Is a fixed position size or a volatility-based one better? We tested both on the same 106 trades in SPY (the S&P 500 ETF). A fixed size puts the same share of the account into every trade. A volatility-based size puts in less when prices swing more.
From $10,000, the fixed size ended at $43,347. The volatility-based size ended at $40,933. Both counts include dividends. Under the volatility-based method, trade sizes ran from 18.2% of the account to 100%.
What is position sizing? Position sizing is the decision of how much of an account goes into one trade. A fixed size uses the same share every time, while a volatility-based size shrinks the position when prices swing more.
The fixed size turned $10,000 into $43,347. The volatility-based size turned it into $40,933.
Volatility-based sizing averaged 85.6% of the account in the calmest group of trades and 25.7% in the most turbulent.
We repeated the test with four settings of the trend rule that decides when to trade. Volatility-based sizing grew faster in three and slower in one, so neither method won consistently.
How does position sizing work?
Position sizing is the choice of how much money goes into one trade. The signal says when to trade. The size says how much.
A fixed size might be 50% of the account every time. A volatility-based size scales the share to recent price swings. Volatility is how far a price moves from day to day. We measure it as the standard deviation of daily returns, which our explainer on standard deviation in investing covers.
A third method sizes each trade by its distance to a stop-loss, the price at which a holder sells to limit a loss. We did not test it.
How did we test fixed against volatility-based sizing?
We wrote the trade rule before looking at any result. The account holds SPY while its close sits above its 200-day simple average. Otherwise it holds cash. Our guide to the 200-day moving average explains the line.
This is a textbook rule, not one of our models.
We read the signal at the close and trade at the next day's close. Each run of days in the market is one trade. The list holds 106 trades across 8,293 daily closes, from 12 November 1993 to 26 October 2026. The last trade is still open, so we value it at that final close.
Of the 106 trades, 47 made money and 59 lost.
The two methods use the same trades and change one thing, the size. The fixed method puts 50% into every trade.
The volatility-based method divides a constant by SPY's volatility on the signal day, then caps the size at 100%. Volatility here is the annualised standard deviation of daily returns over the previous 20 trading days. Annualised means scaled to a one-year figure.
We chose the constant so that the mean size across the 106 trades is also exactly 50%. Both methods then commit the same average amount. Only the spread between trades differs.
This step uses hindsight. We set the constant after seeing all 106 trades, which a trader could not do in advance.
Cash earns nothing in both accounts, with no costs charged. The growth rates are low by design, because the test compares the two methods, not either with SPY itself.
How did the two accounts compare?
The annual rate is the single yearly rate that turns $10,000 into the ending value. Over the 32.9 years, the fixed account grew at 4.55% a year and the volatility-based account at 4.37%. Chart 1 shows both.
Source: YX Insights price data
The volatility-based account led at the end of 1993, 1995, 1996 and 1997. It trailed at the end of 1994, at every year-end from 1998 to 2025 and on 26 October 2026. The widest gap at a year-end came in 2021, when it stood 8.6% below the fixed account.
The volatility-based account moved less from day to day. Its annualised volatility was 5.7%, against 6.1% for the fixed account.
The worst fall is the deepest drop from a previous high in the daily account value, searched over the whole window. For the volatility-based account it was 12.8%, ending on 8 December 1994. For the fixed account it was 12.4%, ending on 13 March 2023.
Volatility measures the typical daily swing. The worst fall measures the single deepest drop. The two can point different ways.
What did volatility-based sizing change trade by trade?
Fifty-six trades came out smaller than the fixed 50% and 50 came out larger.
Chart 2 groups the 106 trades by SPY's volatility on the signal day. In the calmest group (under 12%) the size averaged 85.6% of the account. In the most turbulent (over 30%) it averaged 25.7%.
Source: YX Insights price data
The bars shorten as volatility rises. The dashed line marks the fixed 50% size. Groups with bars past the line traded larger than the fixed size. Groups short of it traded smaller.
The table shows the mean result per trade, which is the position size multiplied by SPY's return over the trade.
SPY volatility at entry | Trades | Mean size, volatility-based | Mean result per trade, fixed | Mean result per trade, volatility-based |
|---|---|---|---|---|
Under 12% | 13 | 85.6% | -0.39% | -0.70% |
12% to 18% | 39 | 57.9% | 2.26% | 2.64% |
18% to 25% | 31 | 40.4% | 2.54% | 2.08% |
25% to 30% | 14 | 32.0% | 1.03% | 0.67% |
Over 30% | 9 | 25.7% | -0.61% | -0.33% |
All trades | 106 | 50.0% | 1.61% | 1.55% |
Result per trade is the share of the account in the trade multiplied by SPY's return over the trade, with dividends counted. The fixed size is 50% in every trade. Source: YX Insights
In the nine trades entered when volatility was over 30%, the fixed size lost 0.61% of the account on average. The volatility-based size lost 0.33%.
In the 13 trades entered when volatility was under 12%, the fixed size lost 0.39% on average. The volatility-based size lost 0.70%. The method that trimmed the turbulent entries enlarged the calm ones.
The worst single trade shows the same effect. The fixed account's worst lost 2.57% of the account.
The volatility-based account's worst lost 3.42%. That was the first trade, entered at the 12 November 1993 close when volatility was 8.5%, so the size hit the 100% cap. SPY fell 3.4% before the trade closed on 30 March 1994.
Does the result hold for other averages and other periods?
We reran both methods with the average set at 100, 150 and 250 days, each time rescaling to a 50% mean size. Chart 3 shows the annual rates.
Source: YX Insights price data
The volatility-based account grew faster in three of the four versions. The fixed account grew faster with the 200-day average.
The method with the smaller worst fall also changed from one version to the next. Volatility-based sizing had the smaller worst fall with the 100-day average (24.0% against 29.5%) and the 150-day average (15.9% against 20.5%). It had the larger worst fall with the 200-day average (12.8% against 12.4%) and the 250-day average (14.7% against 12.1%).
The choice of average moved the worst fall more than the choice of size did.
We tested four averages: 100, 150, 200 and 250 days. Under fixed sizing, the worst fall ranged from 12.1% (250 days) to 29.5% (100 days), a spread of 17.4 percentage points. Switching to volatility-based sizing moved the worst fall by 0.4 to 5.4 percentage points across those same four versions.
We split the 200-day trades in two. In the 60 trades from 1993 to 2008, the fixed account grew at 4.39% a year against 3.84%. In the 45 trades from 2010 to 2026, the figures were 4.73% against 5.31%. One trade sits in neither half.
What does the test leave out?
The test covers one fund, one rule and one window. It makes no forecast.
We set the volatility-based size once at entry. A method that resizes daily would differ.
Our explainer on how to read a backtest covers what a test like this can and cannot show.
Our explainer on what volatility is covers the measure that drove the sizes.
Neither sizing method won consistently. With the 200-day average, the volatility-based account ended 5.6% below the fixed one, at $40,933 against $43,347. With three of the four averages, the volatility-based size grew faster.
Get the next explainer
The Academy is free to read, here on the web. The Long Trend, our free weekly letter, comes by email.
Learn more with YX Insights
This explainer is part of the YX Insights Academy. Each one takes a single idea and checks it against real data.
The same approach runs through everything else we publish:
Systematic Portfolio: ready-made portfolios for Macro & Megacaps and for Commodities. We publish the holdings and every change.
Multi-model Signals: a daily long-or-flat call on every name we cover. Each call shows how strongly our four models agree.
Research: company deep dives, macro commentary and essays on how we test.
Good places to start on the website:
Academy: more explainers like this one.
Start here: get our free weekly letter by email.
Macro & Megacaps Systematic Portfolio and Commodities: the two portfolios in detail.
Track record: live results since 3 August 2026, plus the ten-year backtest.
Upgrade: get the Systematic Portfolio, Multi-model Signals and research in full.
Common questions about position sizing
What is position sizing in trading?
Position sizing is the decision of how much of an account goes into one trade. A fixed size uses the same share every time, while a volatility-based size shrinks when prices swing more. In our test of 106 SPY trades, a fixed 50% size grew $10,000 to $43,347 over 32.9 years.
How do you calculate position size?
The stop-loss method divides the amount at risk by the loss per share at the stop-loss price. With 1% of a $10,000 account at risk, that is $100. If the stop-loss sits $2 below the entry price, the position is 50 shares. Our test set the size from volatility instead.
What is volatility-based position sizing?
Volatility-based position sizing sets each trade's size from recent price swings, so higher volatility gives a smaller position. In our test, the size averaged 85.6% of the account when SPY's volatility was under 12% and 25.7% when it was over 30%. It capped at 100% of the account.
Is volatility-based sizing better than fixed sizing?
Not in this test. The fixed size grew faster in the main version, at 4.55% a year against 4.37%. The volatility-based size grew faster in the other three versions. It also cut the size of the nine trades entered when volatility was over 30%. Those trades lost money on average.
DISCLAIMER: This article is intended for educational purposes only. Any information or analysis in this note does not constitute an offer to sell or a solicitation of an offer to buy any securities. Nothing in this note is intended to be investment advice, nor should it be relied upon to make investment decisions. Any opinions, analyses, or probabilities expressed in this note are those of the author as of the note's date of publication and are subject to change without notice.
