Optimal Hedging Portfolios for Derivative Securities in the Presence of Large Transaction Costs
Posted: 10 Sep 1999
We introduce a new class of strategies for hedging derivative securities taking into account transaction costs, assuming lognormal continuous-time prices for the underlying asset. We do not assume that the payoff is convex as in Leland (J of Finance, 1985), or that the transaction costs are small compared to the price changes between portfolio adjustments, as in Hoggard, Whalley and Wilmott (Adv. in Futures and Options Res., 1993). The Leland number, A, which is proportional to the ratio of the round-trip tansaction cost over the typical price movement during the period between transactions, is a measure of the importance of transaction costs versus hedging risk. If A is greater than or equal to one, standard delta-hedging methods fail unless the payoff of the derivative security is a convex function of the price of the underlying asset. In contrast, our new strategies can be used effectively in the presence of large transaction costs to control simultaneously hedge-slippage as well as hedging costs. These strategies are associated with the solution an "obstacle problem" for a Black-Scholes diffusion equation with Leland's "augmented" volatility, a parameter which depends on the volatility of the underlying asset as well as on A. The new strategies are such that the frequency for rebalancing the portfolio is variable. There are periods in which rehedging takes place often to control gamma-risk and other periods, which can be relatively long, when no transactions are needed. Moreover, instead of replicating exactly the final payoff, the strategies can yield a positive cash flow at expiration, according to the price history of the underlying security. The solution to the "obstacle problem" is often simple to calculate. There exist closed-form solutions for various securities of practical interest, such as digital options.
JEL Classification: G13
Suggested Citation: Suggested Citation