In April 2008, dividend risk nearly wiped out our trading business. A few months earlier, we had founded Android Alpha Fund and our first few months had been a stellar success. Our system was built to capitalize on short-lived pricing inefficiencies and the models that formed the basis for our strategy were proving to be remarkably effective in the real options markets. Then, in mid April we entered into a spread trade involving BPT call options. BPT is a stock that derives its value based on the value of mineral rights in the Prudhoe Bay oil field on the North Slope of Alaska. For our purposes, all you really need to know is that BPT pays a LARGE dividend on a per share basis (about $12 per year back in 2008, paid quarterly).
We had sold and bought an equal number of BPT calls, all of which were deeply in-the-money. At the time of the trade, we had made what we believed was a modest arbitrage profit on the trade. Our plan was to simply hold the options until expiry and then book the profit. Under normal circumstances, it does not pay to early exercise call options, since it means that you need to put up the capital for the stock purchase earlier, thus forfeiting the time value of money.
We woke up one morning in April (April 14th to be exact) to discover that we had been early assigned on all of the short calls we had sold, meaning instead of having a call spread, we now owned a short stock position and an offsetting amount of long call contracts. Now, it turns out then when you are short a stock, on the ex-dividend date for the stock, you owe the dividend on the stock (think of this as the mirror image of owning stock, where you would collect any dividend that the stock pays). In the case of our BPT position, this turned out to be tens of thousands of dollars worth of dividend payments, dwarfing any supposed arbitrage profit that we had made on the original trade.
Just because you are short dividend-paying calls does not guarantee you will be assigned on them. In the extreme, the counterparty will not exercise calls to capture dividends if the call strike price is out-of-the-money as this would not make economical sense. To determine whether you are likely to be early assigned on a call, one can utilize put-call parity. In the special case being considered here, one will reach the conclusion that it makes sense to early exercise the call ahead of the ex-dividend date if the market price of the corresponding put (the put with the same strike and expiry date as the call under consideration) is lower than the value of the future dividend payments (prior to the contract expiring) minus any interest cost associated with early exercise.
We learnt a valuable lesson that day, namely that you need to be fully aware of any dividend paying stocks in your call portfolio, and to take defensive action (close the position, exercise offsetting calls, buy stock etc.) ahead of the ex-dividend date. We have since automated this process for our own portfolio, relying on a daily check against several unique sources of dividend data, since no single source is 100% accurate (another lesson learnt the hard way, but that is for another day).