RSS
Economic perspectives on the determination of harm for white-collar sentencing.
Feb 17th, 2008 by admin
By: Kenneth D. Gartrell
The ongoing wave of white-collar crime prosecutions recently led by a spree of option-backdating cases raises some important economic questions about how to measure harm for sentencing purposes and some equally important questions about the relationship of overlapping civil and criminal proceedings.
There appears to be a search in the legal community for a single measure of harm for sentencing in option backdating cases under the Federal guidelines. The search for a ready standard seems doomed for two reasons. First, that the actual economic harm to the victims of a crime must be measured in terms of the specific facts of each case. Second, that it seems likely the economic methods ultimately have to conform to conclusions reached by the Supreme Court of the United States in Dura Pharmaceuticals (April 19, 2005).
Even though there is a single conceptual measure of economic harm universally acceptable to economists, there is no convenient single method to quantify damages or harm from option backdating. Every case is as different analytically as it is factually. The alternative methods have to be weighed against each other and ultimately reconciled. All the forensic tools of finance, economics and accounting must be jointly weighed with the facts to converge on the most reliable measure of harm. The Dura decision raises this imperative to even higher levels than before.
The difference between the market price of the company’s shares on the backdating and the price determined from but for cash flow analysis would be the correct economic measure of total potential loss regardless of whether a civil or criminal case. It is the only measure of harm that restores the victims to the instant before harm. Perhaps more importantly, it is the only equitable and just measure of harm that would in all cases rationally deter potential perpetrators.
Even though there is little disagreement among economists what constitutes the proper measure of economic harm, there is widespread opinion in the broader legal community. There are also legal/practical considerations that complicate the efforts to define and measure harm.
Until a month ago, the legal community thought that the Reyes sentencing would set the standard for application of Federal sentencing guidelines in option backdating cases (U.S. v. Reyes, No. CR06-556CRB). But, on January 17, 2008, Reyes was sentenced to 21 months and Judge Breyer found no provable harms. Without a measure of harm, the judge had to find another rationale to impose a jail term. It remains to be seen if the decisions can sustain an appeal and, in the meantime, no standard for the determination of economic harm developed from the case.
Prior to the Reyes result there was widespread speculation of how Judge Breyer might determine the measure of harm. It turns out the Court found that it was much more difficult to assess harm than was imagined in the speculation leading up to the final determination on January 17, 2008.
An accounting approach
“Reyes will give us the road map,” Professor Peter Henning speculated in an August interview. Henning suggested that judges might rely on the corporate restatement of earnings as a benchmark of losses even though the restatement of hundreds of millions in losses could have added five years to a prison sentence. Reyes’ company Brocade, for example, restated earnings to show $100 million in noncash expense attributed to options backdating.
Judge Breyer showed wisdom in bypassing the accounting measures. By themselves, the use of the retroactive financial accounting adjustments would not provide a reliable measure of harm. The accounting adjustment for the backdating options is to record an accrued expense for employee compensation allocable to the vesting schedule of the grant. It is only a bookkeeping entry and involves no outlay of cash. There would be no cash flow or valuation implications. The historical operating cash flows would remain unchanged and the proportionate claims to the dividend stream would remain unchanged.
Taken a step further in terms of the Dura decision, the fact of the backdating could not have been “generally known” before the actual grant date (see Dura). Regardless of any restatement, no one can go back in time before the actual grant date. The victims could not have been harmed and the defendant could not have benefited during the time that passed from the backdating until the actual grant date on the basis of the backdating or the accounting per se.
Schemes
As the Supreme Court left open in Dura, however, it is possible for a civil plaintiff to plead that a scheme was planned in advance and harm was done to shareholders from the time of backdating until the actual grant date and beyond because of a fraudulent effort to manipulate earnings. For an options backdating case, such a theory creates substantial complexity and insists on a type of analysis to determine economic harm as would be found in a typical 10-b5 class action.
A scheme, like the one alleged in the Reyes case, is a realistic possibility. As commentators have pointed out (see NERA) the practice of backdating options has been a part of business for many years and there is widespread sophistication in their use. Cases may exist, especially in short windows around earnings announcements and public filings, where backdating was anticipated in advance and was undertaken as part of a scheme to assure that promised immediate earnings targets were met. The expectation on the part of scheming defendants could have been to award each other unwarranted compensation for the attainment of the earlier earnings targets.
Cash flows and market prices
In a scheme, the financial accounting provides no reliable measure of economic harm. What would be required to allege, and later prove, damages (or by inference harm in the sentencing phase of a criminal proceeding) is a factually sustainable but for reconstruction of the manipulated cash flows — including tax effects, transactions costs, compliance costs and all other considerations having material cash flow implications.
In purely economic terms this ex ante cash flow analysis would be an unbiased and proper measure of damages or harm. But in a legal context, it is not possible to stop at that point because the Supreme Court in Dura decided on a need for an ex post determination of economic harm regardless of what can be shown about the but for situation at the time of the presumed harm. This imposes the need to consider the facts of when and where potential losses were “realized” by the plaintiffs or victims. It must be considered when, as the Supreme Court says, “the facts become generally known and as a result the share price depreciates.”
Economics aside, the Supreme Court has decided that there must be a “depreciation” of share price for harm in Section 10-b5 class actions. And, this introduces the further need to reconcile but for cash flow analysis to stock price changes isolated by use of event studies. It follows that because share trading continues between the date of true harm and the dates of share “depreciation” not all shareholders are harmed alike. Hence, some form of trading model has to be employed to quantify economic harm.
Under the rules, this type of analysis can scarcely be avoided. It gets very expensive and very fast. It is possible, moreover, that no causal link will ever be found in the analysis that connects the option backdating to demonstrable harm to the victims. In the most elaborate schemes, with the most planning and with the most sophistication there are many escape routes available.
In the earlier speculation about emerging standard, some lawyers offered the view that stock price drops are inappropriate for backdating cases “ because it would be hard to show backdating affected investor decision.” Hard? Yes. Quite so, but the Dura decision says it needs to be done both to plead and prove harm sufficient for damages in a civil securities case under Section 10(b). Could the requirements for the proof of harm be any less in a criminal matter? Ultimately, we would not expect a difference.
Accounting revisited
Obviously, the complete determination of actual economic harm via cash flow analysis is fraught with complexity. And, this is where it may be feasible to reintroduce the prospect of using the accounting compensation calculations as the measure of harm. On the basis of long-established professional custom and practice, an accounting expert could opine where a scheme is present that economic cash flow analysis is subjective and unreliable and that the ill-gotten gain manifest in the schemed for compensation provides a quantification of harm which is “more objectively determinable.”
The Courts rely on the testimony of both economic and accounting experts. Accountants have an economic interest in having their particular methods applied. Further, in a world of perfect information and low transactions costs, there should be no difference between the economic loss sustained by a victim or plaintiff and the ill-gotten gain enjoyed by the defendant.
A rebuttal to this method exists, nevertheless, because the formal accounting principles concerning the calculation of option-based compensation would at its core only be a shortcut to the ultimate cash flow analysis. The option-pricing model is only as good as its assumptions. The market price used in the model ex ante would have to be the but for price of the security on the date of harm. This, in turn, can only be calculated from the discounted cash flow analysis. It does no good to back cast based on observed prices at the grant date.
Before and after period
An article on Law.com by Pamela MacLean observed “… Northern District of California Judge Charles Breyer, who will sentence Reyes in November, has signaled his unwillingness to use a single stock price drop event to calculate losses for sentencing purposes in a 2003 fraud case. He settled on the average stock price for 60 days during the fraud and the 60-day average after the fraud to show the loss. U.S. v. Grabske, 260 F. Supp. 2d 866…”
Perhaps there is little more than irony in the fact that Dura was a Supreme Court decision to reverse a ruling by the Ninth Circuit “ that loss causation (could be established) …simply by alleging and in the complaint and subsequently establishing that the price on the date of purchase (harm) was inflated merely because of misrepresentation.”
Whatever the thinking or motivation could be behind such an approach the Judge again acted wisely in not adopting it. No sense can be made of it of in terms of modern financial economics. The field has maintained, over decades of research and testing, that stock prices follow a random walk. As such, there is no information in the historical trends of a given a stock that predict the future trend. This finding is a central tenant of the Efficient Markets Hypothesis and it stands behind the entire legal theory of “fraud on the market.”
Hence as Professor Eric Lie shows in his option backdating lookback analysis, anything can happen to the price — even in a shorter 30-day lookback period. A 60-day period does nothing to resolve the fact that stock prices are not structurally stable within a 60-day period, let alone across two concurrent 60-day periods.
Lesser of intrinsic value of compensation or actual cash benefit
A final proposal from the early speculation proffered that the most logical calculation would be the difference between the “backdated price and what the price should have been, then factoring in the number of options actually exercised to purchase stock shares….”
Actually this is just another version of the ill-gotten gains versus cash flow analysis discussed above. The only apparent virtue of the approach is simplicity. But, it still confronts all the same issues as before except for the simplifications that:
1) Reduce the measure of compensation or ill-gotten gain to intrinsic value of the option (which is what I take the language — backdated price versus what the price should have been — to mean), and
2) Reduce the measure of damages further to the gain actually realized by the defendant upon exercise.
The approach would satisfy neither the economist nor the accountant professionally. Accounting has long since moved beyond intrinsic value standards. Economics, furthermore, would find at least two major faults in these restricted metrics. First, even at the money options, due to the elements of risks associated with future events, have positive value even when the apparent intrinsic value is zero. Second, basing the harm on the realized exercised value would be akin to limiting the harm done by a petty thief to what he or she can sell a stolen camera for at the local pawnshop. It would have to be a discount below fair market value due to expedience and other marketing costs. The real loss to the original owner would be fair market value of the stolen camera plus the transactions costs to replace the camera.
Law of One Price
Unequivocally, the ex ante difference between the observed price of the stock affected by option backdating and its but for price is the theoretically correct measure of economic harm done by the backdating and/or the scheme it represents. But because of factual ambiguity and because of the constraints of Dura the analysis requires a many faceted approach. The differences in proposed measures are not differences in kind, but rather differences in perspective.
The differing elements of the preferred but for analysis are not independent. They are all part of the same fabric and only represent alternative routes to the same underlying conclusions. In a perfect economic world with no information costs or transactions costs, there is only one measure of harm. Conforming to the “law of one price,” the cash flow analysis and the accounting analysis will agree as two sides of the same coin — just as the cash flow analysis must agree to the analysis of realized harm from stock price depreciation when the “facts are generally known.”
For any case to be decided in advance based on a pre-selected method is the same as to decide the facts ahead of time. The correct economic approach for the quantification of economic harm is to use all the facts in a forensic fashion to converge on the correct measure of harm with a full understanding of a need for the reconciliation of the methods. When there are apparent differences across methods of analysis there is often compromise at some middle ground. No matter what the pathway is, it cannot be engaged independently. The actual available path in each case is entirely dependent on the facts of that case in an “imperfect world.”.
The laws of economics as interpreted directly by economists, or inferred by way of the customs and practices among accountants, do not change because a particular case is a civil case or a criminal case. How could a situation long endure that treated the same thing in a civil case with lower standards of proof differently from a criminal case with a higher standard of proof? If a certain approach or philosophy is required to meet a lower standard of proof, then it would seem most certainly that the same approach would be required to establish harm at a higher standard of proof.
PUB_Backdating_Part_2_Primer_SEC1502_final.pdf
Sphere It