Quadratic-Variation-Based Dynamic Strategies
分析了一类不依赖随机过程假设（仅需连续性和正性）的动态交易策略，能在“已实现累积平方波动率”达到预定水平时复制衍生品收益，并推广了Black-Jones和Brennan-Schwartz的对冲策略，对投资组合保险有重要应用。
The paper analyzes a family of dynamic trading strategies which do not rely on any stochastic process assumptions (aside from continuity and positivity) and in particular do not require predicting future volatilities. Derivative payoffs can still be replicated, except that this occurs at the stopping time at which the “realized cumulative squared volatility” hits a predetermined level. The application of these results to portfolio insurance is emphasized, and hedging strategies studied by Black and Jones and by Brennan and Schwartz are generalized. Classical results on European-style options arise as special cases. For example, the initial cost of replicating a call or a put under the new method is given by a generalized Black-Scholes formula, which yields the ordinary Black-Scholes formula when the volatility is derterministic.