Collins

1992 T.C. Memo. 478, 64 T.C.M. 557, 1992 Tax Ct. Memo LEXIS 503

• Court: United States Tax Court
• Decided: August 24, 1992
• Docket: Docket No. 11579-90
• Status: Unpublished

Opinion

MARK D. COLLINS, Petitioner v. COMMISSIONER OF INTERNAL REVENUE, Respondent

Collins

Docket No. 11579-90

United States Tax Court

T.C. Memo 1992-478; 1992 Tax Ct. Memo LEXIS 503; 64 T.C.M. (CCH) 557;

August 24, 1992, Filed

---

Decision will be entered under Rule 155.

In one afternoon in July 1988, P, a ticket seller at a New York Off-Track Betting (OTB) parlor, punched up betting tickets for himself on his computer terminal without paying for them. The tickets, which evidenced P's bets on nine horseraces, had a total face amount of $ 80,280; P's tickets for the last two races resulted in winnings of $ 42,175 and a net loss of $ 38,105 for the day. P did not remove any cash from his cash drawer, and transferred his winning tickets to OTB when he turned himself in at the end of the day. In October 1988, P pleaded guilty to grand larceny in the third degree. In 1989, P suffered a judgment obtained by OTB's insurance company on the claim the insurer had paid OTB in partial satisfaction of OTB's loss from P's activities.

1. Held: P realized gross income from theft that did not also constitute income from wagering transactions.

2. Held, further, P's gross income from theft is equal to the fair market value of the tickets he received. The fair market value of the tickets is determined to be $ 80,280.

3. Held, further, P is entitled to a deduction of $ 42,175 for the winning tickets he transferred to OTB. Sec. 165(c)(2), I.R.C.

4. Held, further, P is not entitled to deduct the $ 38,105 loss from his theft income as a loss from wagering transactions, sec. 165(d), I.R.C., nor as a loss from a transaction or transactions entered into for profit, sec. 165(c)(2), I.R.C.

5. Held, further, P is not liable for additions to tax for negligence or substantial understatement of income tax. Secs. 6653(a)(1), 6661, I.R.C.

For Petitioner: James B. Lewis.

For Respondent: Raymond A. Kahn.

BEGHE

---

BEGHE

MEMORANDUM FINDINGS OF FACT AND OPINION

BEGHE, Judge: Respondent determined a deficiency in petitioner's 1988 Federal income tax of $ 9,376, plus additions to tax of $ 469 under section 6653(a)(1)¹ and $ 2,344 under section 6661.

The issues for decision are whether petitioner had taxable unreported income from gambling or theft during 1988, and whether he is liable for the additions to tax for negligence and substantial understatement of tax liability. We hold that petitioner had unreported income, but that he is not liable for additions to tax.

FINDINGS OF FACT

We find and incorporate the stipulated facts and exhibits. We also take judicial notice of the results, as published in the Daily Racing Form (Eastern ed. July 20, 1988) (reproduced in Appendix A to this opinion), of the thoroughbred horseraces petitioner bet on that are the subject of this case.

Petitioner's Theft and Gambling Activities

Petitioner resided in Auburn, New York, at the time he filed his petition. During 1988 petitioner was employed as a ticket seller at the Off-Track Betting (OTB) parlor in Auburn. OTB operates a statewide network of betting parlors that allow patrons to place legal bets on horseraces run in New York State (as well as certain races run elsewhere) without going to the track. Patrons also may cash in winning tickets at OTB.

Petitioner was a compulsive gambler who, while placing bets for OTB patrons in the course of his employment, could not resist the temptation to gamble without paying for the privilege of doing so. On Sunday, July 17, 1988, petitioner decided, as he had on several prior occasions, that he "would like some money" and began placing bets on his own behalf. He punched the bets into his computer terminal, which entered his bets and printed the tickets. ² When he had done so on prior occasions, petitioner either ended up ahead or lost a small amount, which he covered without being detected by anyone at OTB.

Petitioner appears to have employed a variant of a technique for betting on horseraces in which he bet on the favorite (the horse with the lowest odds) in each race, increasing the amount of each subsequent bet until he would pick a winner, which would recoup his prior losses and provide a modest surplus. This time, however, petitioner encountered a series of losses that caused his gambling fever to spiral out of control.

He started by placing three $ 20 bets on one horse in the first race at Finger Lakes Race Track. He bet on this horse to win, place (finish second or better), and show (finish third or better), for total bets of $ 60. This horse did not finish among the top three. Petitioner then tried to recoup his $ 60 loss by betting $ 40 in the second race on one horse to win, place, and show, for total bets of $ 120. This horse also failed to run in the money.

Petitioner bet a total of $ 600 on the third race, repeating the pattern of dividing his bets equally among win, place, and show bets on the same horse. At this point, petitioner was trying to recoup the losses he had caused his employer, which he knew he would be responsible for. He also intended to appropriate any net winnings. He lost again.

Petitioner bet a total of $ 1,500 on the fourth race and lost. He skipped the fifth race, but bet another $ 1,500 on the sixth race and lost. He bet $ 7,500 on the seventh race and $ 15,000 on the eighth. Both times he lost.

With two races left that day, petitioner owed OTB $ 26,280. He bet $ 25,500 on the ninth race. His horse came in third, so one of his three $ 8,500 bets paid off, but it paid only $ 8,925.

By the 10th race, petitioner was down $ 42,855. He made three bets totaling $ 28,500 on one horse in the 10th race. He finally caught the winner, which paid a total of $ 33,250 on his three bets. But petitioner still was behind $ 38,105, and the racing day was over. Petitioner's bets had totaled $ 80,280.

At the end of the racing day, petitioner deposited the winning tickets of $ 42,175 in his cash drawer and told his supervisor, Debbie D'Angelo, that he had "screwed up". D'Angelo called the police, who took petitioner into custody. Petitioner signed an affidavit, prepared by the police, admitting his actions.

After a terminal operator completes his shift, his terminal prints out a summary of sales and redemptions at the terminal, which the terminal operator or his supervisor uses to prepare a settlement sheet. The settlement sheet shows the difference, if any, between the amount of cash in the drawer and the amount of cash that should be there. Petitioner's settlement sheet for July 17, 1988, shows that he began his work shift that day with $ 805 in his cash drawer and ended the day with $ 843.40 in the drawer. His terminal recorded "sales" of $ 84,137. Inasmuch as $ 80,280 of sales were petitioner's bets, legitimate sales amounted to $ 3,857. Petitioner also took in telephone deposits of $ 29. ³ Inasmuch as the total of petitioner's cash inflows (including his opening balance) was $ 4,691, he never had more than that amount of cash in his drawer. On an ordinary day, petitioner could have expected to process between $ 1,000 and $ 3,000 in bets, all placed in cash.

Petitioner also had to pay out on customers' winning bets. His terminal indicated $ 45,951.60 had been "paid" during his shift. Inasmuch as $ 42,175 of the payouts were actually the winning tickets petitioner turned over to OTB without receiving cash for them, legitimate payouts were $ 3,776.60. Petitioner never took any cash from OTB. Nonetheless, petitioner's terminal indicated that the drawer was $ 38,135.80 short.

On October 27, 1988, petitioner pleaded guilty in Cayuga County Court to one count of grand larceny in the third degree, a felony in New York defined as theft of property worth more than $ 3,000. He was sentenced to 5 years' probation, 150 hours' community service, and a $ 100 surcharge.

OTB's Policies on Theft and Employee Betting

OTB is a strictly cash business. Bets on credit are absolutely prohibited. OTB has an equally strict policy against employee gambling on any race within New York, and on any other race for which OTB accepts a bet. Employee betting is a major concern to OTB. Employees are indoctrinated during initial training not to bet and OTB frequently and regularly repeats this message in oral and written communications to its employees.

In numerous instances, patrons have reported OTB employees for betting; OTB investigates these charges and takes disciplinary action against employees who have violated the policy. All OTB employees are duty bound to notify their supervisor when a fellow employee violates the rule against betting. Within limitations set by a collective bargaining agreement, an employee who violates OTB's policy, even with an otherwise legal bet, may be disciplined. In addition, if an employee causes a loss to the corporation, OTB is authorized by State regulations to deduct the amount of the loss from the employee's wages.

To prevent employee theft, OTB's internal auditors conduct random spot checks, sometimes during business hours, to determine whether employees' cash drawers are in balance. OTB also uses undercover investigators to detect employee theft. OTB's total shortages from all sources, including theft, amount to far less than 1 percent of the $ 1.9 billion per year wagered at OTB.

Despite OTB's concerns about theft and employee betting, its employees are not under constant surveillance by their supervisors. OTB's computer system did not have the capability to allow petitioner's supervisor, D'Angelo, to monitor from her location in an adjoining office the entries petitioner was making at his computer terminal. During petitioner's work shift, D'Angelo would not have been aware of unusual activity at his computer terminal unless she could have observed the activity directly. A ticket seller generally has contact with a supervisor at the beginning and end of his shift, if a patron presents a particularly large winning ticket for payment (such as one requiring Federal income tax withholding), or if the employee's cash drawer becomes too full of money or needs to be refilled. Neither D'Angelo nor any of petitioner's fellow employees noticed anything out of the ordinary on the fateful day until petitioner turned himself in. Of course, if petitioner had not turned himself in, his day-end terminal print-out would have revealed his actions to OTB.

Petitioner's actions caused OTB to claim $ 37,588 against its theft insurance policy with Hartford Accident & Indemnity Co., Inc. Hartford paid the claim in full, less a $ 5,000 deductible, and obtained a judgment against petitioner in December 1989 for the amount of its payment, plus costs, totaling $ 36,601.94. Hartford has attempted to enforce its judgment, with little success, inasmuch as petitioner has neither the income nor the assets to satisfy the judgment. At no time did petitioner sign a promissory note or consent to entry of judgment against himself.

The Pari-mutuel System

Petitioner's actions not only caused a shortfall in his cash drawer at OTB. They also had a ripple effect that changed the odds and the payoffs to all bettors who made legal single bets ⁴ on any of the races (other than the fifth) at Finger Lakes that day, whether they bet through OTB or at the track. To understand why, we must explain how pari-mutuel ⁵ gambling works.

The only legal gambling on horseraces in New York is pari-mutuel wagering, in which all bets on each type of wager for each race are pooled into a single fund, called a mutuel pool. For instance, the sum of all bets on all horses "to win" a particular race is placed in a single statewide pool, regardless of whether the bets are placed at OTB or the track. Place and show bets have their own pools. On single bets, 17 percent of the pool is taken off the top to pay the track and a State fund for horsebreeding and horserace promotion. The money remaining in the pool is divided among the winners in proportion to the amounts they bet. Bettors at OTB are also subject to a 5-percent service charge on their winnings.

A bet placed through OTB is electronically transmitted to the track when it is punched in by the ticket clerk. The bettor can cancel the bet and receive a refund at any time until 60 seconds before the race. He can receive such a refund at any OTB parlor in the same geographic region as the parlor that sold the ticket. ⁶ If the bet is not canceled, OTB becomes liable as a matter of law to transfer the amount of the bet to the appropriate mutuel pool. OTB cannot avoid liability on the ground that the ticket was issued in error or without authority. The liability generally is settled at the end of each day with wire transfers from OTB to the track. The track in turn reimburses OTB from the mutuel pool when OTB pays off a winning ticket.

A betting ticket does not indicate to whom it was issued, and the holder is entitled to cash the ticket without proving that he lawfully placed the bet shown on the ticket. Thus, a ticket functions as a bearer instrument, although it is illegal to resell it. If the ticket evidences a winning bet, the holder can cash it at any time until March 31 of the following year. If the bet was a loser, the ticket becomes worthless.

The payoff on a winning bet depends on the amount in the pool and the total bet on the winning horse (or, in the case of place and show bets, horses). As the track and OTB accept new bets before the race, the potential payoffs to all bettors constantly change. As a general rule, as more money is bet on a particular horse, the potential payoff per dollar of bets on that horse decreases. The odds on "win" bets, as well as the amounts in each of the place and show pools, are displayed on toteboards at the track and at OTB, so that horseplayers can determine whether the potential payoff justifies the risk of making a bet.

When the results of a race become official, the payoff to winning bettors (expressed as the amount of the payoff per $ 2 bet) is calculated. For win bets, the payoff is the amount in the pool divided by the total winning bets and multiplied by 2. If $ 1,205 is bet on a particular race, 17 percent, or $ 205, is taken off the top and $ 1,000 goes into the pool that pays the winners. If $ 500 of the original $ 1,205 is wagered by all bettors on the winning bet, each winning $ 2 bettor is entitled to 2/500th of the $ 1,000 remaining in the pool, or $ 4 ($ 3.80 after the OTB service charge on winnings). If a greater percentage of money in the pool backs losing horses (e.g., $ 955 bet on the losers, with only $ 250 bet on the winner), the winners enjoy a correspondingly greater return (e.g., 2/250ths of $ 1,000, or $ 8 ($ 7.60 after the service charge) per $ 2 bet). The size and composition of the pool thus determines the payoffs to winning bettors.

While the place and show bet calculations are more complex, the payoff is likewise determined by the amount of money in the pool and the amount bet on each of the two or three horses that place or show. Bettors on each horse that places or shows divide among themselves the amounts bet to place or show on the horses that finish out of the money. As with win bets, the potential payoff per dollar bet on a horse goes down as more money is bet on that horse.

By increasing the amounts in the pools, each of petitioner's bets lowered the odds and potential payoffs on the horse he bet on and correspondingly increased the odds and potential payoffs on the other horses in the same race. ⁷ Petitioner lost all his bets in the first seven races he bet on, so his bets enlarged the pool of funds available for the winners. However, petitioner's winning bets in the last two races reduced the amounts available to pay legitimate winners. All of petitioner's bets, win or lose, increased OTB's liability to the mutuel pools.

Petitioner's Return

Petitioner employed H&R Block, a nationwide tax return preparation firm, to prepare his 1988 tax return. H&R Block did not ask petitioner, and he did not volunteer, whether he had stolen anything of value during the taxable year. Petitioner reasonably believed that the events described above did not result in income to him, and that they in fact resulted in a substantial loss. Petitioner's timely filed 1988 Federal income tax return, which reported $ 11,980 in gross income from various employers, did not include any of the additional income determined by the notice of deficiency.

OPINION

This case requires us to return to first principles to pursue the often elusive answer to the fundamental question, "What is income?" But before we embark on that quest, we consider two preliminary matters.

Preliminary Matters

Amount in controversy. Respondent's notice of deficiency determined that petitioner realized $ 38,136 in "gambling income". While respondent does not forcefully argue on brief that this amount was gambling income, respondent maintains the correctness of the amount, although it is $ 31 greater than petitioner's net loss from his wagers and is in error to that extent. The error was caused by the $ 30.80 shortfall in petitioner's cash drawer. The following table shows what the flows through petitioner's cash drawer would have been, subtracting the results of his theft and personal gambling activities:

Opening balance	$ 805.00
Legitimate sales	3,857.00
Telephone deposits	29.00
Total inflows		$ 4,691.00
Cancellations	(40.00)
Legitimate payouts	(3,776.60)
Total outflows		(3,816.60)
Correct closing balance		874.40
Less actual closing balance		(843.60)
Difference		$ 30.80

Petitioner would have been short $ 30.80 regardless of whether he gambled. The only way, other than gambling, that the $ 30.80 could have been income to petitioner would have been theft of cash, but the parties have stipulated that he did not take cash. The most plausible explanation of the shortfall is that petitioner made a mistake in selling and redeeming tickets. The mere fact that petitioner mishandled the funds in his drawer does not mean he had income in the amount of the shortfall in his closing cash balance. See Hobson v. Commissioner, T.C. Memo. 1992-312. At most, petitioner had $ 38,105 in net income.

Burden of proof. Respondent's answer adopts the alternative theory that petitioner had theft or embezzlement income of $ 80,280, less the $ 42,175 in winning bets that petitioner constructively repaid to OTB, for net income of $ 38,105. James v. United States, 366 U.S. 213 (1961). Under respondent's new theory, petitioner also made bets of $ 80,280. The tickets petitioner punched up were worth $ 42,175 at the end of the day. Petitioner's bets resulted in a $ 38,105 wagering loss, which, respondent argues, petitioner cannot deduct because he did not have additional wagering income. Sec. 165(d).

Petitioner argues on brief that respondent's new theory asserted in the answer is new matter on which respondent has the burden of proof. Rule 142(a). Because we find that a preponderance of the evidence shows that petitioner had $ 38,105 in net theft income, placement of the burden of proof does not affect the outcome, and we need not opine whether respondent or petitioner has the burden.

Did Petitioner Have Income?

We next deal with the theory of the statutory notice that petitioner had $ 38,136 of "gambling income". Respondent has not strenuously pursued this theory. We conclude that petitioner did not have net income from gambling during 1988. ⁸ Petitioner bet $ 80,280 and finished with winning tickets worth $ 42,175, which he transferred to his employer. Petitioner's $ 38,105 shortfall is not gambling income, but a gambling loss.

The more difficult question is whether the $ 80,280 of bets placed by petitioner resulted in embezzlement or theft income. Petitioner argues that the amount of bets he placed was not income to him because he never had control of money or property in that amount.

The outer boundaries of the notion of income for tax purposes are not well defined, although the core concepts have been evolving over the past 80 years. We recently had occasion to consider the boundaries of the income concept:

Respondent asserts that petitioner embezzled for his purposes, whatever they might be, and therefore the embezzled funds constitute gross * * *income within the meaning of section 61. James v. United States, 366 U.S. 213 (1961).

However, in circumstances where misappropriations do not enrich or benefit the misappropriator, and there is a consensual recognition of an obligation to repay the funds, income does not arise. Beasley v. Commissioner, T.C. Memo. 1989-173; Ocean Sands Holding Corp. v. Commissioner, T.C. Memo. 1980-423; see also Gilbert v. Commissioner, 552 F.2d 478 (2d Cir. 1977), revg. T.C. Memo. 1976-104. * * *

The Supreme Court's decision in James v. United States, supra, firmly established the principle that embezzled funds are income to the embezzler. That decision, however, does not stand for the proposition that all misappropriated funds are gross income of the person who illegally misapplied the funds. The decision necessarily confines taxation of an embezzler to circumstances where the embezzler receives a sufficiently cognizable benefit under the normal principles of income taxation.

The Supreme Court began with the "starting point in all cases dealing with the question of the scope of what is included in 'gross income' * * *[omission in original] that the purpose of Congress was 'to use the full measure of its taxing power'." James v. United States, supra at 218 (quoting Helvering v. Clifford, 309 U.S. 331, 334 (1940)). The Court referred to its oft-quoted language describing the breadth of Congress' intent regarding the statutory definition of gross income: all "'accessions to wealth, clearly realized, and over which the taxpayers have complete dominion'." James v. United States, supra at 219 (quoting Commissioner v. Glenshaw Glass Co., 348 U.S. 426, 431 (1955)). The Court refined this definition by noting that such a gain exists when the "'recipient has such control over it that, as a practical matter, he derives readily realizable economic value from it.'" James v. United States, supra at 219 (quoting Rutkin v. United States, 343 U.S. 130, 137 (1952)). [Hobson v. Commissioner, T.C. Memo. 1992-312.]

This formulation does not require possession of money or property, as petitioner argues, but rather the realization of gain. The inquiry suggested by James v. United States, 366 U.S. 213 (1961), Rutkin v. United States, 343 U.S. 130, 137 (1952), and Hobson v. Commissioner, supra is a commonsense, facts-and-circumstances approach. The approach requires that we answer the following questions: (1) Was there any "economic value" that petitioner could have "readily realized" from what he stole (the realizable value test)? (2) If what petitioner stole had readily realizable economic value, did he have sufficient "control" over it to derive that value therefrom (the control test)?

As to the realizable value test, common sense tells us that a betting ticket has economic value that can be readily realized. Almost all horseplayers purchase betting tickets at face value; if they did not consider the opportunity to gamble at least as valuable as the face amount of the ticket, they would not bet. ⁹

Common sense and experience also tell us that those gamblers who are compulsive, such as petitioner, may misjudge the value of gambling relative to other activities. For some of these individuals, gambling may have a negative "true" value because it can cause misery and shame. See Shaviro, "The Man Who Lost too Much: Zarin v. Commissioner and the Measurement of Taxable Consumption", 45 Tax L. Rev. 215, 238 (1990). Petitioner argues that his compulsion provides a reason to find he had no income. If some people, such as petitioner, are predisposed toward compulsive, self-destructive behavior, that tendency may provide a good reason to regulate or prohibit gambling; it is not a reason to abandon the objective criterion of the market as the mechanism for determining whether the betting tickets that petitioner created and stole had economic value. Cf. United States v. Torniero, 735 F.2d 725 (2d Cir. 1984).

Petitioner might prevail under the realizable value test if his tickets had some unique quality in his hands that drained them of their economic value. However, we do not think they had any such quality. Petitioner stole "opportunities to gamble". See Zarin v. Commissioner , 92 T.C. 1084, 1092 (1989), revd. on other grounds 916 F.2d 110 (3d Cir. 1990). An opportunity to gamble is valuable because it provides a chance to win money. The opportunity to win money is not the only reason a betting ticket is valuable, but it is the overwhelming motivation for most gamblers. The different kinds of excitement associated with gambling are inextricably bound to the possibility of winning money. Gambling is not very interesting without the opportunity to win or lose something of value.

Petitioner appropriated opportunities to win money without paying for them. He said so himself; he punched up tickets on his own behalf because he "wanted some money". He had done so in the past and had occasionally won some money. Petitioner's tickets functioned like regular betting tickets in this regard, and regular betting tickets have readily realizable economic value. We therefore conclude that the betting tickets petitioner stole had value. ¹⁰

The control test asks whether petitioner had sufficient control over the tickets (or, more precisely, the rights evidenced by the tickets) ¹¹ to enable him to realize at least some of their value. ¹² This question we also answer affirmatively. Petitioner's past record of avoiding detection by OTB of his illegal activities convinces us that if he had won one of his earlier bets, he could have covered the amount of his bets with his winnings and taken any excess amount. As explained in more detail in our discussion of the value of the gambling tickets, infra p. 24, even when petitioner placed very large bets in later races, he always had the opportunity to realize winnings that would reduce his liability to OTB. ¹³ As a result, petitioner knew he had an economic stake in the outcome of the races on which he bet. He therefore had the opportunity to derive the same or similar gratifications from gambling as any other bettor.

This case presents the same apparent paradox as Zarin v. Commissioner, supra. The more petitioner lost, the more he gambled, and the more he gambled, the more income he is deemed to have, "thus equating the pleasure of gambling with increase in wealth". Zarin v. Commissioner, supra at 1101 (Tannenwald, J., dissenting). As with our decision in Zarin v. Commissioner, supra, the result is made even more harsh by the restriction of section 165(d), which limits wagering loss deductions to the amount of winnings in the same year. Without that limitation, petitioner would have realized no net income from his peculations.

There is another view of petitioner's activities that resolves the paradox and reconciles the harsh result with the notion of gross income for the purpose of an income tax. The Haig-Simons definition of income states that income during a taxable period is properly defined as the sum of (1) the market value of rights exercised in consumption during the period, and (2) the increase in the value of the store of property rights, or wealth, between the beginning and the end of the period. Haig, The Concept of Income -- Economic and Legal Aspects, in Readings in the Economics of Taxation 54 (Musgrave & Shoup eds. 1959); Simons, Personal Income Taxation 50 (1938). Just as "The Fourteenth Amendment does not enact Mr. Herbert Spencer's Social Statics", Lochner v. New York, 198 U.S. 45, 75 (1905) (Holmes, J., dissenting), the 16th Amendment does not enact the Haig-Simons definition of income; nonetheless, the definition provides a useful, comprehensive formula for measuring gross income under the annual accounting concept.

The Haig-Simons definition does precisely what the above-quoted dissent in Zarin found so paradoxical: it equates "the pleasure of gambling with increase in wealth", but circuitously. Gambling, which may be pleasurable or exciting, is a kind of consumption. Admittedly, it is an odd sort of consumption, because as the tickets are "used up" when the race is run, they have the potential to generate gains that exceed the amount consumed, but this is no more odd than hobby consumption, which is treated the same under section 183 as gambling is treated under section 165(d). The fair market value of rights exercised in consumption counts as much in the income formula as accessions to wealth, so consumption and accessions to wealth are equated for income tax purposes. To say that consumption (or gambling pleasure) and accessions to wealth are treated the same under the Code does not mean that they are the same conceptually. Obviously, there are distinctions between them. The chief distinction is that accessions to wealth increase net worth and are available for future use, while consumption leaves the consumer with only a memory that may fade. The distinctions make no difference. Both consumption and accessions to wealth are parts of income for tax purposes. ¹⁴

Another reason for treating petitioner's bets as income is found in section 61. That section gives meaning to the constitutional phrase "incomes, from whatever source derived", U.S. Const. amend. XVI, by providing a long but nonexhaustive list of examples of income, including compensation and fringe benefits. The teaching of Commissioner v. Glenshaw Glass Co., 348 U.S. 426, 433 (1955), is that Congress, in enacting section 61 and its predecessors, intended to reach everything taxable under the 16th Amendment. Employees who receive noncash benefits from their employers must pay tax on those benefits, as if they had received the cash and purchased the benefits for themselves. Sec. 61(a)(1). ¹⁵

It would be anomalous to tax those who lawfully receive fringe benefits from their employers while allowing those who unlawfully appropriate such benefits to escape taxation. James v. United States, 366 U.S. 213, 220 (1961). If petitioner's employer had given him the tickets as a reward for a job well done, the tickets would have been income.¹⁶ To allow petitioner to avoid inclusion in gross income of the value of the bets he unlawfully placed on his own behalf would be contrary to sound tax policy, creating tax-free benefits for employees who steal from their employers.

How Much Income Did Petitioner Have?

We find that petitioner had $ 80,280 in gross income.

Respondent argues that theft income is measured by the fair market value of the thing stolen. Fair market value is defined as "the price at which the property would change hands between a willing buyer and a willing seller, neither being under any compulsion to buy or sell and both having reasonable knowledge of relevant facts". Sec. 1.170A-1(c)(2), Income Tax Regs.; sec. 20.2031-1(b), Estate Tax Regs; see also sec. 1.1014-3(a), Income Tax Regs.¹⁷ Although we have never expressly announced that the fair market value standard should be used to determine income from theft of noncash items, we employed that test in Bonansinga v. Commissioner, T.C. Memo. 1987-586.

Petitioner argues that fair market value should not apply because many gamblers are compulsive and because some races are fixed. This is like arguing that a $ 3.50 gin and tonic is not worth $ 3.50 because some people are alcoholics and besides, the bartender might water the drinks. Presumably people in both marketplaces act with knowledge of the potential for product defects and act accordingly. The fact that some people are willing to pay a high price for something they are addicted to is adequately explained by market behavior. In a similar vein, petitioner argues that "a $ 10 ticket that pays off at $ 7.80 is not worth $ 10." This argument is specious. The minimum payoff on a $ 2 bet under the pari-mutuel system is arbitrarily set at $ 2.10, even if everyone bets on the winner and the mathematical computation of the payoff is $ 1.66 (after the 17-percent "takeout"). The track must make up the 44-cent difference. Ainslie, Ainslie's Encyclopedia of Thoroughbred Handicapping 190 (1978). A $ 10 bet would pay at least $ 10.50. Petitioner's arguments are insufficiently compelling to cause us to depart from the fair market value standard.

We agree with respondent that fair market value is the appropriate standard. Fair market value is the measure of nonmonetary income in every other context, and we see no reason to depart from that useful and well-developed standard of income measurement. However, our choice of standard is not dispositive; establishing the standard merely leads to the next question: What was the fair market value of the rights to gamble that petitioner created and stole?

We first consider petitioner's arguments that he should be charged with no income because the betting tickets were worthless. First, petitioner asserts that his tickets were worthless because he would not have been able to collect on his winnings. We do not believe that this is an accurate statement of the facts. It would have been difficult to collect on petitioner's bets, but not impossible. Petitioner's statements and actions show he believed he would have been able to collect if he had won. Petitioner admitted that several times he had bet at work without paying, and that sometimes he ended up ahead. Petitioner's admission to the police that on occasion he gambled because he "wanted some money" shows the weakness in this claim. When petitioner started making his bets, he believed he would be able to exchange any winning tickets for cash.

Not only did petitioner subjectively believe he would collect his winnings, but a reasonable person also would have expected to be able to collect on a winning bet, and could have realized winnings without detection in a number of ways. If the amount realized was small, petitioner could have cashed the tickets himself at his work station, as he had done in the past. If the winnings were larger, he could have sold the winning tickets to a friend, perhaps at a discount, or he could have conspired to share the proceeds with an outsider. Since there is nothing on a winning ticket to identify the bettor, the accomplice could have cashed the tickets without complication. Petitioner also could have covered the cost of the tickets and redeemed them himself at a different OTB parlor, where he would not be recognized as an employee.

Petitioner next makes a series of timing arguments. Petitioner argues that at the end of the racing day, indeed by the end of the race, a losing ticket is worthless, and petitioner therefore should not be charged with having received income. If petitioner's actions had occurred after the race was run, we would agree. They did not. As he punched up his tickets, petitioner acquired valuable rights -- to receive a prerace refund on demand, to participate in the horserace as a bettor, and to receive money if his horse finished in the top three.

Petitioner's argument is reminiscent of the discredited "diminution of loss" approach of Bowers v. Kerbaugh-Empire Co., 271 U.S. 170 (1926). The Supreme Court held in that case that when a transaction as a whole results in a loss, an interim gain should not be taxed. Kerbaugh-Empire has "lost its vitality". Zarin v. Commissioner, 92 T.C. 1084, 1091 (1989), revd. on another issue 916 F.2d 110 (3d Cir. 1990). We are satisfied that our Zarin opinion was correct in this regard.

Petitioner urges that our interpretation would lead to incongruous results. A person who finds an Irish Sweepstakes ticket, which turns out to be a loser later that day, first would be charged with income from treasure trove and then would suffer a nondeductible wagering loss, a result petitioner finds fantastic. We are not amazed. Petitioner has described the correct tax treatment of treasure trove. See sec. 165(d); Cesarini v. United States, 296 F. Supp. 3 (N.D. Ohio 1969), affd. 428 F.2d 812 (6th Cir. 1970); sec. 1.61-14(a), Income Tax Regs.; see also Rev. Rul. 61, 1953-1 C.B. 17; Rosenak, "Taxation of Found Property and Other Windfalls", 20 U. Chi. L. Rev. 748 (1952).

Petitioner further argues that we should disregard units of time shorter than a day in reaching our result. Inasmuch as petitioner's losing tickets were worthless at the end of the day, petitioner claims he should have no income from them. Petitioner cites authorities going back to Blackstone and Bouvier for the proposition that two events occurring on the same day are generally considered to occur at the same time for legal purposes. Blackstone, Commentaries on the Law of England 902 (Jones ed. 1915); 1 Bouvier's Law Dictionary 757 (Rawle's rev. 1914).

Petitioner has greatly overstated the application of this principle. When two judgments or liens are recorded on the same day, they are considered to be recorded at the same time,

but this is merely a legal fiction, which does not apply where it is necessary to distinguish between the two parts of a day; and, therefore, it has been said that there is no such general rule of law, but that common sense and common justice sustain the propriety of considering fractions of a day whenever it will promote the purposes of substantial justice; [2 Bouvier's Law Dictionary 1298; citations omitted.]

Petitioner asks us to stretch the legal fiction to cover this case, but to do so would disregard both the form and the substance of what he did, which are more appropriately described by the actual sequence of events. Common sense requires that we not view all the events of July 17, 1988, as occurring at the same time. The order of events is of fundamental importance. Before the race, no one can be certain what the outcome will be, and the price of the tickets reflects this uncertainty. After the race, a betting ticket is transformed into either a winning ticket or a worthless scrap of paper. Whatever the results of the race, the ticket has a different value after the race than before. The running of the race, not the stroke of midnight, is the defining event in the life of a betting ticket. We therefore disagree with petitioner that what he stole had no value.

Petitioner also argues that the value of the tickets, if not zero, was less than their face amount, because petitioner only had access to the money in the till. To this argument we now turn.

Petitioner stole "opportunities to gamble". See Zarin v. Commissioner, supra at 1092. As we said supra p. 19, an opportunity to gamble is valuable because it provides a chance to win money. As long as petitioner's opportunities to win money were not materially different from those of other gamblers, his opportunities to gamble were worth as much as the wagers of legitimate bettors.

In the usual case, an OTB horseplayer's potential payoff is limited only by the size of the mutuel pool and the choices of the other bettors. See supra p. 11. In this case, petitioner argues that his opportunities to win were much more severely limited. Petitioner claims that the most he could reasonably expect to walk away with in winnings was the amount of cash in his drawer, which was no more than $ 4,691 at any time during his shift, and was only $ 843.60 at the end of the racing day. Any winnings greater than the amount of money in the drawer would have required petitioner's supervisor to replenish the cash in the drawer. The moment the supervisor did so, petitioner's scheme would have been exposed, and he would have won nothing. Other OTB bettors did not face this artificial restriction on their winnings.

The problem with this argument is that it is largely a consequence of the way petitioner placed his bets and intended to receive his winnings. As explained in Appendix B, infra p. 50, if petitioner had taken his winnings in the form of new gambling tickets, or if he had split each bet among several tickets, he might have obtained full value, at least early in the racing day when his bets were still relatively small.

We realize that petitioner may not have planned his activities with the cunning we describe in Appendix B. However, the fact remains that petitioner completely controlled the structure of his bets. These were not arm's-length transactions. The fact that he chose to place the bets in a way that made realization of his winnings more difficult is of no tax consequence. To say that his bets were worth less to him because he did not appreciate his circumstances is to say that his psychic income differed from that of other gamblers or embezzlers. See, e.g., Shaviro, "The Man Who Lost too Much: Zarin v. Commissioner and the Measurement of Taxable Consumption", 45 Tax L. Rev. 215 (1990); Dodge, "Zarin v. Commissioner: Musings About Debt Cancellations and 'Consumption' in an Income Tax Base", 45 Tax L. Rev. 677 (1990). It is not for us to open the Pandora's box of psychic income unless and until Congress requires us to do so.

What we do find significant, however, is that the likelihood that petitioner would be caught and be forced to forfeit all his net winnings, whether in the form of cash or betting tickets, increased dramatically with his later bets. This was an economic difference between petitioner's position and that of a legitimate bettor over which he had no control. As long as petitioner's bets were so small that they were obscured by the usual traffic, he ran a minimal risk of being caught. As his bets became larger, swelling to a total of 20 times legitimate sales at his terminal, detection became much more likely, and then inevitable, regardless of whether he won or lost. Even if petitioner had not turned himself in at the end of the day, the day-end activity summary printed by his computer terminal would have shown an extraordinary amount of betting activity. Undoubtedly, OTB would have investigated further.

Detection would not have resulted in total forfeiture of petitioner's ill-gotten gains. Until the day's racing card was completed, any winning bets that petitioner had placed would have reduced his debt to OTB. Even if petitioner's excess winnings had been forfeited, the winnings that petitioner used to reduce his debt to OTB would not have been forfeited, but rather applied against the debt. That is what happened. Because the large volume of petitioner's illegal bets in the last five races virtually assured that he would forfeit any excess winnings, the best petitioner could do with his bets on races 6-10 was to wipe out his debt to OTB.

The effective ceiling on the amount petitioner could win by placing bets on races 6-10 was the amount of his debt to OTB. If, for instance, a legitimate bettor had bet $ 15,000 to win in the eighth race on a horse paying 2:1 odds, this bettor would have had a potential return of the $ 15,000 bet plus $ 30,000, for a total of $ 45,000. Petitioner's debt after placing his $ 15,000 wager ¹⁸ was $ 26,280. Even if petitioner's ticket had entitled him to $ 45,000 if his horse had won, he would have had a reasonable prospect of "collecting" only $ 26,280 by reducing his debt to zero. He could not have received the rest in cash or kind. Considering that petitioner's potential payoff under this assumption would have been only 58.4 percent of a legitimate bettor's, his tickets on this bet would appear to have been worth only 58.4 percent of face value. On the other hand, a bet on a horse at 2:3 odds would have returned $ 25,000 total. All of the winnings could have been used by petitioner to reduce his debt to $ 1,280. Therefore, such a bet would have been worth as much as a bet placed by a legitimate bettor.

We have calculated the potential payoff to petitioner on each of his bets; the logical and mathematical details of our calculations are contained in Appendix C, infra page 54. Suffice it to note here that we have concluded that petitioner bet on the favorite in each race, and therefore his potential payoffs were relatively low. ¹⁹ Our calculations enable us to compare petitioner's potential payoff with the amount of his debt immediately after placing his bets on each race:

Race	Potential payoff	Petitioner's debt
6	$ 2,835	$ 3,780
7	10,825	11,280
8	20,450	26,280
9	33,830	51,780
10	34,200	71,355

Petitioner's potential payoffs on races 6-10 never exceeded the amount of his accumulated debt to OTB. In other words, even if the horses picked by petitioner had won any of these races, he would not have eliminated his debt to OTB. Petitioner did not have a realistic chance of winnings over and above the amount of his debt, so he would not have had the opportunity to take any money out of the till, let alone all of it. The odds, not the money in the till, limited what petitioner could have won, and in that regard, petitioner was just like any other OTB bettor.

In an economic sense, petitioner's tickets were identical to any legitimate betting ticket. Therefore, the value of petitioner's tickets must have been the same as what other bettors paid -- their face amount.

Was Petitioner's Theft Income Also Gambling Income?

We must now consider the character of petitioner's $ 80,280 in income. Petitioner also had substantial losses, reflected by his liabilities to OTB and its insurer. If we were to characterize the $ 80,280 in income as gains from wagering transactions, and to characterize his losses as also from wagering transactions, then petitioner would be able to deduct his losses from his income, resulting in zero taxable income. Sec. 165(d). ²⁰ Otherwise, petitioner has $ 38,105 in net theft income, which cannot be offset by his losses.

We conclude that petitioner's gross income is not properly characterized as income from wagering transactions. To protect the fisc from unwarranted writeoffs of gambling losses, we and other courts have narrowly interpreted the income prong of section 165(d). See Boyd v. United States, 762 F.2d 1369, 1373 (9th Cir. 1985) (floor manager's share of table fees charged to poker players was not wagering income); Williams v. Commissioner, T.C. Memo. 1980-494 (blackjack dealers' tokes were not wagering income); cf. Nitzberg v. Commissioner, 580 F.2d 357 (9th Cir. 1978), revg. T.C. Memo. 1975-228 (loss prong interpreted broadly; poker club's losses from backing shills who played with paying customers were not ordinary and necessary business expenses, but gambling losses, deductible only to the extent of winnings). In common speech, a gain from a wagering transaction means the amount won over the amount bet; it implies that there was a prior wager. Theft of the opportunity to gamble does not result in a gain from wagering because there was no prior wager. Cf. Wilson v. Commissioner, T.C. Memo. 1989-652 (theft of cash subsequently lost by gambling characterized as embezzlement income). Petitioner's gains were from wagering only in the sense that he gambled that he would not be caught.

Petitioner's income was ordinary income from theft. Therefore, petitioner may not deduct his $ 38,105 of gambling losses from the $ 80,280 in gross income generated by his theft of the betting tickets.

Was Petitioner's Loss Deductible as a Loss From an Income-seeking Activity?

Petitioner argues that his scheme was a unitary, though multistep, form of embezzlement in which placing his bets was only the first step. Petitioner planned ultimately to use his winnings to cover the amounts embezzled. The price of his bets therefore can be viewed as a preliminary cost. When the first phase of his embezzlement failed, his scheme fell apart, resulting in a loss equal to the amount he owed to his employer.

We understand petitioner to argue that he should be able to deduct the amount of his loss as a loss from a transaction entered into for profit. Section 165(a) permits deductions for uncompensated losses. For individuals, the deduction is limited to trade or business losses, losses from profit-seeking activities, and casualty losses. Sec. 165(c). Respondent has ruled that an embezzler is entitled to a deduction in the year he makes restitution, on the theory that the restitution payment is a loss from a profit-seeking activity. Rev. Rul. 65-254, 1965-2 C.B. 50 (citing sec. 165(c)(2)).

This argument also fails. While we agree that the loss was a loss from embezzlement, it also was a loss from wagering. Unlike the $ 80,280 in gross income, the loss followed and was caused by petitioner's wagers. Cf. Wilson v. Commissioner, supra (loss of embezzled funds by gambling characterized as wagering loss). Section 165(d) applies specifically to wagering transactions. Specific language in a statute controls more general language. If we were writing on a clean slate, and section 165(d) had never been enacted, we might have been inclined to view petitioner's embezzlement gains and gambling losses as offsetting items. However, the specific limitation of deductions from wagering losses to the amount of wagering gains, sec. 165(d), controls the outcome here. See Nitzberg v. Commissioner, supra at 358 (specific language of sec. 165(d) trumps sec. 162). Petitioner did not have additional wagering gains and thus is not entitled to a deduction for the amount of his "embezzlement" loss over and above the restitution he made by turning over his winning tickets to OTB.

Did Petitioner Nonetheless Realize no Income?

Petitioner argues that the outcome of his case is governed by Gilbert v. Commissioner, 552 F.2d 478 (2d Cir. 1977), revg. T.C. Memo. 1976-104, in which the Second Circuit held that a convicted embezzler did not realize income from his embezzlement. Barring stipulation to the contrary, this case is appealable to the Second Circuit. To the extent that Gilbert v. Commissioner, supra, is "squarely on point", as petitioner argues, we are bound to follow it. Golsen v. Commissioner, 54 T.C. 742, 757 (1970), affd. 445 F.2d 985 (10th Cir. 1971).

Petitioner claims that, like Gilbert, he fully recognized in the year of the theft his obligation to repay his victim, and that therefore he did not realize income from his activities. Respondent concedes only that the $ 42,175 in winning tickets that petitioner left in his drawer at OTB constitutes restitution for which he is entitled to a deduction. Sec. 165(c)(2); Fox v. Commissioner, 61 T.C. 704 (1974); see Rev. Rul. 65-254, 1965-2 C.B. 50. For the reasons discussed below, we disagree that Gilbert controls and hold for respondent on this point.

The facts of Gilbert v. Commissioner, supra, are distinguishable. Gilbert was president and principal shareholder of E.L. Bruce Co., which Gilbert wanted to merge with Celotex Corp.Bruce's board had authorized moving forward with the plan. Gilbert personally bought a 56-percent interest in Celotex on margin and commenced merger negotiations. The stock market declined before the merger could be consummated and Gilbert's stockbroker sent him a margin call. Gilbert directed the corporate secretary of Bruce to satisfy the margin call out of corporate funds, and promptly notified several Bruce officers and directors of his actions and his intention to repay the funds. Ten days after the margin call Gilbert executed interest-bearing promissory notes in favor of Bruce for the full amount, secured by assignment of all his real and personal property. The value of the assigned property was more than sufficient to cover the amount of the notes. Gilbert was convicted of embezzlement. We held that Gilbert had income in the amount taken from Bruce. Gilbert v. Commissioner, T.C. Memo. 1976-104.

The Second Circuit reversed, holding that Gilbert did not realize income:

We conclude that where a taxpayer withdraws funds from a corporation which he fully intends to repay and which he expects with reasonable certainty he will be able to repay, where he believes that his withdrawals will be approved by the corporation, and where he makes a prompt assignment of assets sufficient to secure the amount owed, he does not realize income on the withdrawals under the James test. * * * [Gilbert v. Commissioner, 552 F.2d at 481.]

Similarly to Gilbert, petitioner took the equivalent of money from OTB and intended to repay the face value of the tickets out of his winnings. However, several of the factors mentioned in the above-quoted paragraph are absent from this case. Petitioner could not have been reasonably certain that he would be able to repay the tens of thousands of dollars in tickets he took because he had earnings of less than $ 12,000 per year. He could not have believed that his appropriation would be approved by OTB. He did not make a prompt assignment of assets, and it appears unlikely that he owned sufficient assets to secure the $ 38,105 he owed. His debt was reduced to judgment in 1989, the following taxable year.

Petitioner nonetheless urges that we distill a broader proposition of law from Gilbert. If Gilbert stands for such a proposition -- in view of the many qualifications in the above-quoted paragraph, we doubt it -- it is that an agent does not realize income when he acts in the best interests of the principal by using the principal's funds to cover a liability incurred in the agent's name on the principal's behalf. This proposition does not apply to this case. Petitioner stipulated that he was acting on his own behalf, not OTB's. A narrower reading of Gilbert is that if, during the year of the embezzlement, an embezzler assigns assets to his victim that are more than sufficient to cover the amount embezzled, the embezzler does not realize income. Petitioner made no such assignment.

Petitioner also argues that, like Gilbert, he "consensually recognized" his obligation to repay OTB when he turned himself in. The quoted phrase can be traced to James v. United States, 366 U.S. 213, 219 (1961). Most courts that have opined on the matter have concluded or implied that the Supreme Court meant that if the embezzler makes a bona fide arrangement with the victim in the same taxable year as the theft to repay the embezzled funds, the embezzler is either entitled to a deduction at the time the arrangement is entered into,or the arrangement converts the embezzlement into a loan. See, e.g., Buff v. Commissioner, 496 F.2d 847 (2d Cir. 1974), revg. 58 T.C. 224 (1972); Moore v. United States, 412 F.2d 974 (5th Cir. 1969); Mais v. Commissioner, 51 T.C. 494 (1968). However, we have been able to locate only two reported decisions since James -- our decision in Buff, which the Second Circuit rebuffed, and the Second Circuit decision in Gilbert -- that have allowed an embezzler relief from taxation without actual repayment.

The linchpin of the "consensual recognition" doctrine, which is closely related to the claim of right doctrine, is that the victim must agree in the same taxable year as the theft to treat the transaction as a loan, rather than as an unauthorized taking. Buff v. Commissioner, 58 T.C. at 227-228, 232; Bittker & Lokken, 1 Federal Taxation of Income, Estates and Gifts, par. 6.5 n.19 and accompanying text (1989). The Second Circuit's reversal of Buff appears to add the requirement that the agreement have a reasonable chance of resulting in full repayment. Buff v. Commissioner, 496 F.2d at 847 (embezzler took $ 22,000 and agreed to pay $ 1,000 plus $ 25 per week). ²¹ At best, it was petitioner who acquiesced in OTB's right to relief in 1988, perhaps because he believed OTB was entitled under state regulations to recover the loss from him. This is equivalent to any other detected embezzler's general acknowledgment of his duty to repay, which is insufficient under both our analysis and the Second Circuit's in Buff v. Commissioner, supra. Neither OTB nor its insurance company "consensually recognized" anything during 1988, and the only evidence of petitioner's ability to repay, his 1988 tax return, shows that it was very unlikely that petitioner would be able to repay.

We hold that petitioner is entitled only to a deduction of $ 42,175, as permitted by respondent, resulting in taxable income in addition to that reported on his return of $ 38,105. Petitioner may be allowed a deduction for the amounts he actually repays when he repays them, but not before.

Additions to Tax

Section 6653(a)(1) imposes an addition to tax if any portion of an underpayment was due to negligence or intentional disregard of rules or regulations. Negligence is defined as the "lack of due care or failure to do what a reasonable and ordinarily prudent person would do under the circumstances." Neely v. Commissioner, 85 T.C. 934, 947 (1985). Although petitioner has the burden of proof on this issue, Rule 142(a); Bixby v. Commissioner, 58 T.C. 757, 791-792 (1972), when, as here, the issues were not cut and dried, the addition to tax for negligence has been ruled inappropriate. Govier v. Commissioner, T.C. Memo. 1990-611; Denbow v. Commissioner, T.C. Memo. 1989-92; see also Marcello v. Commissioner, 43 T.C. 168, 182 (1964), affd. in part, remanded in part 380 F.2d 499 (5th Cir. 1967).

We believe that the "consensual recognition" gloss on James, and the fact that Gilbert is precedent that arguably appeared to control this case, Golsen v. Commissioner, 54 T.C. 742, 757 (1970), affd. 445 F.2d 985 (10th Cir. 1971), provided petitioner with tenable reasons for treating his embezzlement and gambling as non-events for tax purposes. Petitioner reasonably perceived his circumstance as a loss. As he stated in his petition, which he prepared and filed pro se, "I disagree because I never actually received any money, my loses [sic] that day were far greater than all for the year 1988." Petitioner did not act unreasonably when he failed to report the income. Petitioner is not liable for the addition to tax for negligence.

Section 6661provides for an addition to tax when the tax shown on a return is substantially less than the taxpayer's true tax liability. An understatement is substantial if it exceeds the greater of $ 5,000 or 10 percent of the tax required to be shown on the return. Sec. 6661(b)(1)(A). Contrary to respondent's argument, the section 6661 addition to tax is not entirely "computational". When computing the addition to tax, the understatement is reduced by the amount of the understatement attributable to items for which the taxpayer had substantial authority. Sec. 6661(b)(2)(B)(i). Substantial authority is a less demanding standard than "more likely than not", but stricter than the "reasonable basis" required to avoid the addition to tax for negligence. Antonides v. Commissioner, 91 T.C. 686, 702 (1988); sec. 1.6661-3(a)(2), Income Tax Regs. Under the regulations, the time for testing whether the taxpayer had substantial authority is at the time the return was filed or at the end of the taxable year, sec. 1.6661(b)(4)(iii), Income Tax Regs., ²² so an authority may be substantial even if it is ultimately held inapplicable. Harston v. Commissioner, T.C. Memo. 1990-538, affd. without published opinion 936 F.2d 570 (5th Cir. 1991).

As we stated in our discussion of negligence, petitioner had tenable reasons for not including and not disclosing the results of his activities on his return. The arguments underlying these reasons had some substantial likelihood of success, although they have not succeeded. Thus, petitioner had substantial authority. The amount of his understatement attributable to items for which he did not have substantial authority is zero. Petitioner is not liable for the addition to tax under section 6661.

To reflect the foregoing,

APPENDIX A

Race Results, Finger Lakes Race Track, July 17, 1988

First Race Mutuel Pool: $ 16,711

Collins' Bets: $ 60

Horse	Odds $ 1
Mac's Boy	$ 10.80
Get Me Luce	3.30
Joaleo	3.10
Holy Shamoly	9.60
Pappa's Song	2.40
Carols Bob	12.40
Copper Sweep	3.90

$ 2 Mutuel Prices	Win	Place	Show
Mac's Boy	$ 23.60	$ 10.20	$ 5.60
Get Me Luce		6.60	3.40
Joaleo			3.60

Second Race Mutuel Pool: $ 17,017

Collins' Bets: $ 120

Horse	Odds $ 1
Fancy High	$ 7.00
Ivory Palace	6.90
Feeling His Oats	4.10
Antibiotic	3.90
Lucky Leads	21.10
Overhaul	14.10
Little Indian	a-1.10
Classic Mistrial	a-1.10

Coupled: Little Indian and Classic Mistrial [coupled horses function as a single betting unit]

$ 2 Mutuel Prices	Win	Place	Show
Fancy High	$ 16.00	$ 7.20	$ 5.20
Ivory Palace		7.20	5.00
Feeling His Oats			4.00

Third Race Mutuel Pool: $ 17,625

Collins' Bets: $ 600

Horse	Odds $ 1
Patch Up's Puddles	$ 7.70
Passline Honey	3.70
Ritzy Laten	9.60
Iron Tiara	2.10
Augusta Faberge	8.00
Lovers Mirage	9.10
Love Cream	2.50

$ 2 Mutuel Prices	Win	Place	Show
Patch Up's Puddles	$ 17.40	$ 7.40	$ 4.60
Passline Honey		6.60	4.40
Ritzy Laten			4.40

Fourth Race Mutuel Pool: $ 22,730

Collins' Bets: $ 1,500

Horse	Odds $ 1
Mistique Mover	$ 3.20
Valdante	3.70
Heaven's Seven	5.30
Twice Gold	4.60
Portentious	1.60
Sir Ramborne	14.40

$ 2 Mutuel Prices	Win	Place	Show
Mistique Mover	$ 8.40	$ 4.40	$ 4.40
Valdante		4.00	4.20
Heaven's Seven			4.70

Fifth Race Mutuel Pool: $ 21,732

Collins' Bets: $ 0

Horse	Odds $ 1
Poker's Judge	$ 3.20
Emerald Lil	5.80
Top Slice	7.20
Noble Viki	1.10
Tacita de Plata	4.40
Batchelorette	12.70

$ 2 Mutuel Prices	Win	Place	Show
Poker's Judge	$ 8.40	$ 3.80	$ 2.80
Emerald Lil		6.20	3.80
Top Slice			3.40

Sixth Race Mutuel Pool: $ 19,562

Collins' Bets: $ 1,500

Horse	Odds $ 1
Jud Star	$ 6.30
Star of Willa Wah	6.10
Yes You Are	5.30
License to Love	6.50
Original Work	6.70
Doc Friday	14.70
Chesapeake Mate	22.50
Wells Express	20.30
Censored	27.50
Mr. Tee and Eee	1.90

$ 2 Mutuel Prices	Win	Place	Show
Jud Star	$ 14.60	$ 8.20	$ 5.80
Star of Willa Wah		9.60	5.60
Yes You Are			4.80

Seventh Race Mutuel Pool: $ 26,164

Collins' Bets: $ 7,500

Horse	Odds $ 1
Orphan Jones	4.10
Tanked	11.70
Armored Knight	4.50
Bomb The Tote	1.00
Rythm King	4.60
Rag Time Boy	8.60

$ 2 Mutuel Prices	Win	Place	Show
Orphan Jones	$ 10.20	$ 6.80	$ 5.80
Tanked		9.80	8.20
Armored Knight			7.20

Eighth Race Mutuel Pool: $ 35,634

Collins' Bets: $ 15,000

Horse	Odds $ 1
Royal Trial	$ 10.00
Roll A Bid	5.50
The Dazzler	4.40
Irresistible Bear	.90
Charlton's Hope	8.00
Mark Neil	7.80
Lord Robdavbri	18.80

$ 2 Mutuel Prices	Win	Place	Show
Royal Trial	$ 22.00	$ 15.00	$ 11.80
Roll A Bid		10.40	9.40
The Dazzler			11.00

Ninth Race Mutuel Pool: $ 52,682

Collins' Bets: $ 25,500

Horse	Odds $ 1
Calabas	7.60
Classic Pleine	2.90
Mr. Oneill	.80
Gold Stark	17.90
Color Of The Law	2.80

$ 2 Mutuel Prices	Win	Place	Show
Calabas	$ 17.20	$ 13.40	$ 2.10
Classic Pleine		$ 7.40	$ 2.10
Mr. Oneill			$ 2.10

10th Race Mutuel Pool: $ 49,449

Collins' Bets: $ 28,500

Horse	Odds $ 1
Bleu Macadamia	$ 0.50
Ridge Rock Rosie	19.60
Looney Binn	11.90
You Fly I Buy	20.00
Continental Katie	7.70
Full Figure	25.20
BooFul	18.10
Exactly Packly	57.50
Miss Deborah Jean	10.60
One Trial	29.30
Turning Leaf	30.00
My Valiant Lady	53.70

$ 2 Mutuel Prices	Win	Place	Show
Bleu Macadamia	$ 3.00	$ 2.10	$ 2.10
Ridge Rock Rosie		6.20	2.10
Looney Binn			2.10

APPENDIX B

Methods of Cashing Large Winning Tickets Without Detection

Petitioner generally placed all three bets for each race on a single ticket. In order to cash in a winning ticket, he would have had to collect the full amount all at once or forgo a portion of his winnings. For example, take the fourth race, when petitioner bet $ 1,500 split among win, place, and show bets. Suppose that he had bet on Mistique Mover, which won the race and paid $ 8.40 to win, $ 4.40 to place, and $ 4.40 to show. Petitioner's ticket would have been worth $ 4,300, but he would have known that to avoid calling the manager to replenish his drawer, he could take only what he expected to have in the drawer in cash at the end of the day. In this case the actual closing balance was $ 843.60. ²³ After the fourth race, he had placed $ 2,280 in bets, which he had to put back in the drawer by the end of his shift, either in cash or winning or canceled tickets. Petitioner could have terminated his venture and taken home net winnings of $ 2,020, if he had the cash to do it. Petitioner presents his options as: (1) put the ticket in the drawer, remove $ 843.60 in cash, and forgo the other $ 1,176.40 in net winnings; or (2) keep the ticket, end up with a drawer $ 2,280 short, and face the music.

It seems to us that there were at least two other possibilities (at least in the first four races). Petitioner could have taken the $ 843.60 in cash and punched up another $ 1,176.40 in bets, taking some winnings in kind rather than in cash. His terminal would have indicated cash flow as follows:

Starting cash	$ 805.00
Legitimate sales	3,857.00
Telephone deposits	29.00
Collins' bets	1 3,456.40
Total inflows		$ 8,147.40
Cancellations	(40.00)
Legitimate payouts	(3,776.60)
Payout to Collins	(4,300.00)
Total outflows		(8,116.60)
Correct closing balance		30.80
Actual drawer balance		2 0.00
Difference		$ 30.80

Compare the cash flow if petitioner had never gambled, supra p. 14. Although the drawer is out of balance, it is in the position it would have been had petitioner not gambled at all, so he could have avoided detection of his gambling, regardless of whether his new bets paid off or not. Petitioner easily could have left OTB at the end of his shift with his new betting slips in his pocket (provided OTB was not otherwise tipped off). If the bets paid off, he could have cashed them in at another OTB parlor, or convinced a friend to do it.

Petitioner could have placed his bets on several tickets by splitting the amounts. He admitted that he sometimes punched up several tickets at a time, although it is not clear that he did so in a manner that would guarantee payment. See supra n.2. Instead of betting $ 1,500 on a single ticket in the fourth race, he could have placed 10 bets of $ 150. ²⁴ Together, the 10 tickets would have paid the same amount as a single ticket -- $ 4,300. But each ticket would have paid $ 430. Petitioner could have cashed in six tickets worth $ 2,580, balancing out the $ 2,280 in unpaid wagers and taking $ 300 in cash, while retaining the other four tickets, now worth $ 1,720, to cash in at another time and in another place, or by somebody else. The ticket could have been cashed in at any time until March 31, 1989. Petitioner's cash flows for the day, as reported by his terminal, would have been:

Starting cash	$ 805.00
Legitimate sales	3,857.00
Telephone deposits	29.00
Collins' bets	2,280.00
Total inflows		$ 6,971.00
Cancellations	(40.00)
Legitimate payouts	(3,776.60)
Payout to Collins	(2,580.00)
Total outflows		(6,396.60)
Correct closing balance		574.40
Actual drawer balance		1 543.60
Difference		$ 30.80

This last method would have had the advantage of not creating unduly large cash inflows and outflows. Large amounts, even though not affecting the bottom line, would have tipped off OTB to a problem and increased the likelihood of detection.

APPENDIX C

Method of Calculating Potential Payoffs

Our valuation theory is that if petitioner's potential payoff (i.e., his debt to OTB) was less than an ordinary gambler's potential payoff, then petitioner's tickets were worth proportionately less than ordinary gambling tickets. This theory requires that we determine the horses that petitioner bet on during the last five races. We then can determine from the race results in Appendix A what the potential payoffs were on those horses. Then we can compare the potential payoff to the level of petitioner's debt, and discount the value of the tickets, if appropriate.

We must determine which horses petitioner bet on. For races 9 and 10, petitioner's bets paid off. In the ninth race, petitioner won $ 8,925. This corresponds to a show bet on any of the first three horses in that race, each of which paid $ 2.10 on a $ 2 bet. Inasmuch as petitioner did not win more than $ 8,925, and petitioner placed all his bets on the same horse, petitioner could not have bet on the horses that won and placed. If he had, he would have won $ 138,975 or $ 31,450, respectively. Therefore, petitioner must have bet on the third-place horse, Mr. Oneill. We will use the actual win and show payoffs on Mr. Oneill in our calculations.

In the 10th race, petitioner won $ 33,250. No horse corresponds to petitioner's usual pattern of betting an equal amount on a single horse to win, place, or show. If he had bet on the favorite, Bleu Macadamia, he would have won $ 34,200, or $ 950 more than the record indicates. One possible explanation is that when petitioner turned in his winning betting tickets, the total payoff was incorrectly calculated. Another explanation is that petitioner deviated from his usual betting pattern, but this does not seem likely. Whatever the reason, the difference is too small to be significant in our calculations. We conclude that petitioner bet on Bleu Macadamia, and we will use that horse's actual win, place, and show payoffs in our calculations. ²⁵

The record does not provide direct evidence of the horses petitioner bet on in races 6-8. However, the facts (a) that in each race he bet on a single horse to win, place, or show, and (b) that his bets did not pay off at all in those races, tell us that petitioner did not bet on the horses that finished first, second, or third in those races. The favorite in each of races 1-8 did not finish among the top three. We have seen that petitioner bet on the favorite in races 9 and 10, and that petitioner had a habit of betting in a simple, consistent pattern. A common, although not invariably successful, betting strategy is to bet the favorite in each race, doubling the bet after each loss. See supra n.19. Petitioner's bets were very large in comparison to the size of the pool. See supra n.7. As a result, we would expect the horses petitioner bet on to have low odds. See supra p. 11. These facts lead us to believe that petitioner bet on the favorites in races 6-8, as well as the earlier races.

On the basis of the foregoing analysis, we conclude that petitioner bet on the favorites in races 6-10, as follows:

		Win bet
Race	Horse	1 odds $ 1
6	Mr. Tee and Eee	1.90
7	Bomb The Tote	1.00
8	Irresistible Bear	0.90
9	Mr. Oneill	0.80
10	Bleu Macadamia	0.50

The odds shown are for "win" bets. They thus provide no direct help in valuing petitioner's place and show bets, since the pools for the various kinds of bets are separate and independent. Place and show odds are not published. However, by examining the actual win, place, and show payoffs on the horses that won the races that petitioner bet on, we can extrapolate the likely amount of petitioner's potential place and show payoffs in a manner sufficiently reliable for the purpose of this inquiry.

Using a rudimentary statistical technique known as regression analysis, we attempted to correlate the odds on each winning horse with the "odds" on the place and show bets on those horses. Inasmuch as the odds on place and show bets are not published, we worked backwards from the payoff to determine the odds. For instance, in the first race Mac's Boy won, paying $ 23.60 to win, $ 10.20 to place, and $ 5.60 to show. By subtracting $ 2 from each payoff and dividing by two, we determined that the win odds were $ 10.80 (as shown in Appendix A), the place odds were $ 4.10, and the show odds were $ 1.80. The odds on the winning horses in the races petitioner bet on are as follows:

Race	Horse	Win	Place	Show
1	Mac's Boy	$ 10.80	$ 4.10	$ 1.80
2	Fancy High	7.00	2.60	1.60
3	Patch Up's Puddles	7.70	2.70	1.30
4	Mistique Mover	3.20	1.20	1.20
6	Jud Star	6.30	3.10	1.90
7	Orphan Jones	4.10	2.40	1.90
8	Royal Trial	10.00	6.50	4.90
9	Calabas	7.60	6.70	.05
10	Bleu Macadamia	.50	.05	.05

After gathering the odds information, we computed the curve (expressed as a mathematical formula) that best describes the relationship between the win odds and the place odds for the winning horses, and then repeated the process to find the relationship between the win and show odds for those horses.

We also determined how well the line or curve fit the data. The fit between the data and the curve is expressed as a "correlation coefficient" between -1 and 1. The closer the correlation coefficient is to 1 or -1, the better the fit between the data and the curve, and the more accurate the curve can be expected to be in predicting the place or show odds.

The curve (rounded to four decimal places) that best expresses the relationship between win and place odds was:

Place = (0.1754) (Win<1.5140>)

The correlation coefficient of this curve was 0.9741. The correlation coefficient was very close to 1, so we are quite confident of the projections to be made from the above formula.

The best curve expressing the relationship between win and show odds was:

Show = (0.1590) (Win<1.0358>)

The correlation coefficient of this curve is 0.5945. Because the correlation coefficient is much lower than 1, we are substantially less confident of the predictions to be made with this formula, but still confident enough to use the formula as a tool in making our calculations. In particular, we note that the two curves predict that a place bet almost always will have a higher potential payoff than a show bet, as we expected.

If we assume that, if petitioner's bets had paid off, the relationships described above would have held, then we can simply plug the win odds on the horses petitioner bet on into the formulae to obtain the place and show odds, and then use those odds to calculate petitioner's potential payoff in each race.

If petitioner's horses had won, we calculate that the odds and payoffs on those bets would have been as follows:

		1 Win		Place		Show
Race	Horse	Odds	Payoff	Odds	Payoff	Odds	Payoff
6	Mr. Tee and Eee	1.90	5.80	0.46	2.92	0.31	2.62
7	Bomb The Tote	1.00	4.00	0.18	2.36	0.15	2.30
8	Irresistible Bear	0.90	3.80	0.15	2.30	0.14	2.28
9	Mr. Oneill	0.80	3.60	0.13	2.26	1 0.05	1 2.10
10	Bleu Macadamia	0.50	3.00	1 0.05	1 2.10	.05	1 2.10

From the above chart, we can calculate the potential payoff of each of petitioner's bets:

		Petitioner's potential payoff
Race	Amount per bet	1 Win	Place	Show	Total
6	$ 500	$ 1,450	$ 730	$ 655	$ 2,835
7	2,500	5,000	2,950	2,875	10,825
8	5,000	9,000	5,750	5,700	20,450
9	8,500	15,300	9,605	1 8,925	33,830
10	9,500	14,250	1 9,975	1 9,975	1 34,200

We now can see that petitioner's potential payoff in each of the subject races was substantially (except for the seventh race) less than the amount of his accumulated debt immediately after placing his bet for each race, showing that petitioner's opportunities to win were not impaired by the small amounts in his cash drawer:

	Potential	Petitioner's
Race	payoff	debt	Difference
6	$ 2,835	$ 3,780	25%
7	10,825	11,280	4%
8	20,450	26,280	22%
9	33,830	51,780	35%
10	34,200	71,355	52%

We elaborate on these results supra p. 34.

We note that we do not think that nine races of data are sufficient to reach a scientifically precise result. To avoid the appearance of unwarranted precision, we have rounded the decimals in the above formulae to four places, and have rounded our results to two decimal places or whole dollars. The exact amounts are not as important as our conclusion that the potential payoffs were smaller, usually substantially so, than petitioner's debt.

---

Footnotes

• 1. All section references are to the Internal Revenue Code as in effect for the taxable year in issue. All Rule references are to the Tax Court Rules of Practice and Procedure.

• 2. During the day in question, petitioner sometimes placed all his bets for a race on a single ticket, and other times on several tickets.

• 3. Although OTB does not permit gambling on credit or with a credit card, it does permit bettors to deposit cash in advance so that they can place bets by telephone. These deposits are called telephone deposits. When these patrons place bets, they apparently call a central office, not the local OTB parlor.

• 4. A single bet is a bet on a particular horse to win, place, or show. An example would be a $ 2 bet on a particular horse to win a particular race. A single bet is to be distinguished from a multiple bet, such as a quinella, which requires the bettor to predict a race's first two finishers (regardless of order of finish), or a daily double, which requires the bettor to predict the winners of two races in a row. Multiple bets promise much larger returns, with a correspondingly lower probability of winning. Multiple bets are segregated from other bets and were unaffected by petitioner's activities.

• 5. Pari-mutuel betting started in the 1860's in Paris. Ainslie, Ainslie's Encyclopedia of Thoroughbred Handicapping 186 (1978). The spellings of "pari-mutuel" and "mutuel" reflect their French origin.

• 6. The Auburn OTB parlor is operated by the Western Regional Off-Track Betting Corp., headquartered in Batavia, New York. There are six regional OTB corporations covering every part of New York State. Other OTB corporations cover the Albany region, the Catskills, New York City, Nassau County, and Suffolk County. The number of parlors in the Western Region does not appear in the record, but there are 298 OTB parlors statewide. The New York legislature created the six corporations as independent entities. They all are overseen by the New York State Racing and Wagering Board.

• 7. As petitioner's bets grew, they accounted for an increasing share of all bets in the pool. The total of petitioner's bets for the 10th race accounted for 57.6 percent of the combined mutuel pool (the sum of win, place, and show pools) for that race, as reported by the Daily Racing Form, reproduced infra Appendix A.

• 8. We note that if petitioner had gambling income as the statutory notice determined, petitioner would be entitled to deduct his gambling losses, resulting in zero net income. Sec. 165(d). See discussion infra pp. 35-36.

• 9. The Third Circuit observed in Zarin v. Commissioner, 916 F.2d 110, 114 (3d Cir. 1990), that the chips the taxpayer used to gamble had no "independent economic value beyond the casino". We made a similar observation about the nature of Zarin's chips in our opinion. Zarin v. Commissioner, 92 T.C. 1084, 1086 (1989). Petitioner argues that, like Mr. Zarin's chips, petitioner's tickets were valueless. Petitioner overlooks our holding in Zarin that "the taxpayer did receive value at the time he incurred the debt" because he thereby became entitled to gamble at the casino. Id. at 1094.

  Even if the Third Circuit's opinion stood for the proposition that casino chips have no value, it would not bind us in this case, which is appealable to the Second Circuit. And even if it did bind us on its facts, this case involves an entirely different kind of gambling. In casino gambling, the gambler and the casino bet against each other, and have substantial latitude, within legal limits, in setting the terms of their wager. In pari-mutuel wagering, thousands of gamblers bet against each other, and the terms of each wager are set by law.

  The Third Circuit's position as applied to casinos is that when the house extended credit to Mr. Zarin, it knew its claim to be unenforceable, and thus the true amount of the debt (and the true price of the gambling enjoyed by Zarin) was only what the casino could have expected to collect. Zarin v. Commissioner, 916 F.2d at 115; see also Shaviro, "The Man Who Lost too Much: Zarin v. Commissioner and the Measurement of Taxable Consumption", 45 Tax L. Rev. 215, 234 n.67, 245, 256 (1990).

  In contrast, pari-mutual wagering is bottomed on the sanctity of the mutuel pool. Real cash must be paid into the pool so that winning bettors can take what they are entitled to, and the integrity of the system is thereby protected. The real cash must come from somewhere; in this case it came from OTB and Hartford. Unlike the casino in Zarin, which merely had forgone a prospective profit it knew it was unlikely to fully realize, the bettors in the pari-mutuel situation have no occasion to cancel a debt incurred by one of them, because everyone pays cash. Thus, there can be no expectation of a post-race purchase price reduction.

• 10. We postpone consideration of whether the value of the tickets was less than face value because of a difference in the amount of potential winnings. See infra p. 24.

• 11. We find petitioner had income in this case not so much because he stole tickets or other property, but because he stole and enjoyed the right to gamble. When we refer to the tickets, we do so as a shorthand way of referring to the rights evidenced by the tickets. Stripped of the rights to gamble, the tickets would have been worthless.

• 12. If petitioner lacked total control over the value, his income would equal only the value over which he did have control. As with other valuation questions, we address this question in our valuation analysis, infra p. 24.

• 13. Although the debt to OTB was not formalized until OTB's insurance company obtained a judgment in 1989, we are persuaded that petitioner knew that he was obligated to pay OTB for any losses that he caused the corporation. In particular, petitioner acted as though the debt was a real one, albeit one he could not actually pay. For all practical purposes, the debt arose and increased as petitioner placed his bets.

• 14. The Haig-Simons definition is particularly appropriate in cases of theft of entertainment, services, and other consumables. We need not and do not reach the question of whether all amounts that might be swept into income under the Haig-Simons definition are includable in gross income under the Internal Revenue Code. See also infra p. 31 (discussing psychic income).

• 15. Of course, Congress has statutorily excluded from gross income or deferred taxation on certain employee benefits, such as health insurance and qualified pension and profit-sharing plans, but these are exceptions to the general rule of sec. 61. Commissioner v. Kowalski, 434 U.S. 77, 83 (1977).

• 16. As an employee achievement award, the tickets might have been excludable under sec. 74(c), to the extent the award did not exceed the limitations on the employer's deduction under sec. 274(j). At most, petitioner could have excluded $ 1,600. Sec. 274(j)(2)(B). The remaining $ 78,680 would have been included in income.

• 17. Respondent would have us hold, as a matter of law, that the face amount of the tickets was their fair market value. Indeed, the New York Penal Law specifies that the value of stolen tickets is their face amount. N.Y. Penal Law sec. 155.20 (McKinney 1988 & 1992 Supp.). While we have found, under the facts of this case, that fair market value equals face amount, we do not adopt respondent's view. Face amount is indicative of the harm petitioner inflicted. The face amount of petitioner's tickets was used to determine how seriously he would be punished (and how much he owes to his employer's insurer). However, the New York courts have recognized:

  The purpose of the statutes fixing the higher degrees of crime is not related to regulating the economic market but to assessing the scale of criminal operations by the persons charged with offenses under the statutes. [People v. Colasanti, 322 N.E.2d 269, 271 (N.Y. Ct. App. 1974).]

  The face amount of the tickets does not necessarily show the value of the benefits to petitioner, and thus does not measure his income for income tax purposes.

• 18. For simplicity's sake, the example in the text assumes that petitioner placed a single "win" bet of $ 15,000 on the eighth race. He actually placed three $ 5,000 bets -- a win bet, a place bet, and a show bet.

• 19. Petitioner appears to have been betting on the "law of averages":

  The limited usefulness of averages becomes painfully noticeable when innocent horseplayers decide to take advantage of the so-called law. Favorites win about 32 percent of races, year after year. That is, one race in every three, when calculated on an annual basis. But this guarantees nothing about the winning percentage of favorites during the few days on which the horseplayer decides that profits will be obtained by betting all favorites, doubling the bet after each loss. For favorites to lose 15 races in succession is perfectly compatible with a year-long winning expectation of .32. Starting with a $ 2 bet, the follower of the law of averages would need $ 32,768 for the fifteenth bet, if still playing the horses. [Ainslie, Ainslie's Encyclopedia of Thoroughbred Handicapping 29 (1978).]

• 20. Sec. 165(d) states: "Wagering Losses. -- Losses from wagering transactions shall be allowed only to the extent of the gains from such transactions."

• 21. At 6 percent interest with interest assessed weekly, it would have taken the embezzler more than 58 years to repay the $ 21,000 balance. At an interest rate greater than 6.19 percent, the principal amount would never be repaid.

• 22. But see Antonides v. Commissioner, 91 T.C. 686, 704 n.9 (1988), expressly reserving the question of whether the regulation validly requires that substantial authority have existed when the return was filed. We need not reach that question now.

• 23. The anticipated amount might have been different.

• 1. Equal to $ 2,280 in initial bets plus $ 1,176.40 in winnings taken in the form of bets.

• 2. Equal to petitioner's actual closing balance of $ 843.60 minus the amount paid to him in cash under this example, $ 843.60.

• 24. He also could have split his bets by type, placing his win, place, and show bets on different tickets. Place and show bets typically have much lower payoffs than win bets, so this strategy could have had the same success as the one described in the text.

• 1. Equal to petitioner's actual closing balance of $ 843.60 minus the amount paid to him in cash under this example, $ 300.

• 25. We also are puzzled that petitioner apparently was not charged the usual 5-percent OTB fee in the computation of his winnings in either the 9th or 10th races. After the surcharge, petitioner should have won $ 8,479 in the 9th race and $ 32,490 in the 10th race. Inasmuch as respondent did not raise this issue, we find it unnecessary to attempt to explain it.

• 1. The odds are expressed as the amount of potential winnings per dollar bet. The odds do not take into account the amount of the bet. For instance, the odds on Mac's Boy in the first race were $ 10.80 per $ 1 bet. A $ 2 bet on Mac's Boy would have returned $ 21.60, plus the $ 2 bet, for a total of $ 23.60, as shown in the chart labeled "$ 2 Mutuel Prices" for the first race in Appendix A.

• 1. All win odds and payoffs, and the other footnoted odds and payoffs, are actual results, not those projected by our analysis.

• 1. All win payoffs, and the other footnoted payoffs, are actual results, not those projected by our analysis.