FLORIDA TAX REVIEW
VOLUME 3 1997 NUMBER 10

More on Accounting for the Assumption of
Contingent Liabilities on the Sale of a Business

Charlotte Crane*

I. INTRODUCTION ................................ 616

HI. SUMMARY OF PROFESSOR HALPERIN'S CONCLUSIONS .... 617

I. WHY IS A BANK LOAN DIFFERENT FROM A CONTINGENT

LIABILrrY .................................... 619

IV. THE PREMISES UNDERLYING SURROGATE TAXATION ..... 623

V. WHAT IF THE SELLER'S DEFERRAL HAD NOTHING TO DO

WITH SURROGATE TAXATION? . . . . . . . . . . . . . . . . . . . . . . 627

VI. WHAT IS AN ASSUMED LIABILITY? . . . . . . . . . . . . . . . . . . . 635

VII. RECONSIDERING THE ALTERNATIVES FOR ACCOUNTING
FOR THE ASSUMPTION OF CONTINGENT LIABILITIES ..... 636
A. Professor Halperin's Preferred Method .......... 637
B. First Alternative Method: Accounting for Price

Adjustments as Amounts are Paid .............. 638
C. Professor Halperin's Second Alternative Method ... 640

VIII. WHAT IF THE EXPECTED VALUE CANNOT BE ACCURATELY

DETERMINED? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .  641

IX. SHOULD THE ACCOUNTING FOR A CASH PURCHASE BE

THE ONLY NORM FOR EVALUATING POSSIBLE

ACCOUNTING TREATMENTS? . . . . . . . . . . . . . . . . . . . . . . . . 643

X. CONCLUSION .................................. 645

* Professor of Law, Northwestern University School of Law. I want to thank

Professor Halperin for his patience in our correspondence over many years on this and related
subjects. Any errors remain my own. This project also benefited from the financial support of
the Linthicom Foundation and the research assistance of Michael D. Cohen.



Florida Tax Review

I. INTRODUCTION

In his article "Assumption of Contingent Liabilities on Sale of A
Business," which appeared in this journal last year,' Professor Daniel
Halperin ably describes the stakes involved in properly accounting for
contingent liabilities assumed by a buyer in asset acquisitions. His primary
conclusion is that a buyer of a business who assumes contingent liabilities
should not be allowed, as some have urged, a deduction upon payment on
these liabilities. Although I agree with much of Professor Halperin's analysis
and almost all of his conclusions, I disagree with some of his underlying
assumptions about the universal nature of contingent liabilities.

Professor Halperin's analysis relies heavily on the related underlying
assumptions that (1) there is always a legitimate reason for deferring a
deduction until the person to whom the contingent liability is owed has been
identified and that person has taken the amount into income and (2) more
often than not, the expected value of the payments to be made on the liability
can reasonably be valued in advance. If either of these assumptions is
unfounded, the problems posed by contingent liabilities are more difficult
than Professor Halperin's analysis might suggest, although his ultimate
resolution of the situation may nevertheless be the best one available.

This comment will explore Professor Halperin's underlying assump-
tions regarding contingent liabilities and examine the stakes involved in
fashioning tax accounting rules that rely upon them. It will attempt to
demonstrate that at least some liabilities, including some contingent
liabilities,2 are different from the liabilities upon which Professor Halperin's
analysis is based. This difference may not easily be accommodated in an
income tax based on realization. Nevertheless, this difference may undermine
both the assumption that there is a legitimate reason for deferring a payor's
deduction until the ultimate payee is known and has been taxed, and may
make it substantially more difficult to establish an appropriate value for the
liability in advance of payment.

1. Daniel I. Halperin, Assumption of Contingent Liabilities on Sale of a Business,
2 Fla. Tax Rev. 673 (1996) [hereinafter Halperin, Assumption].

2. As did Professor Halperin, these comments will ignore those contingent liabilities
that represent payments for which no payor would ever be entitled to an offset against income,
whether as a current deduction or an addition to basis. Such liabilities include federal income
taxes and amounts limited by the provisions of §§ 162(c) and 162(0. The former most likely
should be accounted for at the time of sale by requiring the seller to take an amount into
income without offset. The latter are policy deviations from the results which would be
produced by strict adherence to technical notions of income, and, as such, their implementation
in transactions involving the assumption of such liabilities is more appropriately viewed as a
matter of policy as well.

[Vol 3:10



Accounting for the Assunption of Contingent Liabilities

My disagreement with Professor Halperin about the variations among
contingent liabilities has little practical consequence if (as may well be the
case) workable tax accounting rules cannot be fashioned to accommodate the
differences among such liabilities and other deferred payments. It may also
be true that failing to deal with these cases separately will not produce
intolerable results. Nevertheless, I fear that some readers may ignore his
conclusions because, from their point of view, he seems to be concerned only
with the easy cases for which his assumptions seem reasonable. Since
Professor Halperin's analysis does not appear to have taken these harder cases
into account, readers who are concerned primarily with the harder cases are
likely to reject his analysis too quickly. When Professor Halperin ignores the
variations among contingent liabilities, he may make it too easy for critics to
reject his conclusions as extreme.

II. SUMMARY OF PROFESSOR HALPERIN'S CONCLUSIONS

Professor Halperin would require that buyers assuming contingent
liabilities include the expected value of the liabilities in the basis of the assets
acquired at the time of the acquisition. - If the payment on the liability is
made "as expected," no further deduction would be allowed, even for an
interest component.4 This treatment differs from the accepted treatment of
assumed fixed liabilities, in that when fixed liabilities are assumed there is
both immediate basis available for the use of the loan proceeds and an
allowance for interest,5 frequently taken as such interest accrues, and
certainly taken no later than payment. This proposed treatment for contingent
liabilities is a variant of the device familiar to readers of Professor Halperin's
earlier work involving surrogate taxation.6

Professor Halperin also acknowledges that an equivalent result could
be obtained by allowing the buyer full deductibility of any amounts actually
paid when they are paid, but only after requiring, somewhat artificially,
inclusion of income either at the time of the sale or, with an adjustment for
the passage of time, at some later time. (The equivalence of these choices and
the options available when payments are not made as expected are explicated
more fully below.)

3. Halperin, Assumption, supra note 1. at 688-700, 711.
4. Id.
5. This allowance is generally a deduction, unless the nature of the project is such

that interest should be capitalized. Throughout these remarks. "allowance of a deduction" will
be used, unless indicated otherwise, on the assumption that the item is one that would have
given rise to a deduction no later than when paid if no transfer of assets had occurred.

6. Daniel I. Halperin, Interest in Disguise: Taxing the "Time Value of Money," 95
Yale L.J. 506 (1986) [hereinafter Halperin, Disguisel.

1997]



Florida Tax Review

At the heart of Professor Halperin's analysis is his demonstration that
allowing the buyer a deduction for an assumed contingent liability plus
interest when payment is made would be equivalent to allowing the buyer a
deduction for the liability at the time of the sale. Therefore basis credit should
be allowed only for the liability itself (with adjustment if this allowance is
deferred), with no additional deduction for interest paid. He demonstrates this
equivalence by first pointing out the windfall that would accrue to the seller
of a business if it were allowed a deduction for the interest paid on a fixed,
but as yet undeducted liability, such as deferred compensation. This is the
context in which surrogate taxation has been most readily accepted, and is not
surprisingly Professor Halperin's starting point.

Professor Halperin's reasoning is straightforward and familiar. If an
employer and an employee agree to defer for one year payment of compensa-
tion (Professor Halperin uses $100) for work already performed, the tax rules
should leave them (considered together) no better off than they would be if
the compensation were paid currently. Presumably the employee could have
earned interest on the compensation only at the after-tax rate of interest.
Therefore the employee could have taken the money, paid tax on it and
invested the rest, and, assuming a 40% tax rate and a 10% interest rate, ended
up after one year with $63.60 [net wages: $100 - ($100 x .40)] + [net
interest: ($60 x. 10) - ($60 x .10)(.40)]. Given that the employer which defers
payment need only equate the employee's situation with this $63.60 result,
it need only pay $106 at the end of the year ($106 - ($106 x .40) = $63.60).
Therefore the employer must not be allowed to deduct the compensation as
if it had been paid or to deduct a pre-tax interest amount, because then it
could earn more than it logically should be required to pay the employee. (If
it invests the unpaid $100 at 10%, and pays tax on the $10, it will have $106
to pay the employee. If it has any better tax treatment, it will be able to
deduct the tax savings and earn more.) The employer can be allowed only a
deduction for this $106, the amount originally owed with an adjustment for
the deferral. This adjustment for the deferral gives the employer a deduction
of $106 at the end of the year, a deduction which has the same value to the
employer as a deduction of $100 at the time the compensation was earned.7

The lesson demonstrated is that so long as the employee is not taxed
as the compensation is earned, the employer cannot be allowed a deduction
as the compensation is earned without producing a lower tax burden simply
as a result of the deferral of payment This lowering of the tax burden is

7. Halperin, Assumption, supra note 1, at 678-81 (relying on Halperin, Disguise,
supra note 6, at 511, 522-23).

8. Under Professor Halperin's approach, as is generally the case under current law,
the deduction is allowed when the payment is made. The amount allowed as a deduction is
greater than the amount that would have been allowed at the earlier time. This larger deduction

[Vol 3:10



Accounting for the Assunption of Contingent Liabilities

prevented by giving the employer a deduction for the full amount of the
payment only at the time of payment, and not allowing the employer to
accrue and deduct interest. This deduction of the amount actually paid
effectively compensates the employer for the delay in its deduction because,
assuming unvarying tax rates and the use of the correct discount rate, a
deduction of the future value of any given payment will have the same
economic effect as a deduction of the present value of the payment in an
earlier year.9

This same principle must be followed when the employer's business
is sold in a taxable transaction, and the obligation to pay the employee is
assumed by the buyer. The buyer's liability must be treated in a way that
requires him to set aside an amount no less than the seller would have had
to set aside to provide the employee with an after tax amount of $63.60. The
treatment that produces this result requires the buyer to add the amount paid
to the basis of the assets acquired when it is paid, or in the alternative, to add
the present value of the liability to the basis of assets acquired. In either case,
there should be no further deduction when the amount is actually paid.' 0

M. WHY IS A BANK LOAN DIFFERENT FROM A CONTINGENT LIABILITY?

The deferred payment (and related contingent payment") situations
are to be contrasted, in Professor Halperin's analysis, with the situations
involving the proceeds of an ordinary bank loan. In the case of an ordinary
bank loan, the amount of the proceeds are generally available to be deducted
(or added to basis, depending upon the use of the proceeds) when received
and used; amounts subsequently paid to the lender for the use of its money
are deductible as interest. If a buyer of the business assumes this bank

compensates the employer for the deferral of its deduction. It is in ordinary circumstances the
same as the amount that would be allowed if the employer were allowed a deduction for the
original amount, together with an additional element of interest at the implicit after-tax rate.

9. In the ordinary case, the correct interest rate is the rate that the taxpayer could
earn after-tax. Thus, if the after-tax discount rate is 10%. a deduction of S91 in year I will
have the same economic impact as a deduction of S100 in year 2 or SI10 in year 3. Assuming
a 30% tax rate, such deductions would give the taxpayer a current benefit of S27, S30, or S33,
respectively. The savings in years 1 or 2 could be invested to produce an overall yield of S33
in year three.

10. Halperin, Assumption, supra note 1. at 684-85. The present value for these
purposes is computed at the after-tax rate of return, since the buyer will incur tax liability on
the interest earned if it has set aside a fund to meet the obligation.

11. "All contingent liabilities fall into the second of the two categories described
in Part 1-typified by the deferred compensation arrangement-because the seller is allowed
no deduction, basis, or other tax attribute until the contingency is resolved." Halperin.
Assumption, supra note 1, at 688-89.

19971



Florida Tax Review

liability, the buyer is entitled to add the principal amount to basis, and to
deduct the interest paid or accrued."

Professor Halperin asserts that contingent liabilities assumed in asset
transfers must be treated like deferred compensation because they, like
deferred compensation, have not yet been taken into account by the seller.'3

Ideally, the purchaser should be allowed basis for the expected economic
value of the liability at the time of the purchase. In the alternative, an
allowance for this amount, adjusted for deferral, should be given when the
liability is paid. But in no event should there be a separate deduction for the
interest component of the amount paid."4 If such liabilities had been taken
into account by the seller, as the use of the proceeds of a bank loan are, then
apparently Professor Halperin would allow a deduction for an interest
component.

There is something circular about Professor Halperin's sorting of
liabilities into these two classes: those which have given rise to tax attributes
for a seller/obligor (the classic bank loan that bears interest) and those which
have not (the compensation payment which has been deferred). The fact that
the proceeds have been taken into account by the seller/obligor should not,
in itself, be a reason for distinguishing between these two types of liabilities.
The real distinction between these cases is the prior fact that the typical
lender will have lent after-tax funds, while the typical employee will be
deferring tax on the deferred payment, and will therefore, in effect, have lent
before-tax funds.

12. In the body of the text, Professor Halperin does not consider the question of
how much interest on such liability should be deducted or when it should be deducted. These
issues are considered only in the abstract, in Appendix I. Halperin, Assumption, supra note 1,
at 712-14. In theory, only the present value of the liability, calculated using the after-tax
discount rate of the buyer, should be treated as the liability assumed by the buyer and added
to the basis of the assets purchased. Unless there is a renegotiation of the debt, however, the
terms of the debt will not be restated, and the tax accounting for the assets purchased will
produce imperfect results. If the discount rate faced by the buyer is higher than that faced by
the seller when the debt was first incurred, the seller will have gain relating to the debt
mischaracterized as gain on assets, and the buyer will have an overstated purchase price and
an understated interest cost. If the discount rate faced by the buyer is lower than that faced by
the seller when the debt was first incurred, the seller should have a loss relating to the debt,
and the buyer will have an understated purchase price and an overstated interest cost. (This
summary assumes that the buyer's burden should be the measure of the seller's gain or loss
on its debt position. Other more complicated approaches could be used.)

13. Professor Halperin outlines these two classes of liability in Halperin,
Assumption, supra note 1, at 676-77.

14. Halperin, Assumption, supra note 1, at 684-86.

[Vol 3:10



Accowzting for the Assumption of Contingent Liabilities

One can ordinarily assume that loans are made with funds that are
after-tax to the lender.' 5 When the loan proceeds are transferred to the
borrower, there is no additional value added to the tax base. There has only
been a transfer of previously included values. Therefore the transfer of the
funds can be ignored without omitting any values from the overall tax
base.' 6 As a normative matter, interest on this ordinary bank loan will be
deductible to the payor. There is no reason not to allow a deduction of
interest on amounts that represent earnings on values that have already been
taxed, especially when, in general, these interest earnings can readily be taxed
to their owner, the lender. 7 There is no untaxed principal, and no untaxed
interest. Since both parties will be aware of this treatment, the negotiated
interest can be assumed to be stated at a pre-tax rate.

On the other hand, the typical employee has not yet taken into
income the amount that he is, in effect, lending the employer. To the extent
that this value has not yet been accrued, there has been a value omitted from
the overall tax base. In Professor Halperin's deferred compensation situations,
for instance, the earned compensation should have been taxed to the
employee. Because it has not yet been taxed to the employee, the employer's
deduction is deferred until payment. When the compensation is paid, both the
employer and the employee are taxed in ways that put them in the same
position that they would have been had the compensation been paid and
accrued earlier. The employer is allowed a deduction that includes an interest-
like increment to make up for the deferral of the deduction for the payment,
and the buyer similarly must take into account an amount that includes an
interest-like increment to make up for the deferral of the inclusion. Because
the employer will, in fact, be paying the employee's income tax on the
interest on the not-yet transferred fund, the employer and employee can be

15. Not all loan proceeds are made with values that can be assumed to be after-tax.
For instance, sales with deferred payments that are accounted for on the installment method.
amount to loans of pre-tax values. The interest charge under § 453B. and the rules precluding
installment treatment of recapture gain under this section have substantially limited the
circumstances in which loans can be made with pre-tax proceeds. Similarly, before the 1984
changes to the methods for reserving for bad debts, some lenders could effectively deduct a
substantial part of their loan proceeds, making them effectively pre-tax to that extent.

16. It appears that Professor Halperin would be willing to allow the retransfer of
such after-tax funds, even when a traditional loan is not involved. Professor Halperin simply
assumes that the proceeds of a bank loan can be retransferred to a third party without the third
party taking the amount into income. Halperin, Assumption. supra note 1. at 682 n.19.

17. If interest is not charged to the lender, only the interest deduction should be
subject to the same surrogate taxation device. A deduction should be allowed not as interest
accrues, but only as it is paid or taken into account by the lender. No adjustment needs to be
made to account for the principal, since it has already been taxed to the lender, and has not
been deducted by the lender on the transfer to the borrower.

1997]



Florida Tax Review

assumed to have agreed that the interest-like increment should be accruing at
an after-tax interest rate. Under these assumptions, the allowance of deferred
deductions for principal, with compensation to the employer for this deferral,
is appropriate. 8

In sum, I believe that the criterion Professor Halperin chooses to
focus on, whether the liabilities in question have given rise to tax attributes
for the seller/obligor, does not provide the true distinction between these two
types of liabilities. The link identified by Professor Halperin between the
seller/obligor's tax attributes and the deductibility of interest is useful only
because, in most situations, the rule for allowing the seller/obligor's tax
attributes to be taken into account provides a good substitute test for
determining the likelihood of prior income inclusion by the effective lender.
There is much in the relevant tax accounting rules that conditions the seller's
ability to take a deduction into account upon the likelihood that the ultimate
recipient has taken the amount into income. The fact that the amount has
been taken into account by the seller is only a likely indicator of the fact that
the ultimate recipient has taken the amount into income. Both the granting of
tax attributes to the seller and the allowance of an interest deduction are
conditioned on the likelihood that the ultimate recipient has taken the amount
in question into income, and that, therefore, no amount has been omitted from
the tax base.

Surrogate taxation is only necessary when values that should have
been included in the tax base will remain excluded or their inclusion is
deferred. Deferred payment for values previously created by an employee, in
effect, leaves those values out of the overall tax base during the period of the
deferral. Professor Halperin's surrogate taxation compensates for this
omission from the tax base (and the fact that the taxpayers can earn interest
on what would have been paid in taxes). In Professor Halperin's compensa-
tion example, the $100 of compensation will be omitted from the tax base
from the time that it was owed to the employee until the time that it is paid.
The employer, therefore, can earn the interest on what could be viewed as the
government's share of that $100, or $40 (assuming a 40% tax rate.) The
government can reclaim this lost interest by, in effect, taxing the earnings of
the $100 twice, once as it accrues in the employer's hands, and again on
payment to the employee. Thus, the $10 earnings taxed once at 40% leaves

18. Halperin, Assumption, supra note 1, at 679; Halperin, Disguise, supra note 6,
at 519-24. Professor Halperin's preference for characterizing the deduction as a grossed-up
deduction for principal with no deduction for interest allows one to understand his underlying
analysis more clearly. Unless there is a difference between the interest rate that is appropriate
when compensating the payor for the delay in the deduction and the interest rate that is
appropriate when calculating the interest rate charged by the payee, the results are the same
no matter how the deduction is characterized.

[Vol. 3:10



Accowting for the Assumption of Contingent Liabilities

$6, and taxed again leave $3.60. This $3.60, combined with the after-tax
amount available to the employee after payment, leaves the employee with
$63.60, the same amount she would have had available had the compensation
payment not been deferred. 9

In sum, the real distinction between the bank loan and the deferred
compensation cases is not whether the seller/obligor has been allowed to take
the liability into account, but the prior fact that the typical bank lender will
have lent after-tax funds, while the typical employee will be deferring tax on
the deferred payment. In the case of the bank loan, there is no reason not to
allow the liability to be taken into account immediately and allow a deduction
for interest. In the case of the deferred compensation, the seller's treatment
must reflect the fact that the employee has not taken the payment into
income.

IV. THE PREMISES UNDERLYING SURROGATE TAXATION

As the prior analysis reveals, the starting point for Professor
Halperin's analysis is the need to protect the tax base from inappropriate
omissions that can result from discrepancies in the timing of deductions by
payors and inclusion by payees. The importance of these insights (and
Professor Halperin's contributions to their acceptance and implementation
throughout the Code) cannot be overstated. Nevertheless, it is useful to
consider the premises on which the approach relies, and to consider the
circumstances under which it is appropriate to extend principles of surrogate
taxation beyond Professor Halperin's deferred compensation examples.

The first premise is that there is one ascertainable point in time at
which to take into account any value includible in the income tax base and
that, starting with this point in time, it can be assumed the value could have
generated an additional interest-like return that the payee is in fact claiming.
The second is that all values, once created, remain in the system, either in
their original form or as inputs to the production of some other value, and
that they remain available as a source of an additional interest-like return.
Both of these premises are true frequently enough that they appropriately
constitute a starting point for discussions of timing in tax accounting; neither
are so universally true that the possibility of their nonapplicability should be
totally ignored.

19. If the deferral were for two years, the amount left for the employee after current
payment and taxation would have been $67.42 (S63.60 + (S6.36 - S2.54)). To leave the same
value in the hands of taxpayers, the employer must be taxed on the earnings as they accrue
(leaving $106 at the end of the first year and $112.36 at the end of the second), with the entire
amount taxed to the employee on receipt ($112.36 - (SI 12.36 x .40) = $67.42).

1997]



Florida Tax Review

The limits of the first premise-that there is no ambiguity about the
precise time at which a value can be said to exist, and that an interest-like
return is therefore accruing untaxed-can be demonstrated by moving from
questions regarding employee compensation to self-created assets. Suppose
a farmer harvests 400 bales of hay at a time when hay costs $3 a bale. He
stores the hay in his barn, expecting to feed it to his own livestock. 0

Although there has been $1,200 of value added, it is a value that the tax law
does not attempt to include in the tax base.2 ' (This exclusion may be
theoretically wrong, nevertheless, the tax law has never thought the problem
to be so significant as to require the extraordinary efforts necessary to
overcome the administrative problems inherent in trying to include them.)

The farmer will be taxed on the value that he has added through
producing the hay, but only if and when he sells his livestock. Only at that
point will tax accounting acknowledge the value he has created. No existing
tax accounting rules will attempt to make up for the failure to tax the value
the farmer added to the economy but did not commit to the market, even
though several years may have passed while the hay was stored in his barn.
Furthermore, if he feeds the hay to his son's show pony, the value will never
be accounted for.

Suppose instead that a drought hits a neighboring region. The owner
of a racetrack is attempting to secure as much hay as he can, and contacts the
farmer. The farmer agrees to ship 400 bales the next day for $3.50 a bale.
The farmer has $1,400 (400 x $3.50) of income (no later than when he is
paid), and the racetrack has $1,400 of deduction (assuming the hay will be
consumed within a relatively short period of time. 2) The farmer, like
Professor Halperin's employee, should have income at this point in time, and
any arrangements between the farmer and the racetrack owner to lessen the

20. Hay was deliberately chosen as a crop for which the most substantial
contributions to the added value can be assumed to be the farmer's own land and labor, with
relatively few purchased inputs. Under the assumption that there are few purchased inputs,
questions regarding the timing of the deductions for the purchased inputs can be avoided. If
there are substantial purchased inputs, it is likely that they should not be allowed unless and
until the farmer recognizes gain with respect to the hay.

21. The text does not attempt to address the underlying question whether, as a
normative matter, the income tax should tax all accretions to wealth or all accretions to wealth
that are transferred in the market. Since it is clear that the current tax law omits virtually all
wealth unless it is committed to the market, it seems appropriate to use that norm here. The
ambiguity identified in the text, however, exists even assuming this norm. See, e.g., Noel B.
Cunningham & Deborah H. Schenk, How to Tax the House that Jack Built, 43 Tax L. Rev.
447 (1988).

22. Note that if the farmer does in fact divert self-created value from nonmarket to
market use, he should be entitled to take into account his after-tax investment in such value.
There is no norm, however, for determining whether he is entitled to take such costs into
account as of the time the investment was made, or as of the time of the conversion.

[Vol 3:10



Accounting for the Assumption of Contingent Liabilities

tax burden through payment arrangements are properly subject to surrogate
taxation.

Suppose instead that, on learning of the drought, the farmer suspects
that hay will become even dearer than it is now, and he is unwilling to part
with his hay so soon at such a low price. He agrees to ship the 400 bales
now, but insists on being paid an amount that will vary depending upon the
level to which market price of hay eventually rises. That price will not be
determined until the next taxable year. Should the farmer have income at the
time the deal is struck? Probably, for at this point, the farmer is essentially
the same as Professor Halperin's employee. Assuming that this is correct, for
what period, and for what amount will surrogate taxation be appropriate?
Should we attempt to tax the farmer the same as if he had agreed to sell the
hay at the time when it was first baled? It seems unlikely that the later
decision to take the hay to market should affect the treatment of the time that
passed before the decision to go to market, but it would not be incoherent to
do so. 3 Or should we attempt to tax the farmer as if he had agreed to the
spot price at the time he agrees to sell to the owner, and require surrogate
taxation from that point and measured from that value? This seems far more
appropriate and certainly more consistent with current law. The reason is that,
once the farmer is dealing with the owner, we can assume that he is in a
position to charge interest for the delay in payment and that the racetrack
owner's use of the hay will allow him to pay such interest. Until that time,
we could not be sure that the farmer could command any price, much less a
price with interest, for his hay. After that time, the presumption that he could
charge interest seems more appropriate.2'4 Because the farmer and the owner
have control over the terms of their agreement, it may be appropriate to
assume that they have consciously agreed to defer payment, and that
surrogate taxation is therefore appropriate. But the first premise behind
surrogate taxation (that payees can all command an interest-like rate of return
on all values from their initial creation) is likely to be met only after the
farmer and the owner strike a deal.

The second premise vital to surrogate taxation is that all values once
created are never destroyed. These values are instead always used up in the
course of creating other values or consumed in a way consistent with their
remaining in the tax base. Therefore transferors of value can be assumed to

23. In this example, if the later market transaction affected the taxation of earlier
events, the payment, when ultimately made, would be viewed as a lower price, but with a
higher rate of interest charged.

24. The text ignores one troubling complication. Is any future increase in the price
of the hay a separate asset, which, as a normative matter, need not have been included in the
tax base as of the time the deal was struck, or is it part of the value that the farmer is
effectively deferring?

19971



Florida Tax Review

be implicitly charging interest, because those to whom they have transferred
value will in turn put that value to further productive use?2

If the farmer sells to the racetrack owner, the owner will presumably
buy the hay only if he can expect to use it to create other values. The
farmer's labor has produced the value of the hay, and presumably this value
will be used up in the course of the operations of the racetrack. Without
surrogate taxation, the value of the hay could be kept out of the tax base by
deferring payment to the farmer so long as the farmer is willing to accept the
owner's credit. Without surrogate taxation, both parties can benefit from
arranging their affairs to take advantage of this omission. Therefore, in the
ordinary case, the owner should not be able to take into account the cost of
the hay until the farmer has included it in his income.

But what if the weather changes abruptly and an undetected leak in
the racetrack owner's barn renders the hay useless? From the point of view
of the overall tax base, there are no more values that should be included than
there would be if the farmer had never produced the hay. Although he has
not ultimately received anything in return, the owner must still pay the farmer
next year based on hay prices in the interim. 6 To replicate the tax base that
would have been available had the farmer never produced the hay, we would
have to allow the owner to transfer a fund of cash to the farmer without any
net diminution by tax.

There are no current tax accounting norms that inform us which point
of comparison-as if there had never been any hay or as if the hay continued
to have value-is correct. The problem lies in the failure of income tax
theory to define precisely when realization is normatively appropriate. On the
one hand, the government's claim to a share of the earnings from a fund (the
hay) that no longer exists seems tenuous. If realization is appropriate only
when values are sufficiently concrete that they can be presumed to be
commanding an interest-like return, then the never-any-hay baseline is by
analogy most appropriate for the destroyed-hay situation. On the other hand,
if cash had been paid by the owner initially, the government would have
received its share earlier, and the earnings on this share would not have been
contingent on the fund's ability to produce such a return.

25. Owners of self-created assets could also be assumed to be implicitly charging
themselves interest as they await the sale of the second stage of self-created assets. Thus, in
the normal course, the farmer bales the hay under the expectation that his later livestock sales
will provide him adequate compensation for the labor and incidental costs that he incurs while
baling, plus a modest rate of return for the delay in receiving such compensation. His behavior,
however, is equally consistent with being unable to anticipate a price sufficient to justify the
additional costs of bringing the hay to market.

26. Note that to the extent that the owner actually bought the hay as insurance
against the possibility that he would not be able to obtain hay, the treatment suggested by the
text would undertax the overall transaction.

[Vol. 3:10



Accounting for the Assumption of Contingent Liabilities

Any difference between taxation of the farmer and the taxation of the
employee is attributable to reliance on notions of realization. Professor
Halperin's invocation of surrogate taxation is premised on the notion that
values created through waged labor should be realized as the employee
exchanges labor for a promise to pay. In a realization-based income tax, the
norm for the inclusion of gains on property is generally to ignore new value
(as well as increases and decreases in previously included values) until it is
subject to a market transaction, regardless of the circumstances under which
it arose. The value of self-constructed assets, like the farmer's hay while it
remains in the farmer's barn, does not fit easily into either category. The
destruction of the purchased hay brings the norms for wage income (and the
need to tax the farmer currently) into conflict with the norms under a
realization-based income tax for asset income (which tolerate deferrals of
both gains and losses).

V. WHAT IF THE SELLER'S DEFERRAL HAD NOTHING TO DO WITH
SURROGATE TAXATION?

Professor Halperin ably demonstrates that in the most common
situations, allowing a buyer a deduction for that part of a payment attributable
to a contingent liability in effect accelerates the deduction of the liability
compared to the treatment that the seller would enjoy if the business assets
had not changed hands. When such treatment is allowed and the economic
effect is analyzed taking into account the timing of the deductions involved,
the buyer is effectively allowed to deduct the amount as of the time of the
transfer of assets and liability assumption, when ordinarily the seller's
deduction would have been required to await payment. If an assumed
contingent liability represents a value received by the seller and not yet
included in the tax base, such an acceleration is, as Professor Halperin
demonstrates, inappropriate.

Professor Halperin also acknowledges that there may be circum-
stances in which there is no need for surrogate taxation, because the ultimate
payee is not deferring income and there is no value remaining omitted from
the tax base.27 However, he asserts that these can only be situations in which
the seller's deduction is being delayed without warrant. If the seller's
deduction is delayed without warrant, he asserts, the rule requiring the seller's
deferral should be changed, not the treatment of the liability assumption in
connection with a sale of assets. Implicit in his discussion is the notion that
the sale of a business could not justify a change in the timing of the
deduction:

27. HaIperin, Assumption, supra note 1, at 694.

19971



Florida Tax Review

The potential payees of some contingent liabilities may not
be deferring tax on the corresponding income. Without an
income deferral, the payor could possibly be allowed the
equivalent of an interest deduction. However, if, in the
absence of the sale of the business, the seller's deduction for
the item would have been delayed until payment, it effective-
ly does not get an interest deduction in connection with the
payment of contingent liabilities. If this treatment is consid-
ered unwarranted, it should be corrected whether or not a
sale takes place. I see no reason for a different result merely
because the business has been sold. Thus, if the seller would
not get an interest deduction, neither should the buyer. The
entire payment should be nondeductible.28

This comment should give the reader familiar with the notion that a sale is
ordinarily a realization event some pause. After all, sales do make an
enormous difference in a realization-based income tax, for many losses will
be allowed only upon a sale. If the contingent liability in fact represents a net
loss to the tax base (rather than a deferred payment for value added in
connection with the transaction in question), there is no reason that the
overall tax treatment of the liability should not be more favorable as a result
of a sale.

It is not hard to imagine situations in which sellers have previously
unrealized losses triggered by the sale of a business. Suppose the seller had,
in the course of conducting its business, caused the contamination of the soil
on land it owned. (Assume for the moment that there will be no accompany-
ing obligation to clean up the soil, but that the land is clearly worth less than
it would otherwise be, and that it is worth less than its basis in the seller's
hands.) Surely the sale of the land would be an appropriate time to recognize
this loss in value. Indeed, an argument could be made that an allowance of
the loss only on a sale does not provide the seller with an adequate tax
treatment of the lost value. Perhaps a more technically correct treatment
(absent the traditional limitations of realization) would have been to allow the
seller to take this loss in value into account as the seller took into account its
income from the use of the land. Nevertheless, the seller will in fact take this
loss into account on the sale, and only because of the sale.

It seems to me that there are at least two situations in which an
obligation to pay in the future is tantamount to an unrealized loss that could
appropriately be triggered by a sale of a business. Both involve an appropria-
tion by the seller/obligor of value that has unquestionably been destroyed or

28. Id.

[Vol. 3:10



Accounting for the Assumption of Contingent Liabilities

committed to a use that will produce no future value for the seller or buyer
as of the time of the sale. Suppose, for instance, that instead of contaminating
its own soil, the seller/obligor had inadvertently caused the destruction of
some of the value of its neighbor's land. Although the degree of damage has
not been determined yet and the neighbor's ultimate recourse has not yet been
established at the time of the asset transfer, the seller/obligor clearly will be
ultimately liable for some sort of damages at the time of the sale. How should
the neighbor be viewed for tax purposes? Has she merely transferred new
value to the seller/obligor for which she has not yet been compensated, just
as the uncompensated employee had in Professor Halperin's initial analysis?
If so, Professor Halperin's analysis seems appropriate. Or is it possible that
something less susceptible to Professor Halperin's analysis has occurred?

The assumption that the underlying events that gave rise to the
contingent liability produced value, which will be omitted from the tax base
unless the neighbor is taxed directly or through surrogate taxation, simply
may not hold. The neighbor will likely have suffered a decline in value in her
land. This value may be replaced by the later payment by the seller, perhaps
with interest. The transaction can be viewed as essentially a forced loan of
value from the neighbor to the seller/obligor. So long as the value appropriat-
ed by the seller from the neighbor was an after-tax value, there is no obvious
reason not to accord the entire transaction the same treatment given to a more
conventional loan. Under this treatment, the seller/obligor should be accorded
a basis allowance for the use of the loan proceeds at the time they are
extended, and the interest component should be fully deductible. If the
seller/obligor's actions had caused an absolute loss reasonably measured at
$100 of after-tax values owned by the neighbor, and if the seller/obligor has
no future value left to show for these appropriated values, this amount should
be deductible to the seller. When this amount, plus 10% interest, is repaid to
the neighbor a year later, an additional deduction of $10 should be allowed.
Under Professor Halperin's reasoning, so long as the loan is of after-tax
values, the seller/obligor should be allowed an immediate deduction for the
use of the "loan proceeds" (so long as the equivalent purchase would have
been deductible), and for interest later paid.'

Anything less than this would, under Professor Halperin's own
analysis, result in overtaxation. 30 The rules for taxing any delayed payment

by the seller/obligor should ensure that the same economic burden is imposed

29. Id. at 693-94.
30. Again, the astute reader will discern that this overtaxation is the same

overtaxation inherent in realization limitations on taking losses into account. The argument is
essentially that surrogate taxation is not appropriate when the ultimate recipient has unrealized
losses, or indeed, that surrogate realization of losses is no less appropriate in certain cases than
surrogate realization of income.

19971



Florida Tax Review

on the seller in order to achieve the same net effect on the neighbor as if
payment had been made at the time of the loss. If the seller/obligor had paid
the neighbor immediately upon the destruction of the property, the neighbor
would have received $100, totally offset by basis. If this amount were then
invested at 10% with a 40% tax rate, the neighbor would have $106 at the
end of the second year.

But the neighbor's use of basis in the destroyed property will be
delayed when she is only allowed to use it as an offset to the proceeds she
receives in year two. The rules chosen for the seller/obligor's deduction
should therefore in effect allow a surrogate for the delayed use of the
neighbor's basis in the destroyed property. Without a deduction in the first
year, the seller can set aside only $60 of the $100 that it might have paid
initially, which will earn $6 before tax, and the seller will enjoy a tax benefit
of $40 in the second year. The seller will therefore be able to transfer $106
to the neighbor. But the deferred use of the neighbor's basis will shield only
$100 of the payment, and she will be left with only $103.60. If the sell-
er/obligor is allowed a deduction at the time of the destruction, it can set
aside the full $100, and earn $10, which, when paid to the neighbor, will put
the neighbor in the same position as if she had been paid (and allowed to use
her basis) immediately.

Note that the same result can be reached by acknowledging the loss
when it first occurs by giving the neighbor an immediate deduction for her
loss, without regard for the possibility of payment, and deferring the
seller/obligor's deduction. (This treatment is not generally allowed because
of the possibility of reimbursement. 3 ) A deduction of $100 at the time of
the destruction would have produced a tax benefit of $40, which would earn
after-tax interest of $2.40 after a year has passed. Again, without a deduction,
the seller/obligor will only be able to pay $106 in the second year. But the
receipt of $106 without basis offset in the second year would have produced
a net after-tax receipt of $63.60, for a total after-tax position of $106.

The above analysis simply demonstrates that at least some contingent
liabilities may be the result of absolute losses to the tax base, not just
transfers of values new to the tax base, for which no tax accounting has yet
been made. These liabilities arise under circumstances best understood as a
forced borrowing coupled with a destruction of the loan proceeds. When a
liability arises from the destruction of after-tax values, the transaction could
be given a tax treatment no more onerous than that accorded to bank loans.
As Professor Halperin himself has argued, the fact that the liability is
contingent, whether because the seller's liability is not clear as a matter of

31. See, e.g., Regs. § 1.165-1(d)(2); Halliburton Co. v. Commissioner, 946 F.2d 395
(5th Cir. 1991).

[Vol. 3:10



Accounting for the Assunption of Contingent Liabilities

law or because the extent of destruction cannot be ascertained, should not be
determinative of its appropriate treatment. The determinative factor should be
whether the overall transaction involves a deferral of the inclusion of a value
in the overall tax base (as Professor Halperin's compensation entails) or a
deferral of a deduction for a loss to the overall tax base (as in the case of the
contamination of the neighbor's land.)

In other situations, the loss may not be of values already taxed, and
yet a sale may nevertheless be an appropriate time to take a loss into account.
Reconsider the case in which the seller/obligor had contaminated its own soil
and not that of a neighbor. Assume that there is no obligation to replace the
wealth of another, but the seller/obligor has an obligation to clean up the soil
in the future. The seller/obligor's business would clearly be worth less as a
going business as a result of this future obligation. Barring some special
treatment, however, this loss would be deferred for tax purposes until
payment on the obligation was made.32 Would it nevertheless be appropriate
to take this loss into account at the time the business is sold? Is a sale
supposed to make a difference with respect to the aggregate tax treatment of
such unaccrued liabilities?

Existing law provides surprisingly little guidance with respect to this
question, partly because the occasions in which unaccrued liabilities are in
fact sold or otherwise subject to clear-cut realization events, outside of sales
of businesses, are few. (The fanciful nature of the following should serve as
an indication of how unlikely it is that the issue in question will be starkly
raised, outside of the context of traditional debt instruments.) Imagine that A
was a roofing contractor, and after completing various jobs was obligated to
provide for the repair of many of the roofs in her neighborhood. Assume
further that with each year that passes, it becomes more and more likely that
a roof will in fact fail, and a claim on A will be made. B, also a roofing
contractor, was similarly obligated with respect to all of the roofs in his
neighborhood, 500 miles away. Although both have booked all of the income
associated with these jobs, neither uses an accounting method that allows any
reserve for the future work that may be necessary. Suppose further that they
both decided to move, and swap repair obligations as well as neighborhoods.
Has there been a realization event with respect to these liabilities?

There clearly has been a change in legal obligations: entirely different
people now have claims against A and B. How does the possibility of such

32. For further consideration of the appropriate treatment of this sort of future
liability to the seller, see Halperin, Disguise, supra note 6. at 528-31; Donald W. Kiefer, The
Tax Treatment of a "Reverse Investment": An Analysis of the Time Value of Money and the
Appropriate Tax Treatment of Future Costs. 26 Tax Notes 925 (Mar. 4. 1985): Emil M.
Sunley, Observations on the Appropriate Tax Treatment of Future Costs. 22 Tax Notes 719
(Feb. 20, 1984).

19971



Florida Tax Review

a swap triggering realization of gain or loss with respect to these liabilities
mesh with traditional timing rules for taking into account liabilities? Can such
economic burdens be "realized" as the result of a sale before the traditional
all events test has been met, or before the standard for economic performance
under section 461(h) has been met?33 Can they be viewed as "anti-assets,"
the existence of which should be acknowledged at the time of a sale?'

The answer may well be that they should not. Note, however, that if
A and B had been entitled to payments from their former customers as each
year passed in which their roof did not leak, arguably a swapping of such
rights should be treated as a realization event with respect to an asset.35 The
important point here is that Professor Halperin overlooks the possibility that
the transfer of position that requires a future payment could appropriately be
viewed as a realization event with respect to a position that is tantamount to
an "anti-asset."

Perhaps such "anti-assets" could be acknowledged in those limited
situations in which an unpaid loss to one taxpayer definitely will not give rise
to an increase in wealth to another taxpayer. If such "anti-assets" were
acknowledged, there is no reason not to allow the seller a deduction at the
time of the sale. Returning to the contamination of the land example, the only
real point of dispute should be whether this loss should be measured by an
abstract notion of the decline in the market value of the land (and thus

33. The application of § 461(h) in the hypothetical may be problematic: the taxpayer
has performed, and been paid for, the original roof work, but must, if claims are made, provide
or pay for additional services. The transfer of these additional services will have no tax effect
on the recipients, who are treated as merely having received something for which they have
already paid. (Note that this fact distinguishes the hypothetical from the superficially similar
Mooney Aircraft case, in which the warranty involved the payment of cash. Mooney Aircraft,
Inc. v. United States, 420 F.2d 400 (5th Cir. 1970). Most observers would conclude that the
payment of a Mooney bond would be an event that must be taken into account for tax
purposes, since the recipients were to receive cash. It is much less clear whether there would
be agreement regarding what that tax treatment should have been.)

34. This question is related to, but not the same as the question of whether gains
or losses on previously accounted for liabilities should be taken into account. Section
1274(c)(4) suggests that liabilities associated with single assets are not to be taken into account
upon the sale of those assets. This rule will characterize the gains and losses, but will not
otherwise mismeasure income. This treatment can result in mismeasurement of income when
combined with other accounting anomalies. See, e.g., Charlotte Crane, The Effect of Market
Discount and Premium on the Measure of Corporate Gain on Liquidation Under Section 336,
17 J. Corp. Tax'n 119 (1990).

35. If the position were an asset position, the amount taken into income as a result
of the exchange would become basis, and, if later payments received exceeded this amount,
the excess of payments over basis would be income. If the position were an anti-asset position,
the excess amounts over the amount originally deducted should be deductible at future
realization events.

[Vol 3:10



Accounting for the Assumption of Contingent Liabilities

limited by the pre-contamination value of the land) or by the total economic
burden that the future payments represent. (Note that there is no reason to
assume that these two amounts will be equivalent.) Although rules allowing
a deduction for the loss at the time of the contamination might appropriately
take into account whether the seller/obligor had adequate basis in the land
and in other assets that would bear the economic burden of cleanup, there
should be no such problem at the time of the sale, when, presumably, all of
the gains in properties held by the seller will be recognized. As of the time
of the sale, any value against which the claims might be made can be
assumed to be after-tax.

Note that in the contaminated land situation, there is no particular
reason to look to the tax treatment of the ultimate recipient to determine the
appropriate tax treatment of the overall transaction. This as yet unidentified
taxpayer will not be compensated for values "lent" to the seller/obligor at the
time of the contamination. She will, instead, be compensated for values that
she creates and that are essentially new to the tax base as she performs the
cleanup. There is no untaxed principal owed to her when the future payments
can first be predicted, nor is there any untaxed interest in the interval between
the contamination and the cleanup payments.

These cleanup payments are very different from Professor Halperin's
deferred compensation payment, the tax treatment of which must be deferred
until the recipient takes the compensation into income. The seller/obligor is
not denied a deduction for the future payments only because the taxpayer that
will perform the cleanup has not taken the amount into income. The
seller/obligor is denied the deduction for the cleanup payments primarily
because the loss in value represented by the future obligation is too vague to
be taken into account.

Since the only reason the seller was denied a deduction for such
anticipated payments prior to a sale is because the liability simply was too
amorphous to be "realized" prior to a sale, should a sale, the classic moment
at which gains are to be "realized," also be a presumptively appropriate time
for "realizing" losses?36 Surely the problem cannot be that the loss cannot
be quantified, since Professor Halperin himself relies on being able to
quantify the loss at the time of the sale. Professor Halperin's deferred

36. The reader should not lose track of the fact that although the cash sales price
will of course reflect the economic burden of the cleanup expense assumed by the buyer, this
lowered cash amount will not guarantee a tax result equivalent to loss recognition, because
ordinarily the liability for the future payment assumed will be included in the seller's amount
realized. As Professor Halperin's analysis reveals, tax treatment equivalent to recognition of
the seller's loss at the time of sale can be achieved in several ways. either by explicitly
allowing the seller the loss, or by adjusting the buyer's accounting for payments on the
liability. See Halperin, Assumption, supra note 1.

19971



Florida Tax Review

compensation case clearly does not represent such a loss; the employee
presumably added value to the employer's enterprise and his compensation
is in effect a sharing of that value. To allow a deduction for payment without
including an amount in the employee's income would, in effect, keep that
value out of the overall tax base during the period of deferral.

There are two other types of contingent liability for which Professor
Halperin's insistence on taxation of the seller/obligor to make up for the
failure to tax the ultimate payee may not be appropriate. In these situations,
there may be considerable doubt about whether the ultimate recipient should
ever be taxed upon payment. The first type of contingent liability involves
payment of amounts that, for various reasons, under ordinary circumstances
are not included in the income tax base of the recipient. Potential tort
judgments frequently represent values that have already been destroyed, in the
form of the physical integrity of individuals who have been injured but who
have not yet made claims. These ultimate payees do not have basis in the
ordinary sense in their physical integrity. But they do not have to include
amounts in income when they are compensated for an impairment of this
physical integrity by a third party who acted wrongfully. In such cases,
surrogate taxation of the seller/obligor serves only to limit the exclusion that
is otherwise allowed. Although it is highly likely that this limitation is
appropriate (since otherwise taxpayers can arrange for deferred payment in
order to maximize this exclusion), there is no easy answer to the question
whether, as a normative matter, the tax law treats the amounts lost by tort
victims through "forced loans" to tort-feasors as before- or after- tax values.

The second additional type of contingent liability also involves doubt
about the propriety of insisting upon including an amount in the recipient's
income before a deduction can be allowed to a payor. Many contingent
liabilities may in effect be warranty payments, that is, payments made by the
seller/obligor to remedy some shortcoming in the product transferred,
unrelated to any measure of particular value transferred by or injury incurred
by the ultimate recipient.37 Generally, we do not tax recipients of such
payments. Either they are only being paid a return of capital, or they are
enjoying what is in effect a bargain purchase, since the price that they paid
may not have accurately reflected all of the seller's costs. This tax treatment
may not always be correct, since there may be circumstances in which there

37. Frequently such payments are made when there is a discontinuity between the
value paid to the seller at some point in the past and the value that the seller has promised to
provide in return. The seller may have in effect promised its customers that, in the aggregate,
the value of the units it has transferred will approach the average price paid, but that some
customers will have received units worth less, and will deserve some sort of rebate. When
these ultimate recipients are paid, their receipts may be accounted for not as gains, but as
receipts for an amount for which they have already paid.

[VoL 3:10



Accounting for the Assumption of Contingent Liabilities

is an interest component or a component in the nature of liquidated damages
for lost profits built into the terms of the warranty. But even if it is not
technically correct to leave ultimate recipients totally untaxed, there seems to
be no reason to insist on a tax treatment for warranty liabilities transferred in
asset sales that assumes that warranty payments should be taxed to the
recipients when, generally, that is not the case.

VI. WHAT IS AN ASSUMED LIABILITY?

Professor Halperin correctly points out that perhaps the thorniest issue
with respect to the assumption of contingent liabilities is the identification of
those payments that should be associated with assumed liabilities and those
that should be merely regarded as part of the buyer's operating costs after the
purchase.

The answer, it seems to me, is both clear and unworkable. If total
purchase price must equal total fair market value of assets, and if transactions
that include assumed liabilities are to be accounted for in a way that is as
close to a cash purchase as possible, all commitments to make future
payments that reduced the amount of cash a buyer would be willing to pay
must be counted as assumed liabilities. An assumed liability is therefore any
position with a negative net present value, that is, any unavoidable future
payment with a present value greater than the present value of the benefits
to be enjoyed in the future associated with that payment. (This is simply a
more rigorous version of the test proposed by Professor Halperin, that the
distinction between an assumed liability and the buyer's own liability is
"whether the potential liability relates to past or future income.' ' s) Any such
excess burden can be assumed to have been taken into account by the buyer,
and produced a reduction in the cash price paid.

Any commitment to make a future payment can have this characteris-
tic, regardless of its resemblance to a period or current cost. An unfavorable
loan or rental agreement, indeed, an unfavorable wage rate, is an assumed
liability in this sense. If a buyer could have paid $5,000 a month for very
similar rental space, but must take over a lease for $7,000 a month for the
first two years after the purchase of a business, that buyer will pay about
$17,000 less for the assets transferred than he otherwise would have paid.
With each rental payment, the buyer is effectively paying $5,000 of current
rent, and $2,000 of the price for the assets purchased.

The problem, of course, is in associating future payments with
potential future benefits. For most payments, there simply is no way to trace
payment directly to future benefit, even when the amount of the required
future payment is fixed. The payments to be made to Professor Halperin's

38. Halperin, Assumption, supra note 1, at 690.

19971



Florida Tax Review

employee could be merely compensation for prior completed work, or they
could be designed as an incentive for the employee, or as an incentive for
other employees, or they, more probably, could be some combination of the
three. In this respect, the identification of an assumed liability involves
essentially the same calculation as the identification of the component of any
expenditure that must be capitalized.

The price a buyer is willing to pay for a business will depend not just
on what price the individual assembled assets being purchased could
command if sold separately, but what the costs of future operations will be.
The apparent price of assembled assets will be reduced if the buyer
anticipates greater future operations costs than other buyers might; the
apparent price of assembled assets will be increased if the buyer anticipates
lesser future operations costs than other buyers might. It seems grossly
impractical, however, to try to account for all such discounts and premiums
in ways other than they would have been accounted for had the business
remained in the hands of the seller.

It may make more sense, however, to try to identify as assumed
liabilities those commitments to make future payment that are closely tied to
a prior value transferred to the seller under terms that suggest that most of the
value to be received by the owner of the assets has in fact been received
before the sale of the assets. Only if the value to be received after a sale is
negligible should a commitment to make a future payment be considered an
assumed liability.

VII. RECONSIDERING THE ALTERNATIVES FOR ACCOUNTING FOR THE
ASSUMPTION OF CONTINGENT LIABILITIES

Professor Halperin uses as a starting point the proposition that the
sale of assets in which liabilities are assumed should be taxed no differently
than the sale of assets in which cash is the only consideration given. In a cash
sale, all gain is recognized, and a buyer is allowed cost recovery for the
purchase price. Therefore, in order to replicate the treatment of a cash sale,
an amount must be taken into income (or an amount paid with no accompa-
nying deduction) with a present value equal to the expected value of the
liability,39 and cost recovery must be allowed for this amount as of the time
of the sale. (These items will be offsetting, except to the extent that cost
recovery is delayed and therefore provides deductions with a present value
as of the time of the sale lower than the nominally equivalent income
inclusion.) Therefore, in order to replicate the tax treatment of a cash sale, the

39. In a cash sale, this gain is generally thought of as the seller's gain. In the case
of assumed liabilities that have not yet been taken into account by the seller, however, the
seller's gain will be offset (absent character differences) by the allowance of a deduction for
the amount of the liability.

[Vol. 3:10



Accounting for the Assamption of Contingent Liabilities

expected value of a liability must be reasonably ascertainable. Although, as
explored more fully below, there may be many circumstances in which such
a determination of the expected value is difficult, the alternatives which
require such a determination are worth explicating if only to make clearer the
nature of the errors introduced when such a determination is not made.

A. Professor Halperin's Preferred Method

Under the analysis outlined by Professor Halperin, the assumption of
contingent liabilities in connection with asset sales should be accounted for
by allowing the buyer basis, as of the time of the purchase, to the extent of
the expected value of the liabilities assumed. (This same amount would have
been paid in cash by the buyer had the liabilities not been assumed, and must
be included in the basis to be allocated to the assets purchased in order to
assure that the appropriate cost recovery is allowed.) If the liability is
resolved for an amount equal (in present value terms) to this expected value,
then no further adjustments need to be made.

If, however, the liability is resolved in a way that is unfavorable to
the buyer, the buyer should be allowed an additional deduction. This
deduction, however, should be allowed only to the extent that the resolution
value exceeds the expected value, adjusted for the passage of time. On the
other hand, if the resolution value is less than the expected value, adjusted for
the passage of time, the buyer should recognize income as of the time that the
contingencies are resolved.

Year 1 Year 2 Year 3 Year 4 Present
Value of
cost re-
covery as
of Year I

Payment as Cost recovery 33.33 44.45 14.81 7.4 94.6
expected: allowed
106 in year 2

Payment Cost recovery 33.33 44.45 14.81 7.4 94.6
more than allowed
expected:
200 in year 2 Loss allowed 94 88.6

200-106--94 183.2

Payment Cost recovery 33.33 44.45 14.81 7.4 94.6
less than allowed
expected: 56 52.8
50 in year 2 Gain recognized 41.8

50-106=56

Halperin's Preferred Method: Immediate Basis Inclusion with Limited Adjustment
for Resolution

19971



Florida Tax Review

This method works reasonably well when it is anticipated that there
will be a definite time at which the contingency will be resolved. To the
extent that there is no such time, and therefore possible resolution gains are
not taken into account, the buyer will have been allowed cost recovery on a
price that she has never paid. It also works effectively when there is a high
level of confidence about the ability to identify the payments associated with
assumed liabilities. If payments are not so identified, excess deductions may
be claimed, as payments are misclassified as current costs as they are paid.

To avoid this result, Professor Halperin suggests relaxing the
standards for connecting payments with particular assumed contingent
liabilities. A buyer would be allowed to deduct and add to basis any payment
that would be associated with such assumed liabilities, but "on the dates
projected for the expected payments [as of the time of the sale] the buyer
could be required to forgo deductions equal to the anticipated payments, or
to include an equivalent amount in income, regardless of whether actual
events are consistent with the estimate. 4 °

B. First Alternative Method: Accounting for Price Adjustments as Amounts
are Paid

The problem associated with cost recovery for costs never incurred
could also be solved by postponing the allowance of cost recovery under
Professor Halperin's preferred alternative method, that is, by deferring the
allowance of cost recovery until payments are actually allowed on the
liability. Cost recovery would be allowed for the amounts actually paid, when
paid, to the extent of the expected value of the liability adjusted for the
passage of time.41 A new recovery period would begin with each payment
on the contingent liability. To the extent that amounts are paid in excess of
this expected value, adjusted for the passage of time, deductions would be
allowed.

40. Halperin, Assumption, supra note 1, at 694.
41. Id. at 697-700.

[Vol 3:10



Accounting for the Assumption of Contingent Liabilities

Year Year Year Year Year Present
1 2 3 4 5 Value

as of
Year 0

Payment as Cost recovery 35.32 47.11 15.69 7.84 94.6
expected: allowed
106 in year 2

Payment more Cost recovery 35.32 47.11 15.69 7.84 94.6
than expected: allowed
200 in year 2

loss allowed 88.6
200-106=94 94 183.2

Payment less Cost recovery 35.32 47.11 15.69 7.84 94.6
than expected: allowed (on
50 in year 2 amount actual-

ly paid, and
amount taken
into income)

gain recognized 56 52.8
50-106=56 41.8

Halperin's Preferred Alternative Method: Deferred Basis Inclusion with Adjustment for
Resolution

Evaluating the results of this method in the event that the liability is
resolved in the buyer's favor is more problematic. If the liability is resolved
in the buyer's favor, the buyer has in effect enjoyed a bargain purchase. It is
unclear whether, in terms of realization, this bargain element should be taken
into account by the buyer. In order to replicate the results of a cash sale,
however, this bargain element must be taken into account. If the bargain
element is taken into account, cost recovery on this amount should be
allowed, a complicating detail which Professor Halperin overlooks. In sum,
the alternative preferred method cannot mimic the tax results of a cash sale
unless, when the contingency is resolved in the buyer's favor, two abstract
adjustments are made, one including the favorable resolution in income, and
the other allowing cost recovery on this newly included amount.4

42. The current treatment allowed in the case of sales to which § 338 (outlined in
Temp. Regs. § 1.338(b)-3) applies, resembles this treatment; but because (1) it does not clearly
allow for an increment above the original price resulting from the passage of time, and (2) it
does not allow a deduction for resolutions more burdensome than anticipated, or for gain for
resolutions less burdensome, it will produce results equivalent to a cash sale only when the
burden of the liability is correctly predicted.

19971



Florida Tax Review

C. Professor Halperin's Second Alternative Method

Both the preferred method and the first alternative approach require
relatively complex adjustments throughout the period during which the
contingent liability remains outstanding. The preferred method requires
adjustments only at the time the contingency is resolved. This method will
produce no error in the total definition of the tax base when the contingency
is resolved against the buyer, for the buyer will be allowed a deduction that
(assuming appropriate discount rates are used) reflects the economic burden
the liability ultimately borne. It will, however, produce a relatively large error
when the contingency is resolved in favor of the buyer, since the buyer has
been given basis credit for an amount that has not and may never be paid.
Only if some arbitrary resolution mechanism is introduced will this basis
credit ever be undone. The first alternative method requires the use of abstract
cost recovery, based on the actual characteristics of the assets purchased in
the sale, but with allowances unrelated to the presence of these assets as of
the time cost recovery begins. If the basis for which this abstract cost
recovery is allowed would have been allowed other than through cost
recovery allowances (for instance, because the assets have been resold)
additional abstract adjustments would be necessary.

These complex adjustments could be avoided by using a second
alternative method.43 This method would in effect assume in advance that
the contingency is resolved without the buyer making any payment at all, but
nevertheless allow the buyer cost recovery for the fair market value of the
assets acquired. Under this method, the buyer takes into income the expected
value of the assumed liability at the time of the acquisition and is allowed a
deduction for all amounts subsequently paid.

43. This method is suggested by Professor Halperin as a variant of his preferred
method. Halperin, Assumption, supra note 1, at 696-97.

[Vol 3:10



Accountiig for the Assumption of Contingent Liabilities

Year 1 Year 2 Year 3 Year 4 Present
Value as
of Year 0

Payment as Cost recovery 33.33 44.45 14.81 7.4 94.6
expected: allowed
106 in year 2

Upfront income 100 IJO
inclusion

Deduction for 106 100
payment 94.6

Payment more Cost recovery 33.33 44.45 14.81 7.4 94.6
than expected: allowed
200 in year 2

Upfront income 100 100
inclusion

Deduction for 200 188.6
payment 183.2

14.8133.33Cost recovery
allowed

Upfront income
inclusion

Deduction

Second Alternative Method: Buyer's Income Inclusion and Deduction of ll Other Amounts

This second alternative method has several distinct advantages: first,
it comports with a recharacterization of the liability assumption in which the
seller pays the buyer cash for the liability assumption (giving the seller sale
treatment and the buyer income) and then the buyer uses this cash to
purchase assets from the seller (giving the seller an amount realized and the
buyer basis in the assets so acquired). Because this method merely reflects a
recharacterization of the sale transaction, it could be adopted without enabling
legislation. Finally, as Professor Halperin acknowledges, this method
minimizes the distortion involved in choosing discount rates.

VIII. WHAT IF THE EXPECTED VALUE CANNOT BE ACCURATELY

DETERMINED?

Although assumed liabilities may be identifiable, the fact that they are
identifiable does not mean that they are quantifiable. For instance, a seasoned
buyer of smaller mining operations may be able to predict, in the aggregate,
the upside and downside risks of its purchases. A certain number of
properties will have more remaining reserves than predicted; a certain number

Payment less
than expected:
50 in year 2

44.45

47.1
41.7

19971



Florida Tax Review

will have less. Similarly, a certain number will have more environmental
problems than predicted, but others will have less. The agreement between
a buyer and a seller regarding the sum of these possibilities does not indicate
that they would have agreed to anything about the burdens or benefits
associated with any particular risk involved in the purchase. Indeed, the buyer
may be taking into account the overall risks involved in a series of purchases
separately negotiated with a series of sellers. Assuming the buyer assigns a
particular value to a particular risk in one of a series of asset purchases may
be entirely unrealistic. The buyer may instead, simply by reviewing the way
in which each business has been conducted in the past and using his overall
knowledge of the industry, make aggregate assumptions about the liabilities
he will find that he has assumed.

Two approaches to accounting for contingent liabilities are available
if one assumes that the expected value of these liabilities cannot be
ascertained either as of the time of sale or in retrospect. The first would
require capitalization of all amounts paid with respect to assumed contingent
liabilities as if all amounts paid were in effect part of the purchase price for
assets.44 This approach clearly imposes a greater tax burden than would be
required had cash been paid. But with the introduction of section 197 and the
general availability of fifteen year amortization for all intangibles, the burden
of this extra tax has been substantially reduced.45

The other alternative would allow current deductions for all payments
made with respect to assumed current liabilities.4 6 As Professor Halperin's

44. Current law seems clearly to require this result, although it is ambiguous
whether the capitalization should be as of the time of payment using a new recovery period,
or as of the time of the sale, with or without an adjustment for the passage of time. Case law
has simply required the capitalization of all amounts paid. See, e.g., David R. Webb Co. v.
Commissioner, 77 T.C. 1134 (1981), aff'd 708 F.2d 1254 (7th Cir. 1983); Pacific Transport
Co. v. Commissioner, 483 F.2d 209 (9th Cir. 1973), rev'g per curiam, 29 T.C. Memo. (CCH)
133, 39 T.C. Memo. (P-H) 70,040 (1970), cert. denied, 415 U.S. 948 (1974) and reh'g
denied, 416 U.S. 952 (1974). The regulations under § 338 provide a variant on this approach,
apparently allowing basis when the liability is fixed (rather than when paid).

Thus, current authority suggests directly tying all payments to assets acquired, an
approach that clearly will overstate the value of assets acquired when contingencies are
resolved unfavorably to the buyer.

45. Assuming a 6% after-tax interest rate, a payment technically entitled to an
immediate deduction is reduced by 15-year amortization to a series of future deductions worth
about 64% of the value of an immediate deduction. Without amortization under § 197, the
denial of an immediate deduction would have produced basis only available on a later taxable
sale of assets.

46. This treatment has been proposed, among other places, by the New York State
Bar Association Tax Section. Richard L. Reinhold, Report on the Federal Income Tax
Treatment of Contingent Liabilities in Taxable Asset Acquisition Transactions, 49 Tax Notes
883 (Nov. 19, 1990).

[VOL. 3:10



Accounting for the Assuinption of Contingent Liabilities

analysis reveals, this treatment would produce the same result as allowing a
deduction at the time of the sale for the expected value of the liability. (It
would also provide an automatic adjustment for the resolution value, to the
extent that the amount actually deducted varied from this expected value.) To
the extent that the contingent liability related to arrangements, like deferred
compensation, that involve values that will be omitted from the tax base
unless surrogate taxation is used, the overall transaction will be undertaxed
under this alternative. The benefit to the taxpayer of accelerating the
deduction as an offset to income for the amount paid on the liability would
depend upon the cost recovery that would have been allowed on the assets
acquired had the expected value been known and properly accounted for.47

To the extent, however, that the contingent liability relates to arrangements
in which surrogate taxation is not appropriate, the degree of undertaxation
will be less. (There may be a lesser tax burden as a result of the sale of the
assets than there would have been without the sale, as will generally be the
case when losses recognized on assets as a result of the sale are greater than
gains so recognized.)

From a practical standpoint a comparison of the alternatives for
accounting for assumed contingent liabilities must take into account the
effects of the likely errors involved in the implementation of these alterna-
tives. Under all methods, there will be an advantage to understating, if not
ignoring completely, the expected value of the contingent liability, at least if
one assumes that payments that are not treated as payments on the expected
value of the liability are entitled to current deduction. This advantage will
always be the difference between the value of current deduction for the
expected value of the liability, and the value of the cost recovery that would
have been available had the expected value of the liability been assigned a
proper cost recovery. Indeed, the understatement of a contingent liability and
the ignoring of the fact of the liability assumption involve exactly the same
potential error: avoiding the drag on deductions involved in proper capitaliza-
tion and cost recovery.

IX. SHOULD THE ACCOUNTING FOR A CASH PURCHASE BE THE
ONLY NORM FOR EVALUATING POSSIBLE ACCOUNTING TREATMENTS?

Professor Halperin's analysis began with the proposition that all
transactions in which liabilities are assumed should receive the same tax

47. A current deduction with a present value of I has been allowed instead of cost
recovery with a present value of .64 if the acquired assets were primarily intangibles for which
15-year amortization is available under § 197; about .44 if the acquired assets were primarily
commercial real estate; of undeterminable value if the assets were primarily unimproved real
estate for which cost recovery would be allowed only upon sale.

1997]



Florida Tax Review

treatment as transactions in which only cash is given as consideration. This
starting point assumes that it is possible to identify a cash price that
represents the fair market value of the assets in question, without regard for
the tax treatment anticipated in connection with the future use of the assets.
This may be an ill-conceived notion. The price of virtually any asset will
reflect the tax treatment of the ongoing business anticipated by the buyer.
Although it may be possible to assign to any individual asset a cash price and
a technically correct tax accounting treatment of that cost, it seems unlikely
that this will be possible (and certainly not worthwhile) for sales of business
assets. Each such sale will have associated with it commitments to make
future payments-at both favorable and unfavorable prices. Each such
commitment will produce a deviation from the cash price for the individual
assets, and will increase the likelihood that no technically correct tax
accounting will be possible. To the extent that the most likely tax accounting
will be favorable to the buyer, the buyer may be willing to pay more than
otherwise; to the extent that the most likely tax accounting will be unfavor-
able, the buyer will be willing to pay less than otherwise.

Reproducing the tax treatment of all cash sales therefore may not be
the only criterion for evaluating the accounting treatment afforded assumed
contingent liabilities. Indeed, to the extent that establishing the technically
correct treatment in every case is futile, the possible accounting treatments
should be evaluated primarily according to whether the consequences of
getting the accounting wrong are tolerable.

At least two factors determine whether tax accounting errors will be
tolerable. The first is the degree to which taxpayers can take advantage of
these errors, and the second is the range of disparity in the ability of
taxpayers to take advantage of these errors. If a taxpayer truly has no control
over its ability to take advantage of a tax accounting error, then (although the
error may create a windfall) the consequences of the error are likely to be
tolerable. Similarly, virtually any tax accounting error would be tolerable if
every taxpayer could be assumed to be equally able to take advantage of (or
to be burdened by) the accounting error, so long as a limit exists that
prevents arbitraging to such an extent that no tax base remained.

Most tax accounting errors, however, fall somewhere in the middle.
Some taxpayers are better able to take advantage of favorable tax accounting
errors than others, but not to so great an extent that the benefit of the error
is entirely capitalized in the asset involved. Some taxpayers are disadvantaged
to a sufficient degree that the activities in which they are involved bear a
heavier tax, and perhaps attract less capital than they otherwise would.

Choosing among technically imperfect tax accounting rules requires
predicting which effects are most likely, and which of these effects are most
tolerable. In choosing the tax treatment of assumed contingent liabilities,
predictions need to be made about the taxpayers most likely to be affected.

[Vol. 3:10



Accounting for the Assumption of Contingent Liabilities

Under all of the technically correct methods, taxpayers can obtain real
tax advantages by characterizing portions of the purchase price as contingent
liabilities, and then underestimating the expected value of that liability.48

Given this reality, it may be more appropriate to simply adopt an accounting
treatment that is less favorable than cash to avoid the possibility for abuse of
a more favorable method.

On the other hand, some industries may be able to avoid assumptions
of contingent liabilities in connection with asset transfers because insurance
is more readily available for the risks inherent in that industry. 9 If so, there
may be identifiable classes of taxpayers that are put at a disadvantage because
they may be unable to avoid the consequences of assumed contingent liabili-
ties and the onerous accounting methods that they require. Nevertheless, we
might be willing to tolerate the error if the effects are felt uniformly through-
out an industry and there is little possibility that asset transfers within that
industry will be burdened more than asset transfers in other industries. If a
particular industry is intolerably affected, and especially if a substantial
portion of the liabilities faced by the industry reflect destroyed values rather
than merely transferred values, then selective use of more favorable
accounting treatments may be appropriate.

X. CONCLUSION

Professor Halperin's analysis of the appropriate tax accounting for
contingent liabilities is correct to the extent that all contingent liabilities
represent obligations to pay for values that have not yet been included in the
tax base.

His conclusions are based on the assumption that all contingent
liabilities should be treated similarly for the purposes of tax accounting, and
that none should be given as favorable tax accounting treatment as fixed
liabilities. It is possible, however, that at least some contingent liabilities are
in effect borrowings of after-tax values, and thus have more in common with
traditional bank borrowings than with deferred compensation.

Thus, to the extent that contingent liabilities represent either (1)
obligations to repay for values that were previously included in the tax base
but for which an income offset has been deferred or (2) obligations to repay

48. An advantage could also be gained from underestimating the discount rate used
to compare actual payments with the estimate of expected value. With such an underestimate,
fewer favorable resolutions would be identified as such and more resolutions would be
assumed to be unfavorable.

49. If insurance is available, the overall taxation of the asset transfer will be less
than that urged by Professor Halperin if, and to the extent that, tax accounting for insurance
effectively allows an immediate deduction for a reserve against the liability.

1997]



646 Florida Tax Review [Vol. 3:10

for values that are not ordinarily included in the tax base, Professor
Halperin's proposed treatment may lead to overtaxation. (The degree of
overtaxation involved will depend upon clarification of the as-yet undevel-
oped norm for realization of gains and losses with respect to liabilities
transferred in connection with sales of assets.)

Although it may be possible for some taxpayers to establish that the
liabilities related to their businesses warrant more favorable tax treatment than
that proposed by Professor Halperin, this possibility alone is not enough to
warrant more favorable treatment across the board than that advocated by
Professor Halperin. Indeed, good arguments can be made that the appropriate
treatment of assumed contingent liabilities should be more, rather than less,
onerous than the most technically correct method, especially if all taxpayers
are equally able to arrange the terms of the sale of assets in such a manner
so as to avoid this more onerous treatment.