September 24th, 2014, 1:36 pm
My thought: "better" requires a precise notion of what the objective is. The problem with a lot of the 'volatility estimator' literature is that'better' seems to be defined by comparing estimators for some obviously wrong model process, like Brownian Motion (BM). Since the real world clearly does not evolve as a BM, surely that is not the objective. So, what is the real objective?
Last edited by
Alan on September 23rd, 2014, 10:00 pm, edited 1 time in total.