Betting Against Beta: Leverage Aversion as a Premium
Why might low-beta assets earn more than their market beta predicts?
Funding constraints can lead investors seeking higher returns to bid up high-beta securities rather than borrow against low-beta holdings. Frazzini and Pedersen test the resulting relative premium through a beta-neutral portfolio.
The published gross pattern does not establish a tradable net return after financing, margin, shorting and turnover costs.
Evidence map
| Aspect | Finding |
|---|---|
| What it is | The finding that low-beta assets earn higher risk-adjusted returns than high-beta ones, and the explanation that ties it to a constraint rather than to a risk: investors who want more return but cannot or will not borrow bid up high-beta assets instead. |
| Key result / formula | Frazzini and Pedersen construct a portfolio that is long a leveraged basket of low-beta securities and short a de-leveraged basket of high-beta ones, scaled so that the combination has no net market exposure. |
| Why it matters for backtesting | Two transfers. The first is a hypothesis a user can test on their own instruments: sort by exposure to a common factor and compare risk-adjusted returns across the sorts, which is available on bars once a market proxy is defined. |
What it is
The reported premium is only potentially accessible after borrowing costs, margin requirements, shorting costs and trading costs.
Key result / formula
That portfolio earns positive risk-adjusted returns, and the result appears consistently across many asset classes and many countries, which is unusual for a documented anomaly. The model behind it is explicit: agents facing leverage constraints express their desire for higher returns by buying riskier assets, which flattens the relationship between beta and expected return relative to what the standard model predicts. The theory makes further predictions that the authors test and find supported, including that the strategy should do badly when funding conditions tighten, since that is when constrained investors are forced to reduce leverage, and that the effect should be stronger where constraints bind harder.
Why it matters for backtesting
The second is more important and is about the user's own position: a strategy that harvests this premium is short funding liquidity, so its bad periods coincide with periods when leverage becomes expensive, which is exactly when the user is least able to hold it. That correlation between the strategy's drawdown and the user's own constraint does not appear in a Sharpe ratio and must be stated separately. The general lesson is that a premium explained by a constraint is available solely to someone the constraint does not bind, and a user who needs leverage to harvest it is not that person.
Source
Frazzini & Pedersen, "Betting against beta", Journal of Financial Economics 111(1), 2014, 1-25. Primary source