The Trinity Study: The Paper That Taught America How to Retire

Every retirement plan you've ever seen — every FIRE calculator, every "25x your expenses" rule of thumb, every safe withdrawal rate chart on a financial advisor's desk — traces back to three finance professors in San Antonio, Texas, and a 1998 paper that wasn't supposed to change the world.

Philip Cooley, Carl Hubbard, and Daniel Walz. If you don't recognize those names, you're not alone. They weren't celebrity economists. They weren't hedge fund managers. They weren't the kind of people who give keynote speeches at CNBC conferences. They were professors at Trinity University — a small liberal arts college with an enrollment of about 2,500 — who set out to answer a question so obvious that nobody had bothered to answer it properly: If you retire with a portfolio of stocks and bonds, how much can you safely spend each year without running out of money?

Their answer became the most cited number in personal finance. And the story of how they got there is worth understanding — because it reveals how much of the conventional retirement wisdom you've absorbed is built on their specific assumptions, their specific data, and their specific definition of "success."

The Question Nobody Had Answered

By the mid-1990s, the financial planning industry had a problem. Advisors knew how to help clients accumulate wealth: save early, diversify, keep costs low, let compounding work. But the distribution phase — turning a lump sum into a lifetime income stream — was a black box.

Annuities existed, but they were expensive and opaque. Systematic withdrawal plans existed, but nobody could say with any rigor what withdrawal rate was safe. Advisors used rules of thumb: 5%? 6%? Whatever kept the client happy? The academic literature was thin. Retirement income was the neglected half of personal finance.

Then Bill Bengen, a financial planner in southern California, published his 1994 paper "Determining Withdrawal Rates Using Historical Data." Bengen's method was straightforward: simulate a 30-year retirement using every rolling period in US market history back to 1926, and find the highest initial withdrawal rate that never failed. His answer — roughly 4% — was a genuine breakthrough. For the first time, there was a data-driven answer to the question of how much to spend in retirement.

But Bengen was a practitioner, not an academic, and his paper appeared in the Journal of Financial Planning — a trade publication. It didn't carry the weight of peer review. The financial industry needed a study with academic credibility, broader scope, and a more systematic approach.

Enter Trinity University.

What the Trinity Study Actually Did

Cooley, Hubbard, and Walz published "Retirement Savings: Choosing a Withdrawal Rate That Is Sustainable" in the AAII Journal in February 1998. (AAII is the American Association of Individual Investors, a nonprofit that publishes practitioner-oriented research.) Their methodology was similar to Bengen's but more ambitious. They tested:

• Withdrawal rates from 3% to 12% (in 1% increments), adjusted annually for inflation
• Payout periods of 15, 20, 25, and 30 years
• Asset allocations from 0% stocks / 100% bonds to 100% stocks / 0% bonds (in 25% increments)

For each combination, they simulated every rolling period between 1926 and 1995 using the Ibbotson Associates dataset — the standard historical record of US stock and bond returns. A "success" meant the portfolio had at least one dollar remaining at the end of the payout period. A "failure" meant the portfolio hit zero before the period ended.

The results were striking — and more nuanced than the soundbite that emerged.

The Core Finding: 4% Works (Mostly)

For a 30-year payout period with a portfolio of at least 50% stocks, withdrawal rates of 3% and 4% had a 100% historical success rate. The portfolio survived every 30-year window in modern US history, including retirements that began in 1929, 1937, 1966, and 1973 — the worst starting years on record.

At 5%, success rates fell to roughly 85% for a 50/50 portfolio over 30 years. At 6%, they dropped below 70%. By 7%, the success rate was a coin flip. And at 8% and above, failure was virtually guaranteed for 30-year retirements.

The study's data tables — which you can still find online and which remain remarkably readable — showed something important that gets lost in the "4% rule" shorthand: the relationship between withdrawal rate and success probability is not linear. Going from 3% to 4% costs almost nothing in terms of safety. Going from 4% to 5% costs a lot. Going from 5% to 6% costs nearly everything. The cliff is steep, and it starts around 4%.

The Allocation Lesson

The Trinity Study also provided clear guidance on asset allocation. Portfolios with 0% stocks (100% bonds) had dramatically lower success rates — a 4% withdrawal rate failed in roughly half of all 30-year periods. Portfolios with 25% stocks fared better but still had failures at 4%. Once you reached 50% stocks and above, 4% was consistently safe, and success rates leveled off — there was little additional safety benefit from going to 75% or 100% stocks.

This was the origin of the advice that retirees need meaningful equity exposure. The old wisdom — "put everything in bonds when you retire" — was mathematically wrong. Bonds alone, with their lower expected returns, could not sustain a 4% inflation-adjusted withdrawal over three decades. You needed stocks to generate the returns that offset withdrawals and inflation.

Shorter Horizons, Higher Rates

The study's results for shorter payout periods were less widely reported but equally important. For a 15-year horizon, withdrawal rates of 7% to 8% had high success rates with an equity-heavy portfolio. For 20 years, 5% to 6% was sustainable. This mattered for people retiring later in life, where a 30-year horizon was unnecessarily conservative. The study implicitly endorsed higher withdrawal rates for shorter retirements — a nuance that the "4% rule" flattened into a one-size-fits-all number.

How "4%" Became the Number

The Trinity Study didn't invent 4%. Bengen found it first. But the Trinity Study gave 4% its academic legitimacy. By 2000, the phrase "the 4% rule" had entered the financial advisor lexicon. By 2010, it had become the organizing principle of the FIRE (Financial Independence, Retire Early) movement. By 2020, it was one of the few pieces of financial advice that people across the ideological spectrum — from Dave Ramsey fans to Bogleheads forum moderators — agreed on.

The appeal is obvious: one number, one formula, no complexity. Save 25 times your annual expenses. Withdraw 4% in year one. Adjust for inflation. You're done. It's the closest thing personal finance has to an elegant equation.

But the simplicity that made the rule popular also made it widely misunderstood.

What Everyone Gets Wrong About the Trinity Study

The study's findings have been misinterpreted in consistent, predictable ways that matter enormously for retirement planning:

Success Means Not Zero

The Trinity Study defined success as "the portfolio did not reach zero before the end of the payout period." It did not mean the portfolio maintained its value. It did not mean the retiree had the same inflation-adjusted wealth at the end of 30 years that they started with. It meant the portfolio didn't hit zero — even if the ending balance was one dollar on day 30, year 1. Many of the historical scenarios where 4% "succeeded" involved the portfolio drawing down significantly. In the 1966 retirement scenario — the worst in modern history — the portfolio survived 30 years but was severely depleted by the early 1980s and only saved by the bull market that began in 1982.

US History Is the Only Dataset

The study's data universe is US stock and bond returns from 1926 to 1995. This is a sample of one country during one of the most prosperous centuries in human history. The United States was not invaded, did not experience hyperinflation, did not have its stock market nationalized, and ended the century as the world's dominant economic power. These are the best-case outcomes. The Trinity Study's success rates would look very different using data from Japan (which has still not recovered from its 1989 peak), Germany (which experienced two world wars and a hyperinflationary episode), or the UK (which saw its stock market severely disrupted by nationalizations in the post-war period).

Inflation-Adjusted Means Inflation-Adjusted

The 4% rule says you withdraw 4% of the initial portfolio balance in year one, then increase that dollar amount by inflation each year. You do not recalculate 4% of the current portfolio value each year. This is the single most common mistake. If your $1 million portfolio drops to $700,000 in a bear market, you still withdraw the inflation-adjusted dollar amount (say $43,000 in year three with 3% inflation), not 4% of $700,000 ($28,000). The rule's safety depends on this rigidity — the whole point is that the good years compensate for the bad years. If you reduce spending when the market is down, you're not following the 4% rule. If you withdraw 4% of the current balance, you're not following the 4% rule. Most people think they're following the 4% rule and aren't.

Fees Are Not Included

The study assumes gross market returns — what the index delivered before any costs. A 1% advisory fee, common in the traditional wealth management industry, turns a 4% withdrawal into a 5% withdrawal from the portfolio's perspective (4% for you, 1% for your advisor = 5% total drain). At 5%, the Trinity Study's success rate drops to roughly 85% for a 30-year period. At the 2% total fee burden that's not unheard of in the advisory world (1% AUM fee plus 1% in fund expense ratios), you're effectively withdrawing 6%, where success rates fall below 70%. This is why self-managing your money with low-cost index funds isn't just about saving fees — it changes what withdrawal rate is sustainable.

The 2011 Update

In 2011, the original authors — all three of them — updated the study with data through 2009, incorporating the dot-com crash and the Global Financial Crisis into the historical record. The findings were published in the Journal of Financial Planning under the title "Portfolio Success Rates: Where to Draw the Line."

The update confirmed the original conclusions with one important refinement: adding international stocks modestly improved success rates, though the effect was smaller than many diversification advocates had hoped. A globally diversified portfolio with a 4% withdrawal rate had a slightly higher success rate than a US-only portfolio, but the difference was marginal — a few percentage points. The core finding held: 4% was still the number, and meaningful equity exposure (50% or more) was still required.

The update also addressed a question the original didn't: what about retirees who want to leave a bequest? The authors found that if the goal is to preserve inflation-adjusted principal rather than merely avoid running out of money, the safe withdrawal rate drops meaningfully — to about 2.5% to 3%. This is a critical distinction. The 4% rule is a depletion strategy, not a preservation strategy. It answers the question "how much can I spend without going broke?" not "how much can I spend and still leave my kids the same inflation-adjusted amount I started with?"

What the Trinity Study Never Said

The study never said 4% is optimal. It said 4% is safe, based on the worst historical case. In most historical periods, retirees could have withdrawn 5%, 6%, or more and ended with more money than they started. Four percent is a floor — the number that survived the Great Depression, World War II, stagflation, and every bear market — not a carefully calculated optimum.

The study never said you should blindly withdraw 4% regardless of circumstances. The authors were careful to describe their work as a tool for planning, not a prescription for behavior. Real retirees adjust. They spend less in down years. They take more when markets cooperate. They have Social Security and maybe a pension. They earn income from part-time work. The rigid 4% + inflation formula was a stress test, not a retirement plan.

The study never said 4% works forever. The maximum payout period tested was 30 years. For a 40-year retirement (retire at 50, live to 90), the historical success rate at 4% is lower — maybe 85% to 90%. For a 50-year retirement, it's worse — perhaps 55% to 65%. The Trinity Study was not designed for the FIRE movement's 50+ year horizons. Early retirees need lower withdrawal rates, dynamic strategies, or both.

Why the Trinity Study Still Matters

It would be easy to dismiss the Trinity Study as outdated — a 1998 paper using US-only data, testing withdrawal rates that may not survive forward-looking capital market assumptions. And yet, nearly 30 years later, its core method — rolling historical periods, variable withdrawal rates, multiple asset allocations, probabilistic success rates — remains the standard for retirement income research.

Every major study that followed — Morningstar's annual safe withdrawal rate research, Wade Pfau's Monte Carlo analyses, Big ERN's exhaustive Safe Withdrawal Rate series on Early Retirement Now — is essentially the Trinity Study with better data, more sophisticated modeling, and additional variables. The framework survived because it's good. Not perfect, but good enough.

For individual investors, the lesson of the Trinity Study isn't really about 4%. It's about the value of asking a specific, testable question — "what withdrawal rate would have survived every historical scenario?" — and following the data to an honest answer. The study gave us the number. More importantly, it gave us the method.

If you want to test your own retirement plan against history, we built a Monte Carlo simulator and a retirement withdrawal calculator that extend the Trinity Study's logic into forward-looking analysis with variable spending, Social Security, and other real-world considerations. The tool is free, like everything else we build.

The Trinity Study asked a simple question and answered it with data. That tradition — rigor over intuition, evidence over narrative — is what we're trying to continue at Alistair. Three professors in San Antonio did it first. The least we can do is get the details right.