The rules
Each month, rank stocks by volatility and hold the least volatile group. Ang, Hodrick, Xing and Zhang (2006) sort NYSE, AMEX and Nasdaq stocks into value-weighted fifths (each stock weighted by its market value) on the past month's daily-return volatility and hold for one month. Blitz and van Vliet (2007) use equal-weighted tenths of the world's large companies, ranked on three-year weekly volatility. Baker, Bradley and Wurgler (2011) use cap-weighted fifths ranked on up to five years of monthly returns.
- Universe
- US common stocks via the Market Screener: the 30 with the lowest one-month average daily range of at least 0.8%
- Bars
- daily
- Position size
- 3.2% of equity per stock, at most 30 held. Published: value-weighted fifths (Ang et al.), equal-weighted tenths (Blitz and van Vliet) or cap-weighted fifths (Baker et al.).
- Rebalance
- Daily check; each stock held 21 trading days, so the book turns over mostly in batches about a month apart. Published: monthly rebalance.
What the research found
Ang, Hodrick, Xing and Zhang (2006), over 1963-2000, found that, sorted on total volatility, the least volatile fifth returned 1.06% a month, the middle fifth 1.22% and the most volatile 0.09%. Adjusted for Fama and French's market, size and value factors, the most volatile fifth trailed the least volatile by 1.19% a month (t = -5.92; a t-statistic beyond about 2 either way marks a result unlikely to be chance), and by 1.31% (t = -7.00) when sorted on idiosyncratic volatility, the part those factors leave unexplained. Their 2009 follow-up found the most-minus-least spread on idiosyncratic volatility, after adjusting for world market, size and value factors, at -1.31% a month across 23 developed markets (-0.67% without the US), and the effect significant in every G7 country. Blitz and van Vliet (2007), on the world's large companies over 1985-2006, found the least volatile tenth had a Sharpe ratio (return over cash per unit of volatility) of 0.72 against 0.40 for the universe, and 0.58 against 0.47 in the US, where it earned 6.9% a year over the risk-free rate against the universe's 8.1%, with an alpha (return not explained by market exposure) of 3.3% (t = 1.6). Baker, Bradley and Wurgler (2011), on all US stocks over 1968-2008, found the least volatile fifth beat the market by 1.05% a year (0.62% among the 1,000 largest), with alphas of about 2% a year (t = 2.44, 2.03). Blitz, van Vliet and Baltussen (2020) find that a beta-neutral long-short volatility factor in US stocks, its low-volatility side levered up and its high-volatility side levered down (calm stocks bought with borrowed money on top, volatile ones sold short, for no net market exposure), earned a positive premium in every decade since 1940. Hou, Xue and Zhang (2020) find that idiosyncratic and total volatility fail once the cutoffs for their ten volatility groups come from NYSE stocks alone and returns are weighted by market value (total-volatility spread: -0.39% a month, t = -1.18). Fama and French (2016) find that low-volatility stocks' returns move like those of profitable firms that invest conservatively, which accounts for their high average returns. Jensen, Kelly and Pedersen (2023) find that the low-risk theme replicates. None of these studies deducts trading costs. In these studies the edge is mainly risk-adjusted: a long-only low-volatility book's raw return over the market is small.
Results
| Measure | Orbital test | SPY, same days |
|---|---|---|
| Window | Mar 1, 2017 to Oct 30, 2026 | Mar 1, 2017 to Oct 30, 2026 |
| Total return | +107.64% | +267.14% |
| CAGR | +7.85% | +14.40% |
| Sharpe | 0.36 | 0.60 |
| Discounted Sharpe | 0.06 | — |
| Max drawdown | −31.2% | −33.7% |
| Win rate | 58.9% | — |
| Profit factor | 1.23 | — |
| Trades | 3,502 | — |
Hypothetical: a backtest with modeled costs, not traded with real money.
Hypothetical growth of $10,000
R04 Low-Volatility Stocks · +107.6% vs SPY +267.1%
Rebuild it in Orbital
Paste this into Orbital Chat. The Strategy Developer builds the workflow and asks before it runs any backtest.
On daily bars, use a Market Screener universe of US stocks with the type set to Common stock only, priced at $10 or more, with 20-day average volume of at least 1,000,000 shares and Volatility 1M (average daily range) of at least 0.8%, sorted by Volatility 1M (lowest first), top 30. Buy each stock in the screen with 3.2% of equity at the next open and sell it 21 trading days after entry. Hold at most 30 positions, filling free slots in screen order. Long only; uninvested capital stays in cash. Backtest from March 1, 2017 to October 30, 2026 with $100,000.
Is it real?
- Strategies tested together
- 7
- Beta to SPY
- 0.47
- Alpha vs SPY
- −1.1% a year
- Costs paid
- $16,612, 13% of the gain before costs
- Data
- Daily bars, 2,432 sessions
- Fills
- Next open at market, sized at the fill
Read this before you trust it
The ranking is Orbital's Volatility 1M: the 21-trading-day average of each day's high-to-low range as a percent of that day's close. The papers use the standard deviation of returns; Orbital's measure is total, not idiosyncratic, volatility, with no three- or five-year window as in Blitz and van Vliet or Baker, Bradley and Wurgler. A $10 price floor and a 1,000,000-share average volume floor stand in for size. The 0.8% volatility floor, an addition, keeps out most pending cash takeover targets, but it can still admit some and can leave out the calmest stocks in quiet years. It buys from the 30 least volatile stocks that pass these floors, in equal 3.2% slots, where the papers hold value-weighted fifths (Ang et al.), cap-weighted fifths (Baker et al.) or equal-weighted tenths (Blitz and van Vliet). It does not sell short the most volatile stocks, where most of Ang, Hodrick, Xing and Zhang's effect sits. Each stock is sold after 21 trading days, and at that open its slot can go only to a screen stock the book did not hold at the previous close, so a stock still in the screen is bought back a day later at the earliest and its slot can sit in cash meanwhile. Orders fill at the next open with 0.05% slippage (the assumed trading cost).
Sources