Thursday, October 03, 2002
Back-filling on missed news stories after a couple of days at a conference: how many people out there caught the details behind one of the latest nasty corporate headlines, the troubles at EDS? Well, for those who didn’t, EDS – Ross Perot’s old shop – missed its projected earnings by a mile for a number of reasons, but prominent among them was an over $200 million payment that EDS needed to make to settle derivatives contracts on its own stock.
Now why, you might wonder, would EDS have had such derivatives contracts to settle? Was EDS engaged in a legitimate activity or was it speculating recklessly? Well, this I know something about, because I used to be in the business of structuring, marketing and transacting on precisely these kinds of derivatives.
Through the 1990s, a variety of company used equity derivatives on their own stock to “hedge” risks associated with stock repurchase programs and specifically with their stock options grants. I put the word “hedge” in quotes because, while the use of derivatives did hedge certain of these risks, these contracts were not typically designed to be the best hedges they could be. And the reason has to do with accounting.
Let’s start at the beginning: what kinds of derivatives contracts were out there, and why did companies use them in conjunction with stock repurchases and stock options grants? Broadly speaking, the derivatives contracts came in two forms, puts and forwards, and they were used for two reasons, to reduce the cost of repurchasing stock and to finance those purchases off balance sheet.
Suppose a company wanted to buy back between 1 million and 1.2 million shares of stock over the next year, for whatever reason. The company might simply go into the market and buy all the shares. This would seem to be the logical thing to do. But if the company did that, it would have to put up the money right now, and that might put a strain on liquidity and worsen debt-to-equity ratios. Alternatively, the company could purchase the stock on a forward basis. The company would enter into a derivative contract agreeing to purchase the stock in, say, one year at a price somewhat above today’s price for the stock (that higher price reflects the cost of financing). Then, a year later, the company could do one of several things. The company could purchase the shares per the contract. Or, if the company changed its mind, the company could settle the contract for cash. The derivative contract thus achieved two entirely legitimate corporate goals: first, it provided the company with efficient financing; second, it gave the company greater flexibility be deferring the decision of whether to actually purchase shares while locking in the economics of purchasing. So far, so good.
In many instances, the company could also net-settle the contract, either receiving shares or delivering shares of value equal to the cash settlement amount. This provision was important on the one hand because it allowed the company more flexibility; theoretically, at least, the company never had to come up with any cash to settle the trades. In practice, though, no company whose stock price had collapsed would settle a forward by delivering shares because the dilution of current shareholders would be outrageous. The real reason this provision was included was for accounting purposes. By including this provision, the company would not have to disclose the details of the transaction ANYWHERE on its financial statements. It might choose to reveal the number of contracts outstanding, and most companies did. But it would not have to, for example, reveal how large the dollar liability of a contract that went against the company was, since in theory the company could deliver stock and not have a cash liability. And the company would not have to dilute its earnings-per-share to reflect this potential issuance because the company could theoretically settle in cash. Pretty neat, huh? The end result being that a company like EDS could have a $200 million liability coming due in the next quarter and no one could tell from looking at its financials that this was the case.
Now, I don’t know if EDS was transacting in forwards. They might have been transacting in put options. In these transactions, the company writes puts on its own stock. Why would a company do this? The reason is that writing puts monetizes the view that the company would buy more shares if the stock price declined. Without going into a mathematical explanation of why this is so, let me just say the following (skip the rest of this paragraph if you would prefer to just trust me). A put option only has terminal value if the terminal stock price is below the strike price of the put (the price at which the put holder has the right to sell stock), and this terminal value increases dollar for dollar with the decline in the stock price below the strike price. The initial value of the put is the probability-weighted average of all the potential terminal scenarios for the stock price that fall below the strike price (since those that fall above are worth zero). The current value of the put increases as the stock price drops, but it does not increase dollar-for-dollar because that would imply certainty that the terminal price would be below the strike price. As the stock price goes lower and lower, however, the ratio of the change in put option price to the change in stock price approaches 1:1. Therefore, the owner of a put is in the equivalent position to being short more and more shares (or, better, a larger and larger fraction of a share) as the stock price declines. By the same token, the writer of a put is in the equivalent position to owning more and more shares as the stock price declines.
Whew! The point of all that background is to explain why a company would write puts on its own stock. If a company would buy more shares if the stock price declined, the company can capitalize on this view in one of two ways. Either the company can wait and see if the shares go up or down, and buy more if they go up and fewer if the go down; or the company can write puts. In the latter case, the company will be paid today the expected value of its preferred trading strategy of buying as the stock goes down, and will forego the actual realized value of that strategy. If the stock goes up, the company will have been better off having written puts; if it goes down, the company might have been better off waiting and seeing what happened in the market.
So EDS might have bought forwards or it might have written puts. It doesn’t matter, because in either case the contracts would obligate EDS to either buy shares or otherwise settle their contracts at a loss if the stock price declined sufficiently. And either way there would be no indication on the balance sheet that EDS had this potential liability.
That already sounds pretty bad. Now it gets a bit worse, for two reasons. First, motivationally. Remember, these contracts are used in the context of share-repurchases. So remember that the reason why there was such a huge boom in share repurchase activity in the 1990s was because of double-taxation of dividends (which made investors prefer capital gains to dividends) and because of stock-options issuance (which diluted existing shareholders, an effect that could only be offset by buying shares). And remember that the latter activity ballooned to obscene proportions because stock options were not expensed - the cost of the options never showed up on the balance sheet. But this lack of accounting for the cost of options had a further perverse effect: it encouraged riskier hedging schemes than optimal. Why? Well, think about it: if you had issued a call option to your shareholders, what would be the optimal hedge? The simple answer is: you should buy a call option. After all, you should be able to replicate the risks and returns of one call option with another call option more easily than with some other structure. But buying a call option would cost money, and since the calls that were issued were not expensed, there was no reason to spend money. So instead, companies bought forwards or wrote puts, which further increased their leverage. Instead of acquiring an asset to hedge the liability created when a company issued calls, the company acquired another liability in the form of a put or the downside of a forward.
And now the second reason. Investment banks noticed that, in entering into these contracts, they were taking on a lot of credit risk. After all, if these contracts ever went against the company transacting in them, it would be because the company's stock went down dramatically. And that probably meant the company was in some degree of financial difficulty. In the most extreme case, where the company defaulted and the stock went to zero, the investment banks holding these puts or forwards - and expecting large cash payments from the company in question - would suddenly discover they were holding worthless paper that put them at the back of the line in a bankruptcy proceeding. Investment banks didn't like this at all, so they included provisions in these contracts that forced the corporate counterpart to unwind the transactions if the stock price dropped sufficiently. Now the investment bank would get paid before the company went bankrupt. Investment banks had figured out the seriousness of the potential liability and consequent credit risk, and built a hedge into the contract. No such option was provided to shareholders and general creditors, who were not even made aware of the full existence of the risk, much less its full ramifications.
Accounting matters, and not just because loose accounting lets the crooks in. Much of what I'm describing above is potentially quite legitimate. Some of the biggest companies in America saved their shareholders a lot of money by using these derivatives strategies sensibly for a decade. But bad accounting creates perverse incentives, incentives to take risk and erode economic value for the sake of accounting. Shareholders and creditors need to have a real picture of the assets and liabilities of a company for them to sensibly allocate capital, which is what our free-market system is supposed to be all about.