A link between options timing and taxes?

The Wall Street Journal recently reported on the growing scandal involving the "back-dating" of stock options issued as compensation to corporate executives. The scheme here, according to the news reports and some academic studies, is to issue the options as of a certain past date when the stock was at a lower level. In other words, if the stock is trading at $50 today but $20 dollars two months ago, the firm could issue the options (with a strike price equal to the market price) as of two months ago, allowing the executive to cash out $30 per share immediately. There is nothing illegal about this. So long as the appropriate corporate procedures (e.g., shareholder and board approval) were followed, a firm can issue any options, including back-dated options, it wants.

So why the big fuss?

For one, the failure to disclose this fact, however, would likely be a material misstatement or omission under securities laws. This perhaps explains the investigation that the SEC is turning up the heat on every day. For another, some of the allegations involve claims that corporate procedures weren't followed -- CEOs manipulated documents and/or the board in such a way as to hide the true nature of their compensation. While surely possible, none of this is proved, and we should withhold judgment until it is. Whether ultimately proved or not, this is further evidence for corporate critics, especially critics of executive compensation, that corporate governance is in need of fundamental reforms.

I don't want to rehash all of the arguments regarding the possible issues here, but instead want to consider the impact of a particular tax rule on compensation practices. In the early 1990s, the Congress passed a bill, signed by President Clinton, that capped the allowable tax deduction for executive compensation at $1 million. An important exemption was made, however, for "performance-based" compensation, and stock options are presumptively performance based. This rule severely constrains a firm in the manner in which it can compensate its executives, a tricky task to begin with given the complexities of designing incentive contracts and the firm’s lack of knowledge about an executive’s wealth and private financial dealings.

Consider the following case: a firm that wants to pay an executive $5 million in cash and provide the executive with $10 million worth of stock options (to create incentives to maximize shareholder value), can write this contract just like that, but will be able to deduct only $1 million of the cash compensation. Alternatively, the firm can pay a cash salary of $1 million, give the $10 million worth of stock options, and, make up the $4 million shortfall in cash, by back-dating some options to create an automatic $4 million profit based on the current trading price. (The firm could not as easily just write the make-up options at a strike price that guaranteed a certain and immediate $4 million profit, since they would not, on their face, satisfy the "performance-based" requirement of section 162(m) of the Internal Revenue Code. Of course, back-dating once discovered also does not meet this standard. Expect the IRS to visit this issue with the firms in question.)

Without defending these firms, since we don't know what they did or why, it is a serious question whether this rule encourages this type of misleading behavior. I don't know if this is what some firms were doing, but it is possible that a rule effectively requiring firms to use incentive contracts when they didn’t want or need to, encouraged the deception, whether it was legal or illegal. I tend to think that firms should be able to pay what they want to whom they want in the form they want, so long as it is properly disclosed. Insofar as the tax rule in question biases one form over another or encourages deception, and especially since it has not in anyway constrained executive pay, it must be viewed as a failure.

Comments

[JimOfMN]

Thank you.
It was dumb to pass such a law. The consequences are worse than direct payment because the size of the prize is unknown up front.

These various incentives have also highlighted poor oversight by Boards of Directors AND the way certain CEOs showed the Board only a piece of the compensation package at a time preventing them from seeing the Whole picture

e.g.
Grasso of the NYSE
McGuire of United Health ($1.6 Billion options vested and about $400 million exercised.)

File this under:
Rules that-attempt-to-solve-a-problem gone bad.
or
Consequences of both:
Not Thinking something Through up front
AND
Not fixing when an error becomes apparent.

Other Examples:

1. Most attempts to stop the buying of favor in politics via ad hoc spending and donation rules.

2. Did an exception to CAFE mileage standards & the Luxury tax spawn the conversion of the rough-n-ready Jeep into the deluxe SUV industry?

3. The attempt by Nixon to collect tax from wealthy non-payers via the Alternative Minimum Tax that unfortunately did not index the threshold of taxation to either inflation or the Top X % of wealth.

[Sean]

So, to summarize: when a certain behavior is made illegal (or in this case non-deductible), people who want to engage in said behavior are more likely to break the law than if said behavior is legal.

Tautology much?

[Todd Henderson]

My point is not tautological at all.

The purported goal of section 162(m) of the IRC (the provision that limits deductibility) was to reduce total amounts of executive compensation. This section was passed during a flurry of academic, activist, and media criticism of the levels of executive pay. On this count, the law has been a failure. Pay levels rose dramatically in the wake of the passage. Why? The law also required disclosure of the exact amounts and form of pay. One reason often cited for this is that executives at rival firms now knew what others were making. The Lake Woebegone effect then kicked in, as all CEOs wanted to be paid more than average, and no firm wanted to be considered paying below average. The one-way ratchet here is obvious.

Because cash levels were now capped (for tax purposes), firms had every reason to pay more with non-cash, either perks or equity. Since performance-based pay (e.g., options) was specifically exempted, firms could be expected to pay more of this. The problem is, of course, that the use of options doesn't make sense in all cases or at levels necessary to attract and reward talent, especially given the heterogeneous preferences and wealth profiles of potential candidates. The law thus chose one form of compensation over another per se, without recognizing that this choice – the pay mix of cash and non-cash -- is something that cannot be made at a general level. It should be/needs to be a firm choice, not a government choice.

The law therefore distorts free choice, and, this is the key point, without a reasonable relation between the distortion and any sensible policy end. If the government passed a law capping salary, options, and retirement contributions (or any other form of monetary payment at $10 million per year), it isn't hard to imagine firms "paying" a lot more in perks to meet the exogenously set market wage for certain talent, which might exceed $10 million. If we observe this deception, which is merely designed to pay people what their value is, then we would conclude that this is a bad law. The market will set the wage, and the law should not encourage deception in how that wage is delivered.

So yes, we get more law breaking when things are made unlawful, but that is not my point. The point is that when the law is ill conceived, doesn't have the policy effect intended, and irrationally distorts behavior and encourages cheating, we can say that this law is a failure. Especially since cheating has a way of spreading from one area of behavior or firm culture to another.

* * *

As a post-script, I've talked to a few executive compensation lawyers that advise large firms, and they believe that most if not all of these cases involve the managers deceiving the board, not the company deceiving the government, so to speak. Thus this issue may be limited to a small number of cases. The theoretical point is still somewhat sound, and I believe that it does have the distorting effect described.