30,000 of the October 40 Puts traded late this afternoon, priced at $17.60 or 15cents above parity based on the stock price of $22.55. This type of volume (also seen in deep puts on CEG today) involves a strategy to avoid being assigned on short positions that have no remaining time value and are likely to be assigned in the overnight clearing process. In this case there are 7494 open contracts in the AA Oct 40 puts. If most of the long put holders exercise tonight, the traders with short positions can expect to be assigned on a pro-rata portion of their position. The remaining unassigned position will yield a small profit. For example, if 75% of the puts get assigned, and two market makers happened to be short 3000 contracts each, they could expect to be assigned on 75% of their positions, or 2250 contracts. The non-assigned 750 contracts is a small amount of extra profit to their portfolio.
This afternoon’s volume of 30,000 contracts is actually two market makers each buying 15,000 from each other and then immediately exercising the long position (without letting it close the short on their books). Their trades grow the total OI to 37,494 contracts with a total of 35,620 exercised instead of 5620. While they will still be assigned, the ratio of contracts exercised to their short position has changed, and that is the key to this arbitrage. Instead of a 75% exercise ratio, now 95% of the open interest will be assigned. Each market maker can expect to see 95% of their 18000 short contracts assigned, but this 17,100 contract assignment will be netted against their own exercise of 15,000, resulting in a final net assignment of 2100 contracts each, instead of 2250 (or 70% of their short). If the gain from non-assignment is 5cents per contract, this works out to risk-free profit of $747.12 for each market-maker.
Who loses that money? The participant who was short the remaining 1494 contracts and unable to execute the same trade. If that trader runs the numbers he or she will find they were assigned on 95% of their position, even though only 75% of the incoming open interest was exercised. The strategy works on deep calls prior to ex-dividend date in the same way.
Post Script 10/1/08 OCC data shows that only 35% of the original open interest were exercised- changing the net effect to $1942 of profit per market maker.