Richard Searlata, a market maker who formerly traded options at the Chicago Board of Exchange (CBOE), seeks a discharge in bankruptcy from a $4 million debt owed to the firm who cleared his accounts at the CBOE, Goldberg Securities, Inc. (Goldberg). The bankruptcy court barred Searlata from discharge under 11 U.S.C. § 523(a)(2)(A) but found § 523(a)(6) inapplicable to this case. In re Scarlata, 112 B.R. 279 (B.C.N.D.Ill.1990). The district court reversed the bankruptcy court’s finding under § 523(a)(2)(A), affirmed under § 523(a)(6), and granted Searlata a discharge. In re Scarlata, 127 B.R. 1004, 1011 (N.D.Ill.1991). On appeal, Goldberg argues that Searlata is not entitled to a discharge because 1) he tendered a check for which he did not have sufficient funds in his bank account, creating a false pretense that he had greater funds with which to trade; 2) he misrepresented his trading intentions, falsely stating that he would reduce his risk exposure; and 3) he willfully and maliciously injured Goldberg’s property by trading contrary to his representations and with Goldberg’s money. Although we differ with a portion of the district court’s opinion, we affirm its judgment granting Searlata the discharge.
Facts
Searlata was once a professional market maker at the CBOE. For four years, he traded options on an account through Goldberg. According to the various arrangements between Searlata, Goldberg, and the CBOE, Goldberg received fees on Scarlata’s transactions, as well as interest on money loaned to him; in exchange, Scar-lata received certain services. The most important of these was that Goldberg guaranteed Scarlata’s losses in excess of the equity in his account. As a result, Goldberg was potentially liable for 100% of Scarlata’s losses in excess of his equity, even though it was not entitled to receive a share of his profits. Despite its one-sided exposure, Goldberg, like other clearing houses, had no electronic means of controlling Scarlata’s trades once he was in the pit; when Searlata (or any other market maker) traded, Goldberg relied on him to adhere to a “haircut requirement” which obliges him to have funds in his account to cover potential losses from unliquidated outstanding securities positions. The haircut requirement thus limits traders’ ability *524to take positions in excess of their personal ability to cover trading losses. Goldberg considered a trader’s position to be risky if the trader had a haircut requirement over $100,000, had an equity deficit in his account, or had a haircut-to-equity ratio greater than 2-1. R. 2-2 at 19.1 Based on these three factors, Goldberg would monitor its traders at the end of every day; if the trader’s position was considered risky, the firm might require him to reduce his positions, deposit more capital in his account, or, in an extreme case, liquidate his holdings.
Cut to the morning of October 19, 1987— “Black Monday” — when the Dow Jones Industrial Average lost approximately 22% of its value. The Dow had dropped more than 100 points the previous Friday, leaving Scarlata, as well as numerous other traders, in precarious positions. Scarlata had only $22,000 of equity in his account, but held 84 naked short puts amounting to a $150,000 risk exposure.2 Despite this 7-1 ratio, Scarlata wanted to trade that day. He wrote Goldberg a check for $30,000, even though he did not have sufficient funds in his bank account to cover the $30,000. Scarlata also told Goldberg’s risk managers, the employees who monitor the traders, that he would reduce his positions. Yet Scarlata exponentially increased his exposure during the first rotation on the market floor. Over the course of the rest of the day, Scarlata never managed to reduce his positions. Although he began trading with an exposure of just over $100,000, he ultimately lost approximately $5 million.3 Because it had guaranteed Scarlata’s trades, Goldberg paid the bulk of these losses.
Now Scarlata is in bankruptcy, seeking a discharge from his debt to Goldberg. As discussed above, the district court granted the discharge, reversing the bankruptcy court’s findings that Scarlata had misrepresented his trading intentions and created a false pretense. The district court erred by applying the clear and convincing standard of proof, and the bankruptcy court’s finding that Scarlata misrepresented his trading intentions may not have been clearly erroneous. Nevertheless, Goldberg did not prove that it relied on Scarlata’s representation that he intended to reduce his positions. Further, we agree with the district court that Scarlata did not make a false pretense when he tendered the check. Finally, Goldberg has not explained how or whether the bankruptcy court and district court erred in concluding that Scarlata’s actions were not malicious. Thus, we affirm.
Analysis
In bankruptcy, “exceptions to discharge are to be constructed strictly against a creditor and liberally in favor of the debtor.” In re Zarzynski, 771 F.2d 304, 306 (7th Cir.1985). The burden is on the objecting creditor to prove exceptions to discharge. Minnick v. Lafayette Loan & Trust Co., 392 F.2d 973, 976 (7th Cir.), cert. denied, 393 U.S. 875, 89 S.Ct. 170, 21 L.Ed.2d 146 (1968). Although Goldberg may have proved that Scarlata misrepresented his trading intentions, we do not believe it carried its burden to prove reliance or to prove the applicability of the “willful and malicious injury” exception.
Count I — § 523(a)(2)(A)
The bankruptcy court held that Scarlata had made a “false pretense” by tendering a check for which he did not have *525any assets, and that he had misrepresented his trading strategy by promising to reduce his positions but immediately escalating them instead. 112 B.R. at 287. The district court reversed both holdings. We agree with the district court as far as the check is concerned. In Williams v. United States, 458 U.S. 279, 102 S.Ct. 3088, 73 L.Ed.2d 767 (1982), the Supreme Court held that knowingly passing a bad check is not a “false statement” within the meaning of 18 U.S.C. § 1014. The Court reasoned that “a check is not a factual assertion at all”; a-check “serve[s] only to direct the drawee banks to pay the face amounts to the bearer. ...” Id. 458 U.S. at 284, 102 S.Ct. at 3091. Although Williams involved a criminal statute and construed the word “statement”, its reasoning governs whether Scar-lata’s cheek was a false pretense in this civil bankruptcy case. See, e.g., In re Hunt, 30 B.R. 425, 438 (M.D.Tenn.1983) (“creditor cannot rely solely on the existence of an NSF check ... to establish a misrepresentation for § 523(a)(2) purposes”); In re Horwitz, 100 B.R. 395, 398 (B.C.N.D.Ill.1989) (acknowledging split on this issue, but persuasively noting that contrary cases have ignored Williams). Moreover, even if passing a bad check were a false pretense within § 523(a)(2)(A), Scarla-ta’s check was not bad. The day after writing the check, Scarlata withdrew sufficient cash from his credit cards and deposited it in his checking account. The check did not bounce.
As for Scarlata’s statement that he would reduce his positions, Goldberg needed to prove 1) that Scarlata made.a statement either knowing it to be false or with reckless disregard for the truth; 2) that in making this misrepresentation -Scarlata possessed “scienter, i.e., an intent to deceive” Goldberg; and 3) that Goldberg actually and reasonably relied upon the misrepresentation. In re Kimzey, 761 F.2d 421, 423 (7th Cir.1985). Because Scarlata was making a statement of future intention, it is possible that Scarlata’.s statement was true when made; intervening events may have caused his future actions to deviate from his prior intentions. W. Page Keeton et al., Prosser and Keeton on The Law of Torts § 109, at 764-65 (5th ed. 1984). Scarlata claims that he truly intended to reduce his positions, but that exceptional circumstances prevented him from doing so. The'bankruptcy court found that Goldberg proved Scarlata’s fraud by clear and convincing evidence. 112 B.R. at 286-87. The district court held that this conclusion was clearly erroneous because “[t]here is no evidence in the record to support the bankruptcy court’s conclusion that Scarlata had formulated a plan over the weekend to substantially increase the size of his short put position.” 127 B.R. at 1011. But the district court applied the wrong standard of proof; a few months before the district court’s opinion, the Supreme Court held that “the standard of proof for the dis-chargeability exceptions in 11 U.S.C. § 523(a) is the ordinary preponderance-of-the-evidence standard.” Grogan v. Garner, 498 U.S. at 279, -, 111 S.Ct. 654, 661, 112 L.Ed.2d 755 (1991).4 We are thus faced with a curious question: was the bankruptcy court’s conclusion that Goldberg proved the exception by clear and convincing evidence clearly erroneous even under a preponderance standard?
We need not answer this question. Even if Scarlata did misrepresent his trading intentions to Goldberg, Goldberg did not prove that it relied on Scarlata’s statement.5 The bankruptcy court held that Goldberg had proved reasonable and actual reliance, but it did not specify any facts in support of this conclusion. 112 B.R. at 284-85. To the extent that the bankruptcy court did explain its reasoning, the court stated that “[h]ad Goldberg known that Scarlata ... intended to in*526crease substantially rather than decrease or neutralize his existing position, it would have [sic] not have allowed him to trade freely that day." Id. at 285 (emphasis supplied). This holding was expressly, contradicted by the testimony of Goldberg’s senior risk manager, the only Goldberg witness on the subject of reliance. The risk manager admitted on cross examination that if Scarlata had given him a check for $30,000 on the morning of October 19 and said nothing further, he would not have prevented him from going on the trading floor. R. 2-2 at 86. The plaintiff brought little evidence of reliance to overcome this admission. On direct examination, the risk manager rather formulaically testified that he had “rel[ied] on the truthfulness of Mr. Scarlata’s statement.” R. 2-2 at 58-59. He further stated that he took Scarlata off his list of problem accounts after the representation. R. 2-2 at 44. Goldberg presented no evidence from its other employees to support its claim of reliance.
Furthermore, Goldberg did not present sufficient circumstantial evidence of reliance. As everyone involved recognizes, the morning of Black Monday was exceptionally hectic. Indeed, the risk manager considered it “the busiest morning [he had] experienced as a risk manager at Goldberg up to that point.” R. 2-2 at 84-85. In these harried circumstances, Scarlata was far from Goldberg’s greatest concern. Although he was on the risk manager’s list of problem accounts, so were 24 other traders, and most of them were in much more serious condition than Scarlata. Four had haircut deductions over $1,000,000; 5, over $500,000; 16 between $100,000 and $500,-000. R. 2-2 at 32-33. At least five traders had equity deficits, including deficits of $390,000 and $639,000. “[T]here were a lot of traders that morning that needed definite review.” R. 2-2 at 69-72. Considering the risk manager’s damaging admission that he would not have prevented Scarlata from trading even if Scarlata had not made the representation, we believe that Goldberg did not satisfy its affirmative burden of proving reliance.6 As a result, Scarlata is not barred from discharging his debt under § 523(a)(2)(A).
Count II — § 523(a)(6)
Goldberg also argues that Scarlata’s debt should be barred from discharge under § 523(a)(6), which denies a discharge for “willful and malicious” injury to property of another. The meaning of “malicious” is a subject of considerable conflict among lower courts as well as courts of appeals. Some courts have interpreted malicious to require an “intent to do harm”, while others have applied an implied or constructive malice standard derived from Tinker v. Colwell, 193 U.S. 473, 24 S.Ct. 505, 48 L.Ed. 754 (1904). See generally United Bank of Southgate v. Nelson, 35 B.R. 766, 770 (N.D.Ill.1983) (discussing split in authorities). This circuit has not taken a definitive stand on the question, and it is unnecessary for us to reach that issue of first impression here. Both the bankruptcy court and the district court held that Scarlata did not act maliciously because his acts would not “automatically or necessarily” cause injury to Goldberg. Scarlata’s job involved, at least to a certain extent, trading with Goldberg’s money. And if his *527gambles had paid off, Goldberg would not have been injured. 112 B.R. at 290; 127 B.R. at 1013. Thus, the lower courts apparently believed that Scarlata never acted maliciously to take an enormous risk at Goldberg’s expense, but became gradually embroiled in the crashing market.
Rather than explain whether and why the district court erred in concluding that Scarlata’s acts would not necessarily cause harm, Goldberg instead argues that the district court erroneously applied the specific malice standard rather than the implied malice standard. Goldberg urges us to adopt the implied malice standard which was adopted in Southgate and derived from Tinker. Appellant’s Br. at 40. But even if we were to adopt the South-gate standard, the conclusion of the bankruptcy and district courts would not be undermined. Southgate defined a willful and malicious injury as “a deliberate or intentional act in which the debtor knows his act would harm the creditor[]s interest and proceeds in the fact [sic] of that knowledge” (emphasis added). 35 B.R. at 776. In addition, the Southgate court explicitly adopted its implied malice standard from Tinker, which required that a willful and malicious act “necessarily cause[ ] injury.” 193 U.S. at 486-89, 24 S.Ct. at 509. Therefore, even the Southgate standard requires proof that the debtor took the sort of actions that would “automatically or necessarily” harm the creditor. See also In re Guy, 101 B.R. 961, 982 (B.C.N.D.Ind.1988) (“[M]alice may exist by implication or constructively because the debtor’s actions indicate he knew that his actions would result in injury.”); In re Condict, 71 B.R. 485, 487 (N.D.Ill.1987). As noted above, that is the standard the bankruptcy and district courts applied, and Goldberg has failed to give us a reason to overturn this holding.
The dissent contends that the law in other circuits supports barring Scarlata’s debts from discharge. We believe that the dissent reaches out to decide a difficult issue of first impression. Indeed, the dissent does an admirable job of attempting to discover what Goldberg perhaps ought to have argued.7 But Goldberg has not explained why the district court and the bankruptcy court erred in using the “necessarily causes harm” standard; Goldberg has affirmatively requested that we adopt a malice standard which uses the same language; and Goldberg has not developed any argument which would support analogizing this case to the typical cases that arise under § 523(a)(6).8 We will not reverse the district court’s decision when the appellants “have failed to identify, let alone explain, any error committed by the district court.” Sears, Roebuck & Co. v. The Murray Ohio Mfg. Co., 949 F.2d 226, 228 (7th Cir.1991); see also Luddington v. Indiana Bell Telephone Co., 966 F.2d 225, 230 (1992) (“if an appellant fails, to make a minimally complete and comprehensible argument for each of his claims, he loses *528regardless of the merits of those claims as they might have appeared on a fuller presentation”). We should not decide a difficult question of first impression under these circumstances.
Conclusion
Because Goldberg failed to prove that it relied on Scarlata’s misrepresentation, it may not bar the debt from discharge under § 523(a)(2)(A). And since Scarlata did not take the sort of actions that would necessarily harm Goldberg, and Goldberg has not clearly argued any alternative definition of malice, Scarlata is not barred under § 523(a)(6). Accordingly, the judgment granting the discharge is Affirmed. 9
. The record in this case is confusingly indexed. As a result, we have cited the trial transcript by date instead of docket number.
. Scarlata traded in puts, which are option contracts under which the holder of the option has the right to sell a specified number of shares at a particular price. A trader is "short" if he has sold more options than he has bought. If a trader is short, he or she hopes that the market will move up. Finally, a trader’s position is "naked” or "uncovered” if the trader does not hold countervailing positions which minimize his or her risk. By concluding that Friday with 84 naked short puts, therefore, Scarlata had made a large bet that the market would open higher on Monday morning.
.After making certain adjustments, the bankruptcy court calculated Scarlata’s debt to Goldberg at $4,019,091.87. In re Scarlata, 112 B.R. 279, 287 (B.C.N.D.Ill.1990).
. Although Grogan v. Garner had not been decided when the bankruptcy court made its decision, Scarlata does not argue that Grogan v. Garner should not be applied retroactively to this case. Appellee’s Br. at 20. Accordingly, we assume that the decision does apply.
. Although the district court did not reach the issue of reliance, we may affirm on any basis supported by the record. In re Memorial Estates, Inc., 950 F.2d 1364, 1370 (7th Cir.1991), cert. denied, - U.S. -, 112 S.Ct. 2969, 119 L.Ed.2d 589 (1992).
. The dissent would apparently presume reliance based on the relationship of the parties and what it considers to be the blatancy of Scarlata’s fraud. The dissent argues that Goldberg reasonably relied on Scarlata because his past history as a trader had led Goldberg to believe that he would trade conservatively. Dissent at 530-32. It may be true that relying on Scarlata’s past trading history was reasonable. Goldberg’s burden, however, was not to prove that it relied on Scarlata’s past history, but to prove that it "actually relied on the fake repre-sentation_” Kimzey, 761 F.2d at 423 (emphasis added). Goldberg did not present even minimal evidence that it relied on Scarlata’s representations. All Goldberg needed to do was have one of its risk managers testify that he would have done something differently had Scarlata not made the alleged misrepresentation. But the only risk manager to testify contradicted himself on this point, expressly admitting that he would not have done anything differently. Finally, Carini v. Matera, 592 F.2d 378 (7th Cir.1979), does not support the dissent's argument. Matera does not hold that the relationship between a person who defrauds and his victim can provide otherwise lacking evidence of reliance; it holds merely that once reliance has been established, the parties’ past relation*527ship can indicate whether that reliance was reasonable.
. The dissent cites two sentences from Goldberg's initial brief in which Goldberg squarely presented the issue we decline to reach. Dissent at 534. Unfortunately, Goldberg did nothing more than "present” that issue, and it pursued a different argument in its briefs.
. Section 523(a)(6) cases typically fall into one of two categories: (1) claims by a creditor that the debtor sold collateral subject to a security agreement, and (2) attempts by creditors to have previously-entered judgments against the debtor found non-dischargeable. Kimzey, 761 F.2d at 425. This case obviously does not fall into category (2), as Goldberg has not received a judgment against Scarlata in any other action. The dissent apparently believes that this case is similar to category (1): that in violating the haircut-to-equity ratio, Scarlata violated an agreement and deprived Goldberg of a "secured” interest. But Goldberg has not argued that the analogy of security agreement cases should apply, nor has it proven any of the subsidiary facts that would be necessary for us to come to that conclusion. For example, Goldberg has not proven that the haircut-to-equity ratio has the binding contractual effect of a security agreement. In addition, Goldberg has not proven that Scarlata had a duty to stay within the haircut-to-equity ratio at all points during the trading day; Scarlata’s trading activities were not monitored during the day. As a result, this case is very different from a breach of a security agreement, where the debtor breaches the agreement the moment he misuses the collateral. Therefore, a debtor who misuses collateral has breached the agreement even if the breach is ultimately profitable.
. This opinion has been circulated among all judges of this court in regular active service. A majority did not favor a rehearing in banc on the question of whether this decision has defined the term "malicious" as it is used in § 523(a)(6) of the Bankruptcy Code in such a way as to stand in conflict with other Circuits. Judges Coffey, Ripple and Manion voted in favor of rehearing in banc.