Subscriber ATB raised the following insightful question in response to my comments in the July Monthly Review:
In the July monthly review you state that you don’t want to assume a stock picker’s attitude and try and pick the direction of indexes, but rather make predictions on future volatility. I also agree with this belief that no one can pick the direction of stock movements with any discernible edge. My question is why do you think you can make predictions on volatility with greater success than someone who makes predictions on stock movements.
I should clarify one issue – I’m not of the opinion that no one can predict stock movements with any discernible edge; hopefully some of the studies we’ve profiled on the blog in the past demonstrate otherwise. And I would certainly disagree with the view that it’s not possible to beat a buy-and-hold index fund over time. It’s just that directional theses about price movement aren’t what drives this options newsletter.
I’m not sure whether I can make predictions about volatility with more accuracy than I can make predictions about price; that’s a pretty strong claim, and I don’t think I made that claim in the earlier post. I’m not even sure whether I’m confident I can make predictions about volatility with at least as much accuracy, and I’m not sure whether it matters except in an efficient frontier, portfolio theory sort of sense. (And, consistent with my pessimism about subjective rationality and human agency in general, I don’t think my competence at volatility trading, assuming it’s above average, has anything to do with me personally. Unless I’m just lucky or deluded, my methods should in principle be reproducible by anyone else.) Anyway, one reason we might think that someone might be about as good at predicting volatility as they are at predicting price is that some of the same quantitative tools that are useful when applied to price behavior are also useful when applied to volatility. After all, if we’re able to find reliable correlations and relationships among data points in some time series, there’s no a priori reason to discount those results just because the time series measures volatility rather than price. In fact, I’d say the burden of proof is on the skeptic to offer positive reasons why some particular asset or variable is impervious to attempts at analysis. For example, in my ongoing criticism of the application of traditional technical analysis to VIX charts (here and on Twitter), my main objection is that the notions of support and resistance that make so much technical analysis plausible don’t apply to an untradeable statistic, which is what the VIX is; that said, if presented with sufficiently robust contradictory results, I’d gladly revise that opinion.
There’s another reason to be sanguine about our prospects for trading volatility, one that is relatively specific to the strategy I pursue here: if, as our best collective academic efforts and empirical results suggest, there really is a persistent volatility risk premium in equity index options, a strategy that can capture that premium may have a leg up on strategies that study price behavior alone.