FoxAlgo → Research → Academic Audit → Betting Against Beta Factor in Stocks

Academic audit

Failedequity XS low-beta

Betting Against Beta Factor in Stocks

Trzy bramki próby · prawdziwy tylko, jeśli przejdzie wszystkie trzy i przetrwa kontrargument
Gate 1
Bez survivorship
free
czysty panel
Odrzucony tutaj
Gate 2
Placebo ≥ P95
P0
wyżej niż ~0 z 200 koszyków
Gate 3
Netto po kosztach
RF -0.97
netto ujemny po kosztach
Nieudany
Najgorszy okres 12 mies. (RF)-0.98
dolny limit −1.000
Każda strategia tutaj — także zwycięzcy — traci w najgorszych 12 miesiącach. Głębokość straty to uczciwy kontekst, nie werdykt.
Odrzucony na bramce szczęścia — jego netto nie wypadł lepiej niż losowe koszyki (poniżej linii umiejętności P95).

The betting-against-beta effect argues that low-beta stocks earn higher risk-adjusted returns than high-beta stocks, so the factor goes long low-beta names and short high-beta names to isolate that spread.

What we found

Built naively as a decile long/short (long low-beta, short high-beta), the factor failed our test. Its placebo percentile was 0, meaning it showed no rank skill above randomly formed baskets, and its worst-year risk factor was deeply negative. The reason is structural: this construction is net short market beta, so it gives up the equity premium in every bull year and bleeds over the window. In this simple, unhedged form the famous factor is not investable.

How we tested it
2005–2026 test windowmodelled liquidity-aware costssurvivorship free
  • Survivorship-free 1077-name US common-stock panel, 2005-2026. Realistic modelled costs.
  • Placebo / robustness test: real result vs random baskets or shuffled signals (real vs the 95th percentile of random)
Source: Frazzini & Pedersen (2014), "Betting Against Beta", J. Financial Economics
Read the paper ↗
← The Academic Audit — all 54 studies

Research, not investment advice. “Validated” factor-legs are market-neutral diversifying building blocks with a losing worst year — none is a standalone tradeable strategy. Metrics are cost-aware and modelled (not live fills); the 2005–2026 test window is out-of-sample versus the source paper. Dollar figures are not returns and are omitted by design.