Monday, March 23, 2009

Stock Option Friction Revisited

Last week I showed that the frictional costs of commissions and spreads are higher with stock options than they are with trading stocks. Reader feedback included Mark Wolfinger’s comment that I hadn’t accounted for interest on cash not tied up when investing with options, and Glenn’s comment that option spreads are more reasonable in the U.S. So, let’s do the accounting a little differently and see how it affects the results.

The original scenario was to compare direct ownership of 200 shares of RIM to a stock option-based synthetic version of RIM stock ownership where we buy 200 call options at the money and simultaneously sell short 200 put options at the money.

This time let’s assume that the market gets the option pricing right to account for the interest on cash that gets tied up with stock ownership but isn’t tied up with the synthetic stock option version. We can calculate the friction costs by simply assuming that each trade loses half the bid-ask spread plus whatever commissions are charged.

Stock Ownership

In the case of stock ownership, we buy and later sell 200 shares:

Trades: 2
Total shares traded: 400

Assuming $10 per trade, friction costs are then

$20
+ 200 times the average bid-ask spread.

Synthetic Stock Option Approach

In the synthetic stock option case, we buy 200 call options and sell 200 put options initially. Later we would buy or sell 200 options depending on which type of option is in the money. However, we don’t know in advance what the spreads will be, and so we’ll assume that we wouldn’t tolerate spreads more costly than exercising the option and closing out the stock position. Let’s treat the friction costs as though the option gets exercised to avoid overestimating spread costs on options.

Trades: 3
Stock shares traded: 200
Total options traded: 400 (4 option contracts)

Assuming $10 per trade plus $1.25 per contract, friction costs are then

$25
+ 100 times the average stock bid-ask spread
+ 200 times the average option spread.

Montreal Exchange

The average spreads from the snapshot taken of the Montreal Exchange in the original post were 2 cents for RIM stock and $2.07 for September RIM options at the money. So using the costs worked out above, friction is $24 in the stock case, and $442 in the option case. This result isn’t much different from the roughly $500 I calculated for the option case in the original post.

Chicago Board Options Exchange (CBOE)

The average spreads from a snapshot taken of the U.S. market for RIM last Friday on the CBOE were 1 cent for RIM stock and 12.5 cents for September RIM options near the money. So using the costs worked out above, friction is $22 in the stock case, and $51 in the option case.

It’s clear that for RIM stock, U.S. options have much more reasonable costs than Canadian options. However, the US$51 cost represents about 0.6% of the price of 200 shares of RIM. In addition, this option strategy simulated stock ownership for only 6 months. If we repeat this strategy for another 6 months, the total costs for a year are 1.2%. Giving up 1.2% per year is a big drag on the returns of any strategy. After 25 years, the total bite out of a portfolio would be 26%.

Stock option friction varies considerably, and investors need to pay careful attention to these costs. Option-based strategies tend to be short-term which makes the costs recur frequently. Even when these costs are minimized by trading in issues with low bid-ask spreads, the costs are significant. Stock options have a place for risk management by knowledgeable investors, but the effect of friction has to be factored into the analysis of any investing strategy.

No comments:

Post a Comment