Issues regarding pre-award interest arise in most cases where damages are awarded. Typically tribunals assess damages as of a date in the past, and interest is applied to this amount up to the date of the award. Because the entire arbitration process (from the date of damages, to filing, through to an award) can take a long time to complete, and the rate of pre-judgment interest may be quite high, interest can make up a substantial proportion of total damages. For example, in case ARB/11/26 at the International Centre for the Settlement of International Disputes (ICSID), principal damages totalled US$87.3 million and pre-judgment interest totalled US$85.5 million, close to 50 per cent of total damages (principal + interest). In case ARB(AF)/99/1, principal damages totalled 9.5 million Mexican pesos and pre-judgment interest totalled 7.5 million pesos – 44 per cent of the total.
Interest in financial markets
Interest is used in finance to bring amounts of money forward in time to find equivalent amounts of money at a specified future date. Interest rates may also be used for discounting; that is, to bring amounts of money backward in time. Positive interest rates reflect the ‘time value of money’ – lenders require an inducement to lend money because it requires them to postpone their own consumption or their business use of the money, while borrowers are willing to pay this because they have immediate consumption needs or because they can use the money now for their own business purposes. Together, these economic considerations as they apply to actual and potential lenders and borrowers determine the ‘time value of money.’ Notice that governments and their agencies are among the actual and potential lenders and borrowers.
Apart from the pure time value of money there are a number of additional factors that may determine market interest rates. These include a premium in case the borrower defaults, a premium for liquidity or an adjustment reflecting the tax treatment of the loan.
Interest rates are normally greater than zero. They change over time just like any other market price (and default and liquidity premiums change too). Since the financial crisis of 2008, interest rates have been very low and some interest rates have even been negative. For example, in mid-2016, yields on many German, Swiss and UK government bonds were negative, although US Treasury bond yields were still positive. Even high quality euro corporate bond yields had turned negative. Many economists believe that these very low interest rates largely reflect government and central bank policy in many countries in response to the financial crisis of 2008/9 and its aftermath, as well as reflecting underlying fundamental economic conditions such as poor prospects for growth and productivity that lead to poor opportunities for investment.
Interest rates are used to bring amounts of money backward and forward in time. Arbitration cases typically involve amounts of money at many different times, and these are brought forwards and backwards using different rates such as the cost of capital and the pre-award interest rate (the differences between cost of capital and pre-award interest rate and the reasons why they are different will be explained below).
Suppose a claimant owns or owned an asset that, the tribunal decides, is rendered less valuable by an act of the respondent (the ‘bad act’). We assess how much money the claimant would have expected to receive if the claimant had not taken this act. Since the claimant did take the act, this is a hypothetical situation that is referred to as the ‘but-for world’; in other words, the world that would have been expected to prevail but for the respondent’s bad act. The difference between the money the claimant actually received and the money the claimant would have received in the but-for world is the basis for damages.
However, both in reality and in the but-for world, these amounts of money occur at different dates, and we need to allow for this. Suppose the bad act occurred at date 0. The asset yields, or would have yielded, cash flows at dates 0, 1, 2, … n. The tribunal’s award takes place at date t, which could be before or after date n. Typically, the damages figure is assessed as of the date of the bad act by bringing the cash flows from the asset back to date 0 (this is known as the ‘present value’ as of date 0). The difference in asset value at date 0 between the actual and but-for worlds is damages as of that date. This amount is then brought forward in time using the pre-award interest rate. Further interest (post-award interest) is then added until the payment is actually made, which could be considerably later.
The rate used to bring actual or expected (but-for) cash flows backward in time is typically the cost of capital for the asset. This rate normally includes a risk premium reflecting the fact that the asset is a risky endeavour with uncertain cash flows. The rate used to bring the asset value forward in time to the award date, however, is a different rate – often a riskless rate – as I will discuss below.
Tribunals often award post-award interest in addition to the amount specified in the award. Post-award interest is added until the award is actually paid. The starting point for post-award interest could be the date of the award, or it could be a later date. For example, the tribunal might specify that the award should be paid by a date within a few months of the award (without the addition of post-award interest), but that any further delay would attract post-award interest from that date.
Post-award interest is often applied at the same rate as pre-award interest. This reflects the view that interest is just compensating for the time value of money. An alternative view is that post-award interest should be higher that pre-award interest to discourage late payment. If the respondent has to pay a significant default premium on its borrowing, any post-award interest rate that is lower than its cost of borrowing, such as the risk-free rate, can create an incentive to delay payment.
In the rest of this chapter, I will refer mostly to pre-award interest, but most of the material will apply equally to post-award interest.
Different interest rates that may be used for pre-award interest
A number of different interest rates might be mentioned in the context of pre-award interest. Here is a guide to the most common ones.
The claimant and respondent might be bound by a contract, an alleged breach of which gave rise to the dispute. The contract might specify an interest rate to be used in case of late payment. Many arbitrations concern claimed breaches of bilateral investment treaties (BITs), which sometimes specify interest rates.
Benchmark rates (or reference rates)
Many financial contracts use interest rates that are expressed relative to a benchmark rate. Often the interest rate in the contract is not equal to the benchmark rate but is equal to the benchmark plus a spread. For example, the rate on a floating rate security or a swap could be Libor plus 50 basis points (a basis point is 0.01 of a percentage point, so 50 basis points is 0.5 per cent). Common benchmarks are Libor, Eonia, Euribor and US prime. Sometimes, the benchmark rate is not an interest rate; for example, the benchmark could be inflation as measured by the increase in the consumer price index (CPI), and the contract might specify interest at inflation plus 1 per cent.
These are rates at which banks borrow and lend to each other. These loans can be of various maturities but the majority of interbank loans are overnight. They are unsecured. In the US, the interbank rate is the federal funds rate. In the eurozone it is Eonia (for overnight loans) or Euribor (for longer maturities). There are various other measures of interbank rates that vary according to the currency of the loan, the maturity of the loan, the place where the banks are based, and the method of collecting the data (survey responses or actual transactions). These include Libor, Sonia, USD Libor and Saron. Not all banks can borrow at these rates, but in normal times the major banks can. The federal funds rate is different from the other measures of interbank lending rates in that the US Federal Reserve directly targets this rate, while the other measures are only indirectly influenced by central bank actions.
The risk-free rate
Loans to some governments are generally considered risk-free. For example, the US government is considered certain to repay its US dollar borrowing and the German government is considered certain to repay euro borrowing (notice that the US government can in any case create more US currency to repay such loans, but the German government cannot create euros at will). Thus, the yields on securities issued by those governments are used as a measure of the risk-free interest rate. Finance textbooks often refer to ‘the’ risk-free rate, but there are different rates for different currencies and maturities. As explained below, different currencies naturally have their own different rates. Also, government securities with different maturities have different yields, so the ‘risk-free rate’ needs to specify which maturity is being referred to. Government securities with maturities of less than one year are known as treasury bills (T bills). Government securities with maturities of more than one year are known as government bonds or (in the US) as treasury bonds or treasury notes.
Although not risk-free, interbank rates have at times been very close to risk-free rates. The difference between US dollar T bill yields and US dollar LIBOR is known as the TED spread. In normal times the TED spread is well under 1 per cent, typically around 0.25 per cent to 0.5 per cent, although in times of crisis it is considerably higher (over 1 per cent from summer 2007 to early 2009, peaking at over 4 per cent for a short while in October 2008). Since the TED spread is normally less than 0.5 per cent, tribunals might use interbank rates for pre-award interest on the basis that they represent an appropriate yield with very little premium for risk.
The cost of debt of the respondent
This is the cost at which the respondent could borrow. For example, if the respondent is a sovereign state, it is the yield on the bonds of that state. Typically, the state would borrow in a currency it does not issue, such as US dollars, so this would be higher than the risk-free rate. The cost of debt of the respondent is often used to determine pre-award interest.
The cost of capital
The cost of capital is often used in arbitrations, sometimes in connection with pre-award interest. This could refer to the cost of capital of the project at issue in the arbitration, or the cost of capital for the claimant more generally. Cost of capital means a rate that reflects risk and combines the cost of borrowing with a required return on equity. The most common notion of cost of capital is the weighted average cost of capital (WACC). A company’s WACC is a weighted average of its cost of debt and its cost of equity (assuming the company issues only debt and equity; if other financing such as preferred stock is used, this could be factored into the calculation). The weights correspond to the share of the amount of debt and the amount of equity in the total value of the company’s assets. An adjustment is made for corporation tax, reflecting the deductibility of interest payments or the optimum shares for corporation tax. For a company whose assets are all of similar risk, the cost of capital for the company as a whole is similar to the cost of capital for the project. If the project has a different risk to the company’s other assets, the cost of capital for the project, in principle, requires adjustment, although this refinement is often considered impractical.
Bank deposit interest rates
These are rates offered by banks for deposits. They vary by maturity and currency. For example, if a depositor does not have the right to withdraw within a year, the rate is normally higher than for a deposit that can be withdrawn without notice.
Fixed and floating rates
Some loan contracts are at rates of interest that are fixed over the life of the loan, while others make provision for the interest rate to change. When a loan is at a floating rate, the contract specifies a reference interest rate and a spread relative to the reference. For example, US Treasury bonds are at fixed rates. In the US, many residential mortgages are at fixed rates, while others are at floating rates tied to LIBOR. In other countries, mortgages are typically at floating rates but may be fixed for a period such as five years before reverting to floating rates.
Pre-award interest is often awarded at a risk-free rate based on T bills, or at an overnight interbank rate, or at such a rate plus a spread. Since these short-term yields change daily, this implies that the pre-award interest rate also changes, and is effectively a floating rate.
Real and nominal rates
Interest rates most commonly specify the interest payment in an amount of currency (US dollar, euro, etc.). These are called nominal interest rates. Somebody who lends money in this way can expect to receive a predictable amount of money, but does not know the extent to which inflation will erode the purchasing power of that money. Therefore, some loans specify interest payments that are indexed to a price index such as the CPI. For example, the US government has issued ‘treasury inflation protected securities’ (TIPS). The return on such a loan, expressed in units of currency, consists of the inflation rate plus a real return (strictly speaking, the mathematical relationship between nominal rates, inflation and real rates also includes a ‘compounding’ term, which is usually small and frequently ignored). Unless otherwise specified, rates are normally understood as nominal rates.
The problem of inflation uncertainty for loans expressed in nominal terms is more severe when a loan contract is longer term. For a very short-term loan, such as a T bill, indexing the payment to the CPI would serve little purpose because inflation is unlikely to change much over the course of the loan. So short-term (nominal) rates effectively offer inflation protection anyway.
Real interest rates and real discount rates arise quite frequently in the context of arbitrations. For example, the dispute may concern long-term contracts that are specified in real terms. However, tribunals normally specify pre-award interest in nominal terms. It is never or hardly ever specified in real terms. One reason why this makes sense is that pre-award interest is often awarded at the T-bill rate, at another short-term rate, or at a short-term rate plus a spread, and for short-term rates there is no need for indexing.
Compound and simple interest
With any kind of loan contract, interest is specified as a percentage of the amount borrowed (the principal). With simple interest, the interest is a percentage of the original amount. With compound interest, for each period the interest is added to the amount of the loan and then interest is calculated as a percentage of this new amount. For example, the balance on a loan of US$100 at 10 per cent simple interest accumulates to US$120 after two years with simple interest. Compounded annually, it accumulates to US$121 after two years (after one year, principal plus interest totals 110, and the second year’s interest is 10 per cent of that amount, i.e., 11).
Interest can be compounded at different intervals of time. Compound interest originated around four millennia ago in Babylon. Interest on silver loans was set at 20 per cent; loans were usually for durations under one year with no compounding. However, loans of several years’ duration sometimes incurred compound interest once the accrued interest equalled the principal. At 20 per cent annual interest this implies a compounding interval of five years. To repay a loan of 100 mina would have cost 180 mina after four years, 200 mina after five years and 240 mina after six years (because in the last year the 20 per cent interest rate would have been applied to the new balance of 200 mina, while up until then it would have been applied to the original balance). Nowadays compounding intervals tend to be one year or less. For example, a loan at 10 per cent annual interest with quarterly compounding actually means (although the compounding convention may vary) that every quarter, 2.5 per cent interest is added to the previous quarter’s balance (since 2.5 per cent is one-quarter of 10 per cent). Although the difference made by the compounding interval is smaller when comparing compounding intervals of less than one year, it is worth checking the compounding conventions when using data on interest rates. For consumer loans in the US, the APR (annual percentage rate) does not reflect the true economic cost because it ignores the effect of compounding, while the APY (annual percentage yield) allows for compounding. In other countries, consumer legislation often requires similar disclosure but under a different name to APY.
At any point in time, different fixed income securities with the same credit quality (and any other relevant features such as tax status) but different compounding intervals will trade at the same yield. The yield on a security is a measure of return that makes appropriate allowance for compounding frequency.
Tribunals sometimes award simple interest and sometimes award compound interest. The choice may be determined by legal considerations. For example, legal principles may require that a legal expropriation requires simple interest, while an illegal expropriation requires compound interest. It is not the purpose of this article to offer any guidance on appropriate legal considerations.
However, regardless of such legal requirements, in most situations economists favour compound interest for pre-award interest. The reason is that yields (which are based on compounding) are the economic benchmark for market interest rates. While financial contracts with simple interest do exist and are straightforward to value, the market values of such contracts are ‘reverse-engineered’ to make sure that they trade at market yields. Applying a market yield without compounding would not make the claimant whole.
Interest rates in different currencies
Different currencies have different interest rates. It is important not to take an interest rate that refers to loans in one currency and apply it to amounts of money in another currency.
Arbitrations frequently concern international disputes. In a typical case, the claimant makes an investment in an asset in another country. Different parts of the damages calculation might involve the currency of the country where the investment is made, as well as the investor’s home currency or an internationally used currency, such as the US dollar. Issues arise concerning the interest rates or cost of capital to be used for amounts denominated in the different currencies. There is standard way to handle such issues, known as the principle of uncovered interest parity (UIP). UIP requires that expected exchange rate at future dates depreciates or appreciates to offset the differential in interest rates. For example, if interest rates in US dollars are 1 per cent and in Swiss francs are 0.25 per cent, then the US dollar should be expected to depreciate against the Swiss franc as a rate of (approximately) 0.75 per cent annually. If UIP is applied, calculations carried out in either currency will reach the same result.
However, while cross-currency considerations very often arise in damages calculations, they arise less often in relation to that part of the calculation that relates to pre-award interest. Typically, damages at the time of the bad act are assessed in one currency, they will be paid in that currency, and the pre-judgment interest rate used to bring this amount up to the date of the award (or the date of payment) is derived from market interest rates on that currency’s securities. Cross-currency considerations can arise in some situations, however. For example, if the tribunal decides to use the respondent’s borrowing cost for pre-award interest to be applied to a US dollar damages figure, it needs to check this is an interest rate applicable to US dollar borrowing or to convert it to US dollars using UIP.
Which interest rate is preferred on economic grounds?
There are two theories that are generally offered on economic grounds to determine an appropriate interest rate for pre-award interest.
One theory supports the use of the risk-free rate of interest. The argument is that once the damages sum has been determined to compensate the claimant at the date of the bad act, the purpose of adding interest is merely to bring that amount forward in time. The claimant is not bearing any risk because the amount of damages, from then on, does not rise or fall in line with economic circumstances, unlike the return on risky assets. It is a fixed amount. By contrast, the stock price of an oil company or the value of an oil field are examples of risky assets whose returns depend on a host of economic factors such as oil prices, economic growth, inflation, management skill or geological risk. For example, the economists Franklin Fisher and Craig Romaine argue that:
[I]n depriving the plaintiff of an asset worth Y at time 0, the defendant also relieved it of the risks associated with investment in that asset. The plaintiff is thus entitled to interest compensating it for the time value of money, but it is not also entitled to compensation for the risks it did not bear. Hence prejudgment interest should be awarded at the risk-free interest rate.
The other economic theory of pre-award interest is the ‘forced loan’ theory. We take as a starting point that, if the respondent had paid the damage as of the date of the bad act, the claimant would have been made whole. Effectively, therefore, the claimant has not only been deprived of the use of money since the date of the bad act, but this money has been available to the respondent. It is as if the claimant has made a loan to the respondent. If the claimant had received the money then, and then lent it to the respondent, the claimant would be in exactly the same position they are in. In terms of economic incentives, the forced loan theory ensures that the respondent does not have an incentive to expropriate assets merely as a cheap source of finance.
The difference between the risk-free rate theory and the forced loan theory amounts to a different view of default risk. The risk-free rate theory takes the view that the respondent will pay the award. The reason why the respondent, at the time of the bad act, had to pay more than the risk-free rate for its borrowing is that lenders considered the respondent might default. Under this theory, there is no reason for the tribunal to build a premium into the damages award to offset the possibility that the award might not be paid. Instead, the tribunal sets its award on the assumption that the award will be paid.
Some investment treaties refer to the use of a ‘normal commercial rate’, and some arbitral awards use this phrase, sometimes to justify the use of LIBOR or LIBOR plus a spread. However, all market rates could be described as ‘commercial’; the difference between different rates is that they relate to different risks. The commercial rate for a risk-free loan is not the same as the commercial rate for a risky loan.
What rates do tribunals actually use?
I have described the different interest rates that arise in arbitration cases and given the rationales for two possible choices for pre-award interest rates, namely the risk-free rate and the respondent’s borrowing cost. But what do tribunals actually use for pre-award interest?
Since many awards are confidential, it is not possible to conduct a comprehensive study of all arbitral awards. However, some awards are published. ICSID, a unit of the World Bank, conducts many important arbitrations and publishes its awards.
I have studied all the cases published on the ICSID website. Out of 167 decisions with published awards, I have selected 60 where pre-judgment interest was awarded. Most of the remaining cases are cases where the claimant did not prevail and no damages were awarded (34 cases) or the tribunal ruled they had no jurisdiction (29 cases) (in the remaining 44 cases the award is not available for the analysis in this chapter for a variety of reasons, such as the case was settled or discontinued, there is no English language award document, the award document is redacted, or insufficient detail is provided about pre-award interest).
In 80 per cent of cases (48 out of 60), compound pre-award interest is awarded, which is what economists would normally recommend. However, all but one of the 12 cases with simple pre-award interest are prior to 2010. For 2010 onwards, compound pre-award interest was awarded in 29 out of 30 cases. Thus, in recent cases tribunals have preferred to award compound interest. In the single exception, ARB/07/29, it appears that the tribunal awarded simple interest because the claimant did not request compound interest. In some recent cases, tribunals have cited economic as opposed to legal reasons for preferring compound interest. For example, in ARB/05/18 the tribunal observed that:
Simple interest has the great advantage of simplicity; but it is often a simplicity combined with arbitrariness. When the question is, what amount has the Claimant lost by being wrongly denied payment of a sum on a certain date in the past, in circumstances where the Claimant could have invested an equivalent sum, or could only have borrowed an equivalent sum, on terms of compound interest, the award of compound interest is appropriate. The Tribunal takes the view that an award of compound interest is appropriate in this case.
In ARB/05/24 the award stated that ‘The Tribunal has little difficulty accepting that interest should be compounding. In modern practice, tribunals often compound interest, and the Claimant referenced a number of such awards ... In essence, compounding interest reflects simple economic sense.’ In ARB/07/17:
The Arbitral Tribunal notes that there is no uniform case-law on this matter but considers that compound interest is in the present case to be preferred in order to eliminate the consequences of the conduct which the Tribunal has found to give rise to an obligation to pay damages.
In other cases, reasons given for preferring simple interest include ‘Ecuadorian law prohibits compound interest in the present case’ (ARB/04/19). In ARB(AF)/04/5, the tribunal gave a legal reason for preferring simple interest:
However, since this is not an expropriation case, but rather concerns the appropriate compensation to be paid to Claimants for the injury caused as a result of the Respondent’s breach of the national treatment and performance requirements obligations under Chapter Eleven, the Tribunal’s view is that simple interest is appropriate in the present case.
Turning to the rates awarded, we can see that the interest rates fall into three categories. In 23 of the 60 cases, pre-award interest is a base rate (a market rate of some kind) plus an adder (or spread), such as LIBOR plus 1 per cent or US Treasury bills plus 2 per cent. In 14 cases, pre-award interest was simply a number specified by the tribunal, such as 4.5 per cent or 9 per cent. In the remaining 23 cases, interest was a base rate (a market rate) without a spread.
Of the 46 cases with a base rate (with or without addition of a spread), 13 cases used the US Treasury bill rate, and 27 used an interbank lending rate (including BRIBOR, EURIBOR, LIBOR and ROBOR). As noted above, interbank rates are normally close to T bill yields, and since the adders chosen by tribunals are bigger by comparison, for practical purposes the 13 cases based on a T bill rate can be viewed as not too dissimilar to the 27 cases with an interbank rate.
Some cases stand out for levels of pre-award interest that are sharply higher than a risk-free rate. In ARB (AF)/00/2, the tribunal awarded pre-award interest of 6 per cent at a time (May 2003) when T bill yields were in the region of 1 per cent. In ARB/07/16, the tribunal awarded pre-award interest of 9.11 per cent based on the risk-free rate plus an equity market risk premium. By way of illustration, a 6 per cent pre-award interest rate will give total damages (initial damages plus pre-award interest) that are about 50 per cent higher than a 1 per cent pre-award interest rate, if the award date is eight or more years later than the date of the bad act (1.068/1.018=1.47). If we compare a pre-award interest rate of 9.11 per cent to a 1 per cent interest rate, the same is true after five or more years (1.0915/1.015=1.47). These calculations assume compounding. So interest rates that are substantially higher than the risk-free rate will lead to damages that are heavily increased by the addition of pre-award interest, if several years have elapsed between the date of the harm and the date of the arbitral award.
Tribunals give diverse reasons for their choice of pre-award interest. In some cases it appears that the claimants have provided a calculation in their claim that incorporates a particular pre-award interest rate, and respondents have not specifically challenged this calculation, so the tribunals had used the interest rate applied by the claimants. In other cases, tribunals have used risk-free rates or interbank rates as appropriate market benchmark rates. For example, as stated in ARB/05/24:
The purpose of interest is to ‘compensate the injured party for not having had the use of the money between the date when it ought to have been paid and the date of the payment.’ It is therefore appropriate that the rate of interest represents a reasonable and fair rate that approximates the return the injured party might have earned if it had had the use of its money over the full period of time ... The Tribunal observes that it is common in investment treaty cases to tie the interest rate to LIBOR – although in the present case, where the currency is euros, it is more appropriate to use EURIBOR. This represents an objective, market-orientated rate, well suited to ensuring that the consequences of the breach are indeed wiped out.
The forced loan theory is explicitly referred to in several awards, for example, in ARB/07/23 as follows:
Claimant has argued for compound interest at a rate of 9.34% based on the rate that Respondent paid to private and public creditors in 2006 and on the notion of a coerced loan from Claimant to Respondent. Respondent has suggested a pre award interest rate equivalent to six-month LIBOR plus two percentage points. The Tribunal disagrees with the coerced loan rationale of Claimant to arrive at the proposed rate of interest. ... The Tribunal considers that the rate proposed by Respondent is a commercially reasonable rate.
In ARB/11/26, it appears that claimant’s expert applied the forced loan approach, while the respondent’s expert applied a ‘country risk’ approach leading to a quantitatively similar, but smaller rate, and the tribunal chose a rate that was close to both those proposed:
In examining the use of an appropriate ‘borrowing rate’, the Tribunal notes that Claimants (making reference to the language of the Portuguese Treaty) have argued that the interest rate should be equivalent: ‘to the rate Venezuela would have had to pay to borrow money in April 2008 (9.75%).’ Taking a different approach, Respondent’s expert, … discusses the use of such rates in other awards which are then supplemented by a factor covering political risk and other macroeconomic factors (Country Risk Premium). … Comparing this rate with the 9.75% borrowing rate for the government of Venezuela propounded by Claimants, the Tribunal concludes that 9% is a reasonable and fair rate for pre-award interest.
Some awards refer to interest rates specified in treaties. In ARB/09/16:
The relevant standard for purposes of compensation would be that provided in Article 5 of the BIT, ‘namely the market value of the investment adjusted for interest calculated on the annual LIBOR basis’ ... The Tribunal finds it appropriate to use the LIBOR rate of interest as specified in Article 5.
ARB/10/13 uses similar reasoning.
Finally, it is worth noting that tribunals appear to be aware of the issues concerning the appropriateness of different interest rates for different currencies. ARB/05/24 states ‘The Tribunal observes that it is common in investment treaty cases to tie the interest rate to LIBOR – although in the present case, where the currency is euros, it is more appropriate to use EURIBOR.’ In other cases where awards were made in Canadian dollars, Mexican pesos, etc., the tribunals have explicitly recognised the need for a rate applicable to the currency.
- James Dow is Professor of Finance at London Business School.
- Fisher, Franklin M. and R. Craig Romaine.; Janis Joplin’s Yearbook and the Theory of Damages. Journal of Accounting, Auditing & Finance, Vol. 5, Nos. 1/2 (1990), p. 146.