You've watched your balance go up for years. What it pays you once you stop working is a different number, and you don't find out until you pick a rule for taking the money out. On a million dollars over 30 years, the tested rules answer anywhere from $39,000 a year to $57,000 a year. The best retirement withdrawal strategy, according to the two biggest studies on the question, doesn't start with that rate at all.
What the rules pay on a million dollars
Morningstar runs the same test every year. A million dollars, 40% in stocks and 60% in bonds, 30 years, then try the main ways of drawing it down. They score every method on six measures, and those six come down to two questions. How much do you get to spend, and how much is left? No method won all six, which means the best strategy has no answer until you say what you want the money to do.
Take the plain version, where you pull a set amount the first year and give yourself a raise for inflation after that. It leaves the most money behind at year 30 of anything tested and your payment barely moves, but on how much you actually get to spend it finishes near the back of the field. Morningstar's own words: retirees running it may well underspend during their lifetimes. For somebody retiring now they put it at 3.9%, which on a million dollars is $39,000 a year.
That's the careful end. Guardrails start at 5.2%, or $52,000 a year: you set an upper and a lower line around your withdrawal rate, and adjust your spending when the portfolio crosses one. The percentage of balance approach started highest of anything tested at 5.7%, and produced the wildest swings in the study. The RMD method takes a slice of the balance sized to your age, starts at 4.7%, and produces some of the highest lifetime spending, which is the same reason it leaves the least behind. A 30-year TIPS ladder steps outside the market entirely and supported 4.5%.
The bigger numbers cost you in bad years
If 5.7% is sitting right there, why would anybody take 3.9%? Because of what those rules do to you when the market falls. Michael Kitces took Guyton-Klinger, the classic guardrail system, and ran it back through real market history. Retire in 2007, into the financial crisis, and the rules would have cut your real spending by 28%. Start in 1965 and ride the stagflation years and you'd have been cut 54%. That's half your income gone from a rule you picked because it let you start at $52,000. The careful approach really does leave money you could have spent. The alternative asks for a 28% pay cut in your 70s, on a budget that's already set.
The strategy that keeps winning starts with income
Stanford's Center on Longevity ran a bigger version of the same fight with the Society of Actuaries, comparing 292 retirement income strategies, and they named a winner: the spend safely in retirement strategy. The order of it is the whole finding. The first move is to delay Social Security until 70. The second is to take your withdrawals on the RMD formula out of low-cost index funds or target-date funds. That beat 291 other strategies.
The first move is buying yourself more guaranteed income. Waiting to claim from 67 to 70 takes your benefit to 124% of what it would have been, for life, adjusted for inflation every year after that. Morningstar reached the same place from a different direction: in their claiming tests, delaying everything until 70 produced the highest lifetime spending they measured, about $83,000 in the first year against $75,000 for claiming at 67.
It costs something. Bridge those three years out of your portfolio and Morningstar found a median balance about $120,000 lower, with the break-even around age 84. That's their own caution against assuming delay always wins. The order holds. Income decision first, withdrawal rule second.
Why the order matters, in Morningstar's numbers
Same million dollars. You're running guardrails at 5.2%, so in year one you pull $52,000, and say your Social Security is $36,000. Together that's $88,000 to live on. Now the market drops 30%. You already took your $52,000 out, so the drop hits what's left and you land at $663,000.
Your year two withdrawal was supposed to be about $53,000 after inflation. Against $663,000 that's an 8% rate, and it trips the guardrail. The rule forces a 10% cut, down to just under $48,000, and you would feel it. Except that's only the portfolio piece. Your Social Security didn't drop, it rose with inflation to about $36,900, so your year two income is about $84,800 against $88,000. An overall dip of about 3.5%.
Same crash, a 10% cut on paper and a 3.5% dip in real life. The rule is identical, so it isn't what's doing the work. The only difference is that $36,000 of the $88,000 shows up whether the market is up or down, about 40% of the bill covered by something that doesn't care what your shares are worth. The other 60% is the gap, and the gap is the only part the rule ever touched. The more of your bill that arrives without you selling anything, the less any rule can do to you.
Real retirees barely touch the money
David Blanchett and Michael Finke went through the Health and Retirement Study, about 20,000 Americans over 50. Income that arrives on a schedule, Social Security, a pension, an annuity, they spend about 80% of it. Out of savings, a 65-year-old couple pulls about 2% a year. The whole field is arguing over 3.9 versus 4.7 versus 5.7, and real retirees take 2%. Selling is a decision you have to make over and over, and every time you make it the market hands you a reason to wait. A payment that lands in your account works like a paycheck, and a paycheck feels safer to spend.
Shrink the gap while you're still working
Take some of what the growth years built and turn it into payments that arrive on their own. The guaranteed ones first, so Social Security, which you already own, and a pension if you've got one. After that, CDs, high-yield savings, bonds and TIPS ladders, dividends, rentals, private mortgage notes.
Here's where I sit, and it's one reason I think about retirement this way. I left the Air Force at 33 because rent from my properties covered what I spent every month. My gap was zero, and that's what let me leave my job early. The head start was real. I started buying properties at 25 with VA loans while I was single, and that's not the same as being 48 with a mortgage, a 401k, and kids to take care of. What transfers is getting the monthly gap to zero.
There's a cost. When you move money out of growth you give it up, and across 30 years that's real money. Tilting toward dividends can concentrate you into fewer companies. Like every investment, each of these carries risk: dividends can get cut, rentals can sit empty, and a private mortgage note is only as good as the person paying it. That same study found people spend under half of their dividends and rent, against 80% of the guaranteed kind.
So add up the income that already arrives without you selling anything, work out what a year of your life costs, and take one away from the other. Whatever's left is your gap, and that gap is the only number a withdrawal rule is responsible for. Shrink it with income that comes in whether you're working or not, then run whatever rule you like on the rest.