Kohler Company, Plaintiff-appellee, v. United States of America, Defendant-appellant, 468 F.3d 1032 (7th Cir. 2006)

Kohler, the well-known manufacturer of plumbing products, brought suit for a refund of federal income taxes. It won on summary judgment, 387 F. Supp. 2d 921 (E.D. Wis. 2005), and the government appeals.

In 1986, Kohler decided to build a plant in Mexico that it estimated would cost at least $29 million. It needed pesos in order to pay for land, building contractors, and other inputs. How to get them?

Now it happened that Mexico had defaulted on its foreign debt, and in an effort to restore its credit had adopted an ingenious "debt-equity swap" program (pioneered by Chile). The program entitled a foreign company that wanted to invest in Mexico, and therefore needed pesos, to purchase defaulted Mexican dollar-denominated debt on the open market and then swap it with the Mexican government for pesos that could be spent only in Mexico rather than exchanged for dollars. International Business Corporation, Debt-Equity Swaps: How to Tap an Emerging Market 1 (1987). The program enabled the Mexican government to retire some of its foreign-owned debt without having to pay "hard" money—that is, foreign currency, or, what would amount to the same thing, pesos convertible to foreign currency.

Bankers Trust, the American bank, owned Mexican debt in the face amount of $22.4 million. This debt traded at a substantial discount because of Mexico's default, fiscal instability, and general lack of creditworthiness. As a result, Kohler was able to buy the debt from Bankers Trust for only $11.1 million, slightly less than half its par (face) value. The bank preferred the bird in the hand (11.1 million U.S. dollars) to two birds, consisting of claims against the Mexican government, very deep in the bush.

Kohler knew that under the terms of the debt-equity swap program the Mexican government would swap the $11.1 million debt that Kohler had bought from Bankers Trust for $19.5 million worth of pesos as calculated at the then current market exchange rate of 2245 pesos to the dollar. The qualification in "as calculated at the then current market exchange rate" is critical. If for one reason or another that was not the right exchange rate to use for this transaction, the pesos that Kohler received may not really have been worth $19.5 million. That they were worth less is shown by Mexico's willingness to offer $19.5 million in pesos for debt that Kohler had purchased for only $11.1 million. Mexico had to compensate Kohler for accepting pesos that came with restrictions that reduced their dollar value. The pesos had to be spent in Mexico on projects approved by the government and could not be freely converted to dollars or other foreign currencies until 1998. So although the market exchange rate was, as we said, 2245 pesos to the dollar, Kohler received a rate of 3939 pesos to the dollar, which is what turned $11.1 million of dollar debt into $19.5 million in pesos. Kohler did however use all the pesos to pay for real estate and other costs that it incurred in building its plant.

On its federal income tax return it treated the purchase of the debt and its sale to the Mexican government as a wash, yielding no taxable income, just as if the government had paid it $11.1 million in dollars rather than paying it in pesos. The Internal Revenue Service disagreed with this treatment and instead added to Kohler's taxable income for 1987, the year of the transaction, the difference of $8.4 million between the price that Kohler had paid Bankers Trust for the Mexican debt and $19.5 million.

One might have thought that the way to account for Kohler's purchase of Mexican debt would have been to add $11.1 million to the basis of Kohler's investment in the Mexican plant, so that if it ever sold the plant the difference between on the one hand the sale price and on the other hand the sum of $11.1 million and all the other costs of the plant would be the taxable income attributable to the sale. Then if the Mexican government's purchase of $11.1 million in debt from Kohler for $19.5 million in pesos was a windfall for Kohler, reducing the real cost of the plant, Kohler would realize a greater profit from the eventual sale of the plant than it would have realized otherwise, and that profit would be taxable. Even if the plant was never sold, the windfall would give Kohler higher profits (presumably taxable) on sales of the plant's output because the deductions from taxable income that it could take for depreciation of the cost of the plant would be lessened by the $8.4 million reduction in its basis.

An alternative way of accounting for the swap would have been to accept Kohler's argument that the value of the debt that it purchased was unascertainable at the time of purchase and treat the exchange of the debt for the peso account as a swap yielding no taxable income. Any capital gains that resulted in the future from Kohler's use of the pesos to purchase goods and services for its project would be taxable. So if it used the entire amount to buy real estate and construction services before any change in the exchange rate, it would be deemed to have realized a capital gain of $8.4 million ($19.5 million minus $11.1 million) on the purchase.

Still another alternative would be to deem the difference between the two amounts a contribution of capital to Kohler's enterprise by the Mexican government. Such a contribution would not be included in Kohler's gross income, 26 U.S.C. § 118(a), though it would be recorded on Kohler's books as having a zero basis, 26 U.S.C. § 362(c), and so could not be depreciated. Although this approach was adopted in the nearly identical case of G.M. Trading Corp. v. Commissioner of Internal Revenue, 121 F.3d 977 (5th Cir. 1997), we are dubious about it. Compensation for a "specific, quantifiable service" cannot be classified as a contribution to capital, United States v. Chicago, Burlington & Quincy R.R., 412 U.S. 401, 413, 93 S. Ct. 2169, 37 L. Ed. 2d 30 (1973)—and the Mexican government, to the extent it "overpaid" Kohler for the bonds, was buying a service from Kohler: retirement of a part of Mexico's foreign debt. See Scott A. Shane, "A U.S. Policy Toward Debt-Equity Swaps," 16 J. Soc., Pol. & Econ. Stud. 287 (1991); Morris B. Goldman, "Debt/Equity Conversion; A Strategy for Easing Third World Debt," Heritage Foundation Reports 1 (Jan. 21, 1987).

The court in G.M. Trading thought the purpose of the Mexican debt-equity swap program was to encourage foreign investment in Mexico. That was a purpose, but it was secondary to Mexico's desire to retire its foreign debt—the service for which it paid Kohler by exchanging dollar debt for pesos. In deciding at what rate to exchange foreign debt for pesos, moreover, Mexico ranked projects according to their investment value, and Kohler's type of project was rated below several others, such as projects designed to privatize state industries. International Business Corporation, supra, at 56-57; Morgan Guarantee Trust Company, "Debt Equity Swaps," World Finance Markets 14 (June-July 1987). The debt held by companies that planned to use their pesos for the investments most favored by the government was redeemed in pesos at par. Remember that the par (face) value of the debt that Kohler bought from Bankers Trust was $22.4 million, or 50.4 billion pesos at the market exchange rate of 2245 pesos per dollar. Kohler was offered only 87 percent of this amount (43.8 billion pesos). Mexico would not have gone out of its way to encourage Kohler's project had it not been for the opportunity to retire some of its foreign debt. In fact it was Kohler—whose decision to build the plant predated the swap program—that approached the Mexican government about initiating a swap, rather than vice versa.

No doubt the government's motives were mixed, as indicated by the fact that some companies that tendered dollar debt for redemption in pesos were given the less attractive exchange rate of 3399 to the dollar, compared to Kohler's 3939; their projects were not the kind of foreign investment that the government especially wished to attract. Kohler's project was what is called a "maquiladora," a project whereby (in the usual case) a plant imports raw materials into Mexico for processing into finished products that are exported. Thus, as a further condition of the swap, Kohler promised to export at least 20 percent of the output of its plant, which would earn dollars for Mexico, which wanted to encourage foreign investment that would build its dollar holdings. That condition doubtless induced the favorable exchange rate that Kohler received, and maybe the difference between that rate and the bottom rate of 3399 pesos per dollar, translated into dollars, could be considered a contribution to capital by Mexico.

There is no need to pursue the issue. The parties have taken none of the paths we've laid out. (The second—the wait-and-see approach—strikes us as the most practical, as it involves no conjecture.) They treat the sale of the Mexican debt for the peso account as just that—a taxable sale—consistent with the rule that an exchange of "materially different" things (the Mexican dollar debt for the pesos) is an event in which profit or loss is realized. 26 U.S.C. § 1001(c); 26 C.F.R. § 1.1001-1. Cottage Savings Ass'n v. Commissioner of Internal Revenue, 499 U.S. 554, 556, 111 S. Ct. 1503, 113 L. Ed. 2d 589 (1991), is illustrative: "a financial institution realizes tax-deductible losses when it exchanges its interests in one group of residential mortgage loans for another lender's interests in a different group of residential mortgage loans."

The parties quarrel only over the value to Kohler of the exchange when made. The quarrel has driven them to take opposite positions, both untenable. Kohler argues that it had no gain from the sale at all, while the Internal Revenue Service argues that the entire difference between the $19.5 million in pesos that the Mexican government gave Kohler and the $11.1 million that Kohler had paid to buy the debt that it swapped for the pesos was taxable income to Kohler. Kohler's position is untenable because $11.1 million in Mexican foreign debt was worth more to it than to Bankers Trust. It wanted pesos; Bankers Trust did not. Kohler argues absurdly that if it gained from the purchase, the bank must have lost, and why would it sell at a loss? Most transactions produce a gain to both parties—that is what induces the transaction.

Yet the pesos were not worth the full $19.5 million at which the Mexican government valued them for purposes of the exchange, because they were not convertible into dollars or any other currency. They could be used only in Mexico and in fact only to build the intended plant. Had Kohler decided not to build the plant, because of changed conditions after its purchase of the debt from Bankers Trust, it would have been battered by the severe inflation that afflicted Mexico throughout the 1980s. That is why we suggested earlier that the dispatch with which Kohler spent its pesos would determine the actual value of the exchange to it (the "wait-and-see" approach). A dollar restricted to being used to purchase the currency of a country in the throes of a financial crisis is worth less than a dollar.

How to choose between adversaries' valuations when both are manifestly erroneous? The conventional response would be that the party with the burden of proof (in the sense of the burden of persuasion) would lose. And that is Kohler—and would be, by the way, even if it had not paid the additional tax assessed by the IRS but instead had challenged the deficiency in the Tax Court. Tax Ct. R. 142(a); Kikalos v. Commissioner of Internal Revenue, 434 F.3d 977, 982 (7th Cir. 2006); Leo P. Martinez, "Tax Collection and Populist Rhetoric: Shifting the Burden of Proof in Tax Cases," 39 Hastings L.J. 239, 257-60 (1988).

But Kohler argues that it needs no evidence, citing United States v. Davis, 370 U.S. 65, 82 S. Ct. 1190, 8 L. Ed. 2d 335 (1962), a superficially similar case won by the taxpayer. Pursuant to a divorce settlement, Davis agreed to transfer stock to his wife in exchange for her surrender of her marital property rights. In effect he bought those rights for the value of his stock, just as Kohler in effect bought pesos from the Mexican government for $11.1 million, since the money it paid Bankers Trust was the only outlay it made to get the pesos. The Court in Davis held that the only taxable gain on the transaction was the difference between the market value of the stock and the taxpayer's basis—not the difference between the value of the wife's marital rights, corresponding to the pesos that Kohler acquired in this case, and the taxpayer's basis. The Court reasoned that "absent a readily ascertainable value" of the acquired property, it should be assumed to be equal in value to what the taxpayer had paid for it. Id. at 72, 82 S. Ct. 1190. Otherwise, as the Court explained, the wife would not know, if she should later sell the stock, what her basis was—that is, what she had paid in exchange for the stock by giving up her marital rights. Id. at 73, 82 S. Ct. 1190.

But the Court merely assumed, it did not hold, that the wife's marital rights could not be ascertained with sufficient precision to enable a calculation of the taxpayer's "real" gain (or loss). The Court of Claims had held that because in its view the value of those rights could not reasonably be ascertained, their exchange for the taxpayer's stock was not a taxable event. The Supreme Court, assuming—but not ruling on—the soundness of the Court of Claims' finding on ascertainability, held that the exchange was still a taxable event, only one in which the only gain realized was the difference between the market value of the stock (it was publicly traded—it was DuPont stock) and the taxpayer's basis in the stock. Id. at 71-73, 82 S. Ct. 1190. In other words, if property received in an exchange cannot be valued, the taxable gain is limited to the difference between the sale price and the seller's basis.

The problem in our case is different. It is what to do when the value of the property exchanged may well be ascertainable but has not been ascertained. To permit the Internal Revenue Service to place an arbitrary value on difficult-to-value property obtained in a transaction and require the taxpayer to prove that it was worth less—and exactly how much less—would place an unreasonable burden on taxpayers. Suppose a lawyer and a dentist bartered legal services for dental services and the IRS assessed the legal services as worth only $10,000 and the dental services as worth $1 million and so assessed $990,000 in additional taxable income to the lawyer. The government would have to present some evidence in defense of its extravagant assessment before the burden of production and persuasion would shift to the taxpayer. This conclusion is implicit in cases that hold that when the IRS makes a "naked" assessment, which is to say one "without any foundation whatsoever," the taxpayer does not have to prove what the assessment should have been. United States v. Janis, 428 U.S. 433, 440, 96 S. Ct. 3021, 49 L. Ed. 2d 1046 (1976); see also Helvering v. Taylor, 293 U.S. 507, 55 S. Ct. 287, 79 L. Ed. 623 (1935).

So here, the government's assessment was undeniably excessive because it took no account of the restrictions that the seller of the pesos (the Mexican government) had placed on the purchase. Among the restrictions is one that we haven't mentioned yet: Kohler was forbidden to trade its pesos with Mexicans for dollars (Mexico didn't want dollars going out of the country), so that if it had decided against building the Mexican plant and had no other use for pesos it would have had to exchange them for dollars with other foreign companies planning similar or (as judged by Mexico) inferior projects. If, for example, a company was contemplating a project that the Mexican government thought so desirable that it would redeem the company's Mexican debt at par ($22.4 million in pesos versus the $19.5 million in pesos that Kohler received), the company could deal directly with the government rather than buying Kohler's pesos. It would buy those pesos only if Kohler gave it a discount that would make the buyer as well off as if he had dealt directly with the Mexican government.

We think the Internal Revenue Service had either to prove against all probabilities that its assessment was correct or pick a number that was prima facie plausible—a number somewhere in between $11.1 million and $19.5 million. Its effort, by means of an expert witness, to prove that the pesos were indeed worth $19.5 million fell pathetically short of the mark. The expert had not attempted to calculate the discount that a purchaser of restricted pesos would have demanded. Kohler's efforts to show that the pesos it received from the Mexican government were worth the same as the debt it had exchanged were equally pathetic. Kohler was committed, though apparently not irrevocably, to a project that would cost more in pesos than the pesos it was obtaining from the Mexican government. Although the pesos obtained in the swap wouldn't be spent all at once, the government had guaranteed that until they were spent they would earn interest at a high rate and be guaranteed against any devaluation of the peso (though not against inflation). Given Mexico's parlous financial situation, the transaction was not riskless to Kohler. But Kohler would not have paid $11.1 million to obtain pesos from the Mexican government had it not thought that the government's offer to give it 75 percent more pesos than it could have bought on the open market for $11.1 million ($19.5-$11.1=$8.4 ÷ $11.1 =.75) would yield it a profit.

The same thing can be worth more to one person (Kohler) than to another (Bankers Trust); that is the basis of market transactions. To a holder of Mexican debt that had no use for pesos, the debt was worth only half its face amount; to someone like Kohler who needed a great many pesos, the debt was worth more. How much more? Not $8.4 million more; and we have said that before a taxpayer can be required to disprove an extravagant evaluation the Internal Revenue Service must present some evidence to support it. The Service presented no evidence that could have persuaded a rational factfinder that the pesos Kohler got from the Mexican government in exchange for the debt it surrendered were worth $19.5 million. The Service could have justified a more modest estimate yet one well above $11.1 million, but clinging stubbornly to its untenable valuation it suggested no alternative to $19.5 million. It played all or nothing, lost all, so gets nothing.

AFFIRMED.