How Leverage and Quality Explain Buffett's Historical Returns

AQR’s Buffett’s Alpha connects Berkshire’s historical returns with its holdings and financing. The researchers find that inexpensive leverage and a preference for high-quality businesses explain a substantial part of those returns.

Source dates and applicability

Buffett’s Alpha was published in 2018, studies 1976–2017, and has a 2019 correction. Leverage and float costs here are historical estimates. As of October 2026, financing and factor exposures remain useful analytical concepts; the sample averages do not describe Berkshire’s current balance sheet or establish returns available today.

1. Berkshire Uses Leverage

Using total assets to equity as its measure, AQR estimates that Berkshire, which uses insurance float, has effectively operated with average long-term leverage of 1.7:1.

Table III of the paper estimates the average annual cost of that insurance float at approximately 1.72%, about 3 percentage points below the contemporaneous T-bill rate. The researchers also estimate that float accounts for roughly 35% of Berkshire’s liabilities on average.

2. Leverage Plus Low Volatility

Buffett has long favored a basket of safe stocks (lower beta and greater safety), quality stocks (high-quality businesses), and cheap or value stocks.

AQR describes these preferences using two factors:

• BAB (Betting Against Beta): a preference for buying low-beta stocks and avoiding high-beta stocks.
• QMJ (Quality Minus Junk): a preference for high-quality companies, with characteristics such as greater profitability, stronger growth, greater financial soundness, and higher dividends or buybacks.

When the researchers control for Berkshire’s exposure to BAB and QMJ, particularly in its public stock portfolio, the public stock portfolio’s alpha falls to a statistically insignificant level of approximately 0.3% annually.

This suggests that a large part of Buffett’s excess return can be understood as holding safe, high-quality, cheaply valued stocks over the long term, then amplifying their returns with leverage.

3. Why This Strategy Can Work

AQR’s theoretical starting point is that leverage constraints and margin requirements alter the structure of risk premiums. When many investors cannot use leverage effectively, they become more willing to buy high-beta assets to obtain higher expected portfolio returns.

As a result, high-risk assets may become overpriced while low-risk assets become undervalued. Investors able to use leverage can then establish positions that buy low-beta assets and sell high-beta assets.

The parts that are truly difficult to replicate are usually obtaining financing of comparable quality and enduring fluctuations over long periods. Selecting low-beta stocks is only part of the process.

AQR concludes by noting that Buffett’s return history also includes difficult stretches. What matters is his ability to stick with the strategy over the long term and survive periods that would force other investors to liquidate.

4. References

1. Andrea Frazzini, David Kabiller, and Lasse Heje Pedersen (2018), Buffett’s Alpha, Financial Analysts Journal 74(4): 35–55; published journal version. The leverage and financing discussion and Table III support 1.7:1 leverage, a 1.72% float cost, the roughly 3-percentage-point T-bill spread, and float’s roughly 35% liability share. The investment-style discussion and Table IV support BAB, QMJ, and the public portfolio’s approximately 0.3% annual alpha. The conclusion discusses persistence and forced-liquidation risk.
2. Financial Analysts Journal (2019), Correction, 75(1): 2–3. Corrects the information-ratio distribution in Table I, Panel B, rather than the Table III and IV results cited here.