The market doesn't care when you retire. You should.

Hi {{first_name|Investor}} -

A reader sent me a DM this week with a question I've been turning over for days:

"I'm 56 and hoping to wrap things up around 63. My 401(k) looks better than I ever expected, but every time the market wobbles I start sweating. How do I protect what I've built without giving up the growth I still need for the next 30 years?"

I get the shorter version of this question constantly from readers a little further along the path — "I just stopped working six months ago and I'm watching my balance every morning," or "I'm three years in, the easy years are behind me, and I can't tell if I should change anything." Same worry, different stage.

The technical term for it is sequence of returns risk — the unfortunate math that says when your bad years happen matters more than how many bad years you have. Two retirees with the same average return over 30 years can end up in completely different places, just because one of them caught a 30% drop in year two and the other caught it in year twenty-two.

The first one runs out of money. The second one dies wealthy. Same average return.

This issue is a set of questions to ask yourself if you're inside what I think of as the fragile decade — the five years before retirement and the five years after. It's also a quiet reframe of how long that retirement might actually need to fund.

One note: I'm writing from a US point of view. The specific tools below are US-flavored; the questions are not.

When you're accumulating, a market crash is a sale. You're buying more shares at lower prices, and time does the rest. When you're decumulating — selling assets to fund living expenses — a crash is the opposite. Every dollar you withdraw at the bottom is a dollar that can never participate in the recovery. You're locking in the loss in real time.

The five years on either side of your retirement date are where this asymmetry hits hardest. Get unlucky in that window and a perfectly reasonable plan can fail. Get lucky and a mediocre plan can look brilliant. The difference isn't skill — it's timing you don't control.

Which is why the goal in the fragile decade isn't to maximize returns. It's to survive sequence. The five questions below are how you stress-test whether you're built for that.

This is the foundational number. Before any other tactic, the question is whether you can simply not sell equities during a drawdown — wait it out, let the recovery happen, and refill from gains on the other side.

The conventional answer is two to three years of essential expenses held in cash and short-term Treasuries (government bonds maturing in under two years, the safest yield-bearing assets available). Some planners go to five. The right number depends on how much of your spending is fixed versus discretionary, and how willing you are to cut the discretionary side temporarily.

Personally, I run with three years. The buffer lives in short-term Treasuries and a high-yield savings account (HYSA — the online-bank version of a savings account, currently paying meaningfully more than what a traditional brick-and-mortar bank pays). The rule I follow: when markets are down, I draw from this bucket and don't touch stocks at the bottom. When markets are up, I sell first from positions with long-term capital gains (LTCG — profits on assets held more than a year, taxed at the lower 15–20% federal rate rather than ordinary income rates) and use those sales to refill the bucket. That way it's full going into the next drawdown, instead of empty when I need it most.

If your honest answer to this question is "less than a year," the entire rest of this issue can wait. Build the buffer first.

Most people imagine retirement as a binary event — working one Friday, not working the next Monday, with a portfolio that magically reallocates itself in between. The version that actually works is gradual.

Two named frameworks worth knowing:

The Bond Tent. Coined by Michael Kitces and Wade Pfau, this is the practice of increasing your bond allocation in the five years before retirement and the five years after, then decreasing it again as you settle into retirement. The shape, plotted out, looks like a tent — peaking at retirement, when sequence risk is highest, and falling back as you age into a longer time horizon. Counterintuitive, but the research holds up: the years right around retirement are when defensive ballast does the most work.

The Rising Equity Glide Path. Same authors, same insight, different lever. After the bond tent peak, you let your equity allocation drift back up over time. The longer you've survived sequence, the less sequence risk you have left, and the more you can afford to hold growth assets for the back half of a long retirement.

The thing both frameworks have in common: they treat allocation as something that moves through the fragile decade, not something you set once and forget.

The famous 4% rule — withdraw 4% of your starting balance in year one, adjust for inflation thereafter — was built on historical worst-case scenarios. It's a useful planning anchor, but it assumes you keep withdrawing the same inflation-adjusted dollar amount whether the market is up 30% or down 30%. Real retirees don't do that. Real retirees adjust.

One important caveat: the 4% rule was stress-tested against a 30-year horizon for someone retiring at 65. The longer the runway, the less margin a 4% start leaves. I'm 40, so any plan I build has to assume something closer to a 50-year runway — and at that horizon, I anchor lower. My personal starting point is 2–3%, not 4%. If you're targeting a traditional retirement age, 4% is a reasonable anchor. If you're aiming earlier, the math demands more conservatism.

The cleanest formalization of "adjust" is Guyton-Klinger guardrails — a dynamic withdrawal rule that gives you a small raise after great years and asks you to trim 10% after bad ones. The trim is uncomfortable for one or two years, then you reset. The math says this small amount of flexibility lets you start retirement with a higher initial withdrawal rate (often 5%+) without significantly raising the risk of running out.

The question to ask yourself isn't "what's my withdrawal rate." It's "how much of my spending could I cut for two years if the market dropped 30% the year I retired?" If the answer is "almost none," your plan is more fragile than it looks. If the answer is "20-30% of discretionary," you have far more room than the 4% rule alone suggests.

Sequence risk shrinks dramatically when a meaningful chunk of your essential expenses is covered by income that doesn't depend on selling assets at all.

The three sources worth thinking about:

Social Security, which is inflation-adjusted by design. Conventional advice says delay claiming as long as possible — waiting from 62 to 70 increases your monthly benefit by roughly 77%. That math is real, but so is the break-even point: you generally need to live into your late 70s or early 80s for delayed claiming to pay off in lifetime dollars. My own view leans the other way. Claiming earlier and letting that income offset portfolio withdrawals during the fragile decade is itself a sequence-risk hedge — the smaller monthly check is designed assuming you'll supplement it with portfolio income, and pulling less from your portfolio in the years it's most vulnerable is exactly the move this whole issue is about. Health, marital status, and survivor benefits all factor in, but the "always delay" default deserves more skepticism than it gets.

TIPS ladders — Treasury Inflation-Protected Securities held to maturity in a rolling sequence. You buy TIPS that mature in the years you'll need the cash, and the principal adjusts with CPI (the Consumer Price Index, the government's main inflation measure) along the way. It's the closest thing the U.S. government offers to a personal inflation-hedged paycheck.

Annuities, specifically a Single Premium Immediate Annuity (SPIA) — a contract where you hand an insurer a lump sum in exchange for a guaranteed monthly payment for life. This is one I'll be transparent about: annuities don't work for me. My read is that the same lump sum, invested in a diversified market portfolio, will likely produce more lifetime income — with the upside of continued growth and the option to leave a remainder to heirs — than what an insurer will pay out after their fees. But that math depends on having a strong stomach for volatility. For readers whose risk profile genuinely demands a guaranteed paycheck — or whose family circumstances make peace of mind worth more than expected return — a plain-vanilla SPIA covering essential expenses is a legitimate sequence-risk hedge. Just know what you're trading for it: loss of liquidity, inflation exposure unless you pay extra for a CPI rider, and the layer of insurer fees baked into the contract.

The question isn't whether to use any one of these. It's whether your essential, can't-skip expenses — housing, food, healthcare premiums, utilities — are covered by income that survives a bear market without you having to sell anything.

This is the question I'd most want my reader to sit with, because it's where the conventional wisdom is quietly going stale.

Standard retirement planning models assume a 30-year horizon for someone retiring around 65. That number traces back to actuarial tables (the statistical life-expectancy tables that insurers and pension funds rely on) built on 20th-century medical data. It treats the future like a continuation of the past.

The future may not cooperate. A few data points worth holding together:

GLP-1 drugs — semaglutide, tirzepatide, and the next generation behind them — have already started showing measurable effects on cardiovascular mortality, not just weight. The first wave of long-run mortality data is beginning to land.

AI-accelerated drug discovery. AlphaFold cracked protein structure prediction; the platforms built on top of it are now compressing early-stage drug development timelines from years to months. The pipeline of cancer, Alzheimer's, and metabolic-disease therapies coming through clinical trials over the next decade is qualitatively different from anything we've seen.

Earlier detection. Multi-cancer early detection (MCED) blood tests are moving from research into general use. Catching tumors at stage 1 instead of stage 3 doesn't add a few months to a life — it can add decades.

None of this guarantees you personally live to 100. But the probability distribution of how long today's 55- and 60-year-olds will live is shifting to the right, and the planning models most advisors use haven't caught up. The honest move is to plan as though you might need your portfolio to fund 35 or 40 years instead of 30 — and let the bond tent come down faster on the back end so you keep enough equity exposure to fund that longer tail.

The risk isn't dying with too much. The risk is outliving the plan.

Three of the questions above point toward frameworks you can actually implement. Pulling them together:

The Two-Bucket Strategy. The simpler version of bucket-based withdrawal, and the one I run myself.

Some planners prefer a three-bucket version with an additional middle layer of intermediate Treasuries, TIPS, and high-quality bond funds for extra ballast. That's a reasonable variation, but the two-bucket approach keeps the moving pieces minimal and forces more of your portfolio to stay in growth assets — which matters more, not less, the longer your runway.

Roth Conversion Window. The years between retirement and Required Minimum Distributions (RMDs — the IRS-mandated amount you have to withdraw from pre-tax retirement accounts each year, starting at age 73) are often a low-tax-bracket window. Earned income drops, but RMDs and Social Security may not have started yet. Converting traditional IRA dollars (pre-tax savings, taxed when you withdraw) to Roth (post-tax dollars that grow and withdraw tax-free) during this window is one of the few moves that simultaneously reduces sequence risk and improves the long-tail math. A CPA can help you size this against your specific bracket — it's almost always worth the conversation.

All three readers are circling the same question: how do I think about a portfolio that has to last longer than the rules of thumb were built for? The conventional advice — "get more conservative as you age" — was written for a shorter life and a calmer drawdown phase than most of us will actually experience.

The work-optional version is different. You add ballast going into the fragile decade, not for the rest of your life. You build a floor under your essential expenses so the stock portion of your portfolio is free to do its job — compounding over a horizon that may be longer than you've been told to plan for. And you stay flexible, because the difference between a plan that works and a plan that fails is often just the willingness to trim spending for one or two uncomfortable years.

The market doesn't care when you retire. The five questions above are how you stop needing it to.

My e-book is almost ready. Fill out this 2-question survey if you want to be notified about the launch.

Stay disciplined - Koh

Disclaimer: Nothing in this newsletter constitutes investment advice or a recommendation to buy or sell any security. Every reader's situation is unique — your age, family circumstances, tax bracket, health, and goals all change the math materially, and what works for me may not work for you. Withdrawal strategies, annuities, Social Security claiming decisions, Roth conversions, and TIPS ladders involve meaningful tax, legal, and individual financial considerations. Please consult a licensed financial advisor, CPA, and/or attorney before acting on any of the strategies above. Always do your own research.