We Ran the 60/40 Through a Full Risk Teardown
The most recognized portfolio in retail investing — 60% stocks, 40% bonds — put through a backtest, a risk decomposition, and a correlation matrix together. What the historical data shows about where its return, its drawdowns, and its risk actually came from.
The 60/40 — 60% stocks, 40% bonds — is the most recognized portfolio in retail investing, the default that a generation of target-date funds and robo-advisers was built around. It is also easy to talk about and hard to actually look inside. So we pointed three of the tools at it at once: a backtest, a risk decomposition, and a correlation matrix. This is a teardown of one specific, well-known mix over one window — a look at what the historical data shows, not a recommendation to hold it. Every number below comes from the same engines the product runs for any portfolio you give it.
The worked example is a plain VTI 60%, BND 40% book (US total-market stocks and US total-market bonds), rebalanced annually, measured against 100% equity (SPY) over the 10-year window (2016–2026) (as of August 15, 2026). Naming the mix and the yardstick explicitly matters: everything here is relative to them.
The backtest: less return, much less pain
Over the 10-year window (2016–2026) the 60/40 portfolio returned 186.7%, against 316.3% for 100% equity. Holding 40% bonds gave up return in a decade that was kind to stocks — that part is unsurprising. The more useful number is the ride: the 60/40's worst peak-to-trough decline was 24.5%, versus 33.7% for 100% equity, and its Sharpe ratio (return per unit of volatility) was 0.64 against 0.72. Whether that trade — less return for a shallower drawdown — is one you'd want is a judgment the data can inform but not make.
| Metric | 60/40 portfolio | 100% equity (SPY) |
|---|---|---|
| Total return | 186.7% | 316.3% |
| Worst drawdown | 24.5% | 33.7% |
| Sharpe ratio | 0.64 | 0.72 |
What the 40 did in the last three drawdowns
The point of the bond sleeve is what it does when equities fall. The underwater curve below traces the 60/40's decline from its running peak across the three drawdowns in this window — the late-2018 selloff, the 2020 crash, and the 2022 stock-and-bond decline. In 2018 and 2020 the bonds cushioned the fall, which is why the portfolio's worst drawdown (24.5%) is far shallower than 100% equity's (33.7%). 2022 is the exception the teardown makes visible: bonds and stocks fell together, and the '40' did much less cushioning than its reputation promises. The chart shows all three episodes on one axis.
What you're actually holding: 60% of the weight, most of the risk
A 60/40 split describes the money, not the risk. Stocks are several times more volatile than bonds, so the equity sleeve drives the portfolio's ups and downs far out of proportion to its 60% weight. The exposure engine makes this explicit by decomposing total portfolio risk by holding: VTI is 60% of the weight but 93% of the risk; BND is 40% of the weight and only 7% of the risk. On this measure the portfolio behaves like a much more equity-heavy book than '60/40' suggests — its weighted-average beta to the market is 0.65, and its effective number of independent bets (1.9, from a concentration index of 0.52) is closer to one than to two.
| Holding | Share of weight | Share of risk |
|---|---|---|
| VTI (US stocks) | 60% | 93% |
| BND (US bonds) | 40% | 7% |
Why it usually works: the two sleeves don't move together
The reason the 40 helps at all is correlation — how much stocks and bonds move together day to day. Over this window their correlation was 0.16. A number near zero (or below it) is what lets bonds rise, or fall less, when stocks drop, and it is the entire mechanical basis for the shallower drawdown. It is also not a constant: the 2022 episode above is what a stretch of positive stock-bond correlation looks like from the inside, which is why a single historical coefficient is a description of the past, not a guarantee about the next decline.
See how correlated your own holdings are
Correlation and concentration — two of the ideas in this teardown — are what the free Portfolio X-ray runs live. Pick up to three well-known names and it shows how correlated they are, how concentrated the mix is, and how many independent bets you really have, using the same underwater view shown above. It runs equal-weight on a curated set of names, so treat it as a quick diversification check rather than the whole teardown: the custom weights of a 60/40, the per-holding risk-share decomposition, and the historical backtest live in the full app.
Method & caveats
All figures are for the VTI 60%, BND 40% book, rebalanced annually, over the 10-year window (2016–2026) as of August 15, 2026; a later re-run rolls the window forward and shifts every number. Returns use adjusted close (dividends and splits reinvested) and are gross — no fees, trading costs, slippage, taxes or idle-cash yield are modeled, so a real account would differ. The risk decomposition is computed on the exposure engine's covariance basis; correlation is computed on adjusted-close daily returns. This is one illustrative portfolio over one window against one benchmark — an illustration of what the historical data shows, not a general claim or a recommendation, and past results do not predict future returns.