Do Common Factors Really Explain the Cross-Section of Stock Returns?
Alejandro Lopez-Lira, Nikolai L. Roussanov – 2020
https://www.youtube.com/watch?v=uwgaIYQ-RfA
Quote
The central insight of asset pricing theory is that only systematic risk should be rewarded with an average return in excess of the risk-free rate. In particular, the arbitrage pricing theory (APT) of Ross (1976) posits that certain securities earn higher expected returns than others only because they are more exposed to common (i.e., undiversifiable) risk factors. Conversely, the expected excess returns of portfolios hedged against all systematic risk should be zero…
Excess returns can only come from systematic risk, that is to say, risk factors that are common across stocks and therefore can’t be diversified away. These systematic risks drive time series variation in returns. Intuitively, you get paid for taking risks that were forced upon you by the market, any any risk that you can somehow get around via diversification shouldn’t pay you anything. If you somehow manage to hedge against all the systematic risks, your return should be zero. This is in some sense a “no free lunch” theorem – there is no (excess) return without systematic risk. “Alpha” is in some sense the violation of this principle, represented as a numeric value.
Relevant to Where Is the Alpha in AI?