A reflection principle for a random walk with implications for volatility estimation using extreme values of asset prices

In this paper, we derive a reflection principle for a random walk with the symmetric double exponential distribution. This allows us to come up with the closed form solution for the joint probability of the running maximum and the terminal value of the random walk. Based on this new theoretical result, we propose an extreme value estimator for the variance of the random walk that is not just approximately unbiased but exactly so. In simulations, we find that this estimator continues to be unbiased even when intraday mean reversion is present, as captured by the Binomial Markov Random Walk model. On the empirical side, we find that this estimator works well in a variety of global stock indices, including the S&P 500 Index, in the sense of being unbiased relative to the “usual” estimator, i.e., the sample variance of the daily returns.