The Latte Factor, Recalculated

The latte factor has been argued the same way for twenty years. One side says a daily coffee costs you a million dollars in retirement. The other says that is nonsense and you should look at your rent. Both sides are arguing about the same two options: buy the coffee, or do without it. There is a third option, and it changes the arithmetic more than either camp’s talking points do.

The short version

The million-dollar figures come from return assumptions of 10% to 12% with no inflation adjustment. At a defensible real return the number is large but not life-changing. And the debate is posed as coffee against no coffee, when for a daily cafe habit most of the money is available by brewing the same drink at home. Fund the plan first, then decide how you want to drink your coffee.

Where the million-dollar number comes from

David Bach popularized the idea, and his own site is specific about the math: “Starting with $5 a day may not seem like a lot, but if you combine that savings with 10 percent annual interest, you’d have $948,611 40 years later.”¹ Suze Orman put it more sharply in 2019, telling CNBC that buying coffee out meant “you are peeing $1 million down the drain as you are drinking that coffee.” That calculation assumed a 12% annual return.²

Both figures are nominal, before taxes and fees, and before inflation. A 12% assumption drew a blunt assessment from David Blanchett of PGIM DC Solutions, who called it “absolutely nuts” in later CNBC coverage.² It is worth being fair to Bach here. He has consistently said the coffee is a metaphor and that the savings rate is the number that matters. The headline figure is what travelled, and the headline figure rests on the return assumption rather than on anything about coffee.

What five dollars a day becomes at a real return

Run the same contribution at real returns instead, with the contribution rising alongside inflation so every figure lands in today’s dollars.

Real return	After 30 years	After 40 years
3%	~$88,000	~$139,000
5%	~$124,000	~$225,000
7%	~$178,000	~$376,000

$5 a day invested monthly at a constant real return, contributions indexed to inflation.

At a 5% real return, forty years of $5 a day is roughly $225,000 in today’s money. That is a serious sum and it is a long way from a million. The gap between the two framings is entirely the return assumption and the inflation adjustment, which is why the argument keeps producing numbers that feel unfalsifiable.

The price is no longer five dollars

The other reason the classic figures are stale is the price itself. A cafe latte averaged $5.54 in the second quarter of 2026, across 7,479 observed menu prices collected by the point-of-sale platform Joe.³ Treat that as industry data rather than an official statistic: the Bureau of Labor Statistics publishes no index for coffee bought away from home, so cafe drinks are folded anonymously into a broad restaurant category.

Seven cups a week at that price is about $2,017 a year. Five days a week is $1,440. Those are the numbers worth putting into a calculator, and they run 10% to 30% above the figures that circulate.

The option the argument leaves out

Here is the part both camps skip. Coffee is not only a cafe purchase. The same drink can be made at home, and after the recent run-up in bean prices a home cup still runs about $1.50 all in, counting beans, a filter, and a modest kit amortized over five years of daily brewing.⁴

At seven cups a week, the cafe version costs $2,017 a year and the home version costs $546. Switching frees about $1,471 a year. Giving up coffee altogether frees the full $2,017. So on those prices the switch captures about 73% of the money the argument is about, while you carry on drinking coffee every morning.

Invested at a 5% real return over forty years, that $1,471 a year becomes roughly $182,000 in today’s dollars. Going without coffee entirely gets you to about $249,000. The extra $67,000 is real, and it is what abstinence buys you over substitution across four decades. Whether that trade is worth it is a question about your life rather than about arithmetic, which is the point at which a spreadsheet should stop talking.

Those figures depend entirely on the two prices and the return you assume. Put your own in.

Where the household budget goes

None of this makes coffee a large line item. The Bureau of Labor Statistics puts average annual spending per consumer unit at $78,535 in 2024, of which housing is $26,266 and transportation is $13,318. Those two together are just over half of all spending. Food away from home, which includes every restaurant meal and not only coffee, is $3,945.⁵ A consumer unit is not the same thing as a household, and these are means rather than medians, so a few high spenders pull them upward.

Our guide on growing income faster than expenses works through why the gap between those two lines dominates any single category. Nothing here disturbs that. Coffee is a small lever whether or not you pull it well; this guide is about the fact that the lever has three positions rather than two.

Why the framing persists

James Choi surveyed the 50 most popular personal finance books against what economists teach, and found a consistent split on saving. “In contrast to the emphasis on smoothing consumption in economic theory, popular authors advise smoothing savings rates.”⁶ Lifecycle theory says saving should track where you are in your earning life. Popular advice says start now and keep the percentage fixed.

Choi is not simply scoring the professors as correct. He notes that popular advice “tries to take into account the limited willpower individuals have to stick to a financial plan, and its recommended actions are often easily computable by ordinary individuals.”⁶ That is the honest defense of the latte factor as a device. Telling someone to optimize lifetime consumption is correct and useless. Telling them to redirect one visible daily purchase into an automatic transfer is approximate and gets done.

The substitution framing keeps that behavioral advantage. It is one concrete change, it is easy to compute, and it does not depend on anybody’s willpower holding for forty years, because you are not asking them to give anything up.

Frequently asked questions

Is a daily coffee really costing me a million dollars?

No. Figures near a million come from 10% to 12% nominal returns with no inflation adjustment. At a 5% real return, $5 a day for forty years is around $225,000 in today’s dollars. Large, and a very different claim.

Should I stop buying coffee?

That is a preference question, not a math question. If the cafe visit is worth its price to you, buy it. If you are buying it on autopilot and would be just as happy with a cup you made, the substitution is the cheapest change available, because it captures most of the money without removing anything you enjoy.

Does brewing at home actually save that much?

At a $5.54 cafe cup against roughly $1.50 at home, seven cups a week frees about $1,471 a year. Bean prices rose sharply through 2025 and 2026, so the ratio between the two is narrower than it used to be, though the dollar difference per cup has barely moved.

Where does this sit against housing and transportation?

Well below them. Housing and transportation are about half of average spending; food away from home in total is under $4,000 a year. Fix the large recurring costs first. Coffee is worth doing after those, not instead of them.

What return should I assume?

Use a real return, meaning after inflation, and keep it conservative. The table above spans 3% to 7% because the honest answer is a range. Any projection that needs 10% or more to sound impressive is telling you about the assumption rather than about the spending.

Key takeaways

The million-dollar figures are assumption-driven. They require 10% to 12% nominal returns with no inflation adjustment.
At a defensible real return the number is smaller. Around $225,000 in today’s dollars for $5 a day over forty years at 5% real.
The debate skips substitution. For a daily cafe habit at $5.54 against a $1.50 home cup, brewing at home captures most of the money at stake.
Coffee is still a small lever. Housing and transportation are about half of average household spending.
Automate the difference or it does not exist. Money that stays in checking has not been saved, whichever option you pick.

Related guides

Grow Income Faster Than Expenses the case that the gap between income and spending dominates any single category, including this one.
Should Med, Law, and MBA Students Save Money? the lifecycle argument in the case where it bites hardest, with the same Choi paper doing the corrective work.
Savings Rate vs. Investment Returns which of the two levers moves your outcome more, and when that answer flips.

Sources

Bach, D. The Latte Factor: Why I Wrote This Book Now. Source of the $5 a day, 10% annual return, $948,611 over 40 years figure, and of the position that the savings rate is what matters.
CNBC Make It (March 28, 2019). Suze Orman: Spending money on coffee is like throwing $1 million down the drain. The underlying calculation assumes a 12% annual return. David Blanchett’s “absolutely nuts” assessment appears in CNBC’s later coverage of the same claim.
Joe. State of Coffee, Q2 2026. National average latte menu price of $5.54 across 7,479 observed prices. Industry data from a point-of-sale platform; no government series for cafe coffee exists.
BeanBench. Why Coffee Got Expensive, and What a Home Cup Costs Now. Per-cup cost of home brewing at current bean prices, and the change in the home versus cafe ratio since 2024. Sister site, same team, no affiliate relationship.
U.S. Bureau of Labor Statistics. Consumer Expenditures, 2024. Average annual expenditures per consumer unit $78,535; housing $26,266; transportation $13,318; food away from home $3,945. Means, not medians.
Choi, J. J. (2022). Popular Personal Financial Advice versus the Professors. Journal of Economic Perspectives 36(4), 167–192. On savings-rate smoothing versus consumption smoothing, and on limited willpower and computability.