comparison between the value of the settlement and the probable costs and benefits of litigating. In re Energy Coop., 886 F.2d at 927. As we have noted, in making this comparison, the bankruptcy court considers the litigation's probability of success, complexity, expense, inconvenience, and delay, including the wasting of assets that comes from extended litigation. In re Am. Reserve, 841 F.2d at 161. We have already touched on most of these factors. We must also be satisfied that the bankruptcy judge independently evaluated the settlement and did not simply accept the recommendation of the trustee. Depoister, 36 F.3d at 587.

LaSalle argues the bankruptcy court could not have independently concluded the settlement was in the best interests of the estate because the court was not given the proper information. LaSalle contends the Hospital improperly compared the value of the settlement to the value of the creditors' claims against the estate (between $10 and $14 million). With improper comparative information, LaSalle argues, the bankruptcy court could not have properly exercised its discretion.

There is no basis in the record to support LaSalle's argument. The bankruptcy court's opinion discusses at length the possible outcomes of the litigation between the Hospital and Desnick without reference to the validity or value of the creditors' claims against the estate. The bankruptcy court specifically disclaimed any consideration of the creditors' claims: "[T]he court need not determine whether . . . claims [filed by creditors against the Hospital] are valid or not. It must determine the reasonable range of likely outcomes of the litigation against these settling parties . . .." (emphasis added).

LaSalle also notes that only about $1.7 million of the settlement will be available to pay unsecured creditors and argues that such a paltry sum can hardly be in the best interests of the estate when the Hospital stood to recover over $80 million. We have already rejected LaSalle's contention that the Hospital could have recovered over $80 million; the bankruptcy court reasonably concluded that even $34 million was a stretch. The preliminary solvency reports and audited tended to support's position. Given the expense and uncertainty of litigation and the possibility of a poor outcome, this settlement is hardly bad for the unsecured creditors. There is now $1.7 million more for them to recover, and the estate still had live claims to press against other defendants.

LaSalle cites In re Remsen Partners, Ltd., 294 B.R. 557 (Bankr.S.D.N.Y. 2003), to support its argument about the effect of the settlement on unsecured creditors. But Remsen does not stand for the proposition that courts should reject out of hand settlements that leave little for unsecured creditors. Id. at 566-67 ("This does not mean that a settlement must be denied if it would not result in a recovery by general unsecured creditors."). The Remsen opinion merely noted that little or no benefit for unsecured creditors will cause a court to look more carefully at the trustee's recommendation. Id. at 567. In Remsen, the trustee had not alerted creditors that they would receive no benefit from the proposed settlement. The bankruptcy court concluded that the trustee's failure to provide this information gave rise to an inference that more creditors would have objected had they been informed. Id. The same inference does not arise here. The creditors' committee, whose views count but are not controlling, see In re Am. Reserve, 841 F.2d at 161-62, was satisfied with this settlement even knowing only $1.7 million would be available to unsecured creditors.

LaSalle also attacks the settlement for releasing the sixteen other defendants (besides Desnick) when none of them paid a penny. Of the sixteen other defendants, however, twelve were Desnick-controlled entities and the other four were former directors or officers of the Hospital whom Desnick had agreed to indemnify for then-losses. Not surprisingly, release of these sixteen other defendants was one of Desnick's terms for settling. The claims against the four officers and directors of the Hospital were identical to the claims against Desnick. If the estate won on its claims against Desnick and the officers, it would recover only once, likewise if it won against either Desnick or the officers. Desnick could hardly be expected to settle the claims against himself and remain on the hook for the same claims against the others.

Moreover, like the claims against Desnick, the claims against the officers and directors depended on proof of insolvency, and the bankruptcy court's evaluation of the apparent strength of Desnick's defenses applies equally to the four officers and directors. There was no significant additional recovery to be had against the sixteen additional defendants released, and if it had not released them, the Hospital could not have settled with Desnick.

Finally, LaSalle contends the bankruptcy court relied too heavily on the trustee's recommendation and that of the creditors' committee. That is simply not true. The bankruptcy court's decision reflects a command of the facts and the law and an independent and informed decision about the value of the claims surrendered and the likelihood of success in litigation. Although the judge commented on the recommendations of the trustee and the creditors' committee, there is nothing in the record to support LaSalle's assertion that the court simply "rubber-stamped" those recommendations. The bankruptcy court's approval of the settlement was not an abuse of discretion.

AFFIRMED.