hedging - is beta of a portfolio always meaningful?

Consider the following strategies:

• a stat arb strategy with no overnight exposure, but significant market exposure intraday.


• a market timing model which is always long or short the market.


• etc

is it meaningful to consider the betas of strategies like these? Or should we ignore beta when the portfolio returns have low (near zero) correlation to market returns? how do you use beta?

Answer

I would split the question into two sub-questions:

1. Is market beta useful at all?


2. Is market beta useful for high-frequency strategies that are fully hedged EOD?

With regards to the first question, I would summarize the hundreds of papers on the subject as: yes, but not as much as it was initially believed. The reason being that multi-factor models are empirically superior to CAPM and intertemporal CAPM, and in these models the market factor explains relatively little volatility, and in some cases it is entirely omitted, as the volatility is entirely captured by industry factors. Yet, risk models are not universally adopted, especially outside of equities, and moreover market beta can still be useful to assess risk premia (as is usually done in investment banking) or to tactically hedge portfolios.

With regards to the second question: it depends on the time scale. If you estimate beta using daily returns, you cannot use this loading to hedge intraday exposure to market risk, since this risk is not captured by the estimation interval. So you'd have to use 30- or 15- minute returns. This is not trivial however, because asynchroneity effects and induced autocorrelation of returns (Epps' effect). I don't know of commercial high-frequency factor models. I am not sure that statistical arbitrageurs use in house models, but there are many technical hurdles to overcome.