The standard advice on debt payoff versus investing goes like this: look at your debt's interest rate, look at the return you'd reasonably expect from investing instead, and send your extra dollars wherever the number is bigger. A 3% mortgage loses to a stock market averaging 7-10% a year, so you invest. A 22% credit card balance beats almost anything you could earn in a brokerage account, so you pay it off. It shows up in standard personal-finance advice: risk tolerance is presented as one factor to weigh alongside the rate comparison itself. It's not bad advice. It's an incomplete one.
The rate comparison is, at bottom, an opportunity-cost calculation, and it's built on a hidden assumption worth naming out loud: it treats the two numbers as if they're the same kind of number. They aren't. Paying down a debt at a fixed rate is a guaranteed return — you know exactly what you're getting, because the alternative was paying that same rate in interest indefinitely. Investing at an "expected" 7-10% is not guaranteed at all; it's a long-run average built from decades that included both extraordinary gains and multi-year stretches of losses. Comparing a certain number to an average of an uncertain range isn't apples to apples, it's apples to a probability distribution.
The honest risk-free comparison isn't the stock market's long-run average — it's the return on the safest instrument actually available to you: the yield on the 10-year Treasury. At the time of writing, in mid-2026, that yield sat at 4.57%. The number itself will move, and it will be a different number by the time you're reading this — but the principle underneath it doesn't move: compare your debt's rate to the real risk-free rate available to you on the day you're deciding, not to an optimistic equity forecast. A debt sitting above that risk-free rate isn't just "probably" a good payoff target — it's a guaranteed return that beats the safest asset in the market, with zero variance. That's a much stronger case for payoff than comparing it against a long-run stock average.
Follow that logic and an obvious objection shows up: if most mortgages and plenty of auto loans sit below the Treasury yield, does the same reasoning tell you to stop investing and just buy Treasuries instead of stocks? No — the Treasury comparison is a floor, not a ceiling. It marks the minimum a debt's rate has to clear before payoff becomes the arithmetically safer move, not the bar your investment returns need to hit. Below that floor, investing still has the stronger case, because the extra expected return over Treasuries is compensation for volatility you're choosing to accept. Above it, paying off the debt stops being a bet on the market underperforming and becomes plain arithmetic. That's the real dividing line: a 3% mortgage and an 8% personal loan aren't the same decision wearing different numbers, and the honest version of this comparison already tells you which side of the line each one sits on.
One adjustment sits above this whole comparison and overrides it: an employer 401(k) match. A dollar that becomes two dollars is a guaranteed, unbeatable return even after accounting for the vesting schedule many employers attach to matched funds, and it gets funded before extra debt payoff or extra investing in nearly all cases — no Treasury yield and no double-digit credit card rate beats a 100% return, even factoring in the vesting period many plans attach to it.
Everything above assumes you're still accumulating — still earning a paycheck, still years away from needing the money. That assumption breaks the moment you retire early. While you're accumulating, a bad market year is something you can wait out, because your income keeps buying shares at the lower price and time keeps working in your favor. In drawdown, a bad market year forces you to sell shares to cover expenses regardless of the price, and there's no paycheck to replace what you sold — that's sequence-of-returns risk, and it does its worst damage in the years right after you stop working. That changes what "long-run average" even means for you. A 30-year-old comparing a loan to an expected equity return has forty-plus years for bad decades to wash out. A 45-year-old fifteen years into a planned 40-year retirement doesn't have that runway the same way — money already earmarked for near-term withdrawals isn't sitting in a long-run average at all, it's exposed to whatever the market happens to do in the specific years you need to spend it. For that person, paying off a debt at even a moderate rate can be worth more than the number implies, because it lowers the fixed costs a bad sequence has to cover — and reducing what you're forced to withdraw in a down year does more for a portfolio's survival than chasing an uncertain extra point of return.
The practical test, then, isn't whether your debt's rate looks "good" in the abstract. It's how many years that money will actually sit exposed to the market before you need it, and how much room you have if those years turn out to be the bad stretch instead of the average one. Someone with decades of future contributions ahead can afford to lean on the long-run average — the bad years have time to be absorbed. Someone five years from retirement, or five years into one, is playing on a much shorter clock, and on that clock the guaranteed side of the comparison gets more valuable at exactly the moment the uncertain side gets more dangerous.