Is Locking In Today's TIPS Yields Market Timing?

In 2008, Jason Scott, William Sharpe and John Watson set out to price the 4% rule. Before arguing about withdrawal rates, they asked what the thing a retiree actually wants costs: a dollar of inflation-adjusted spending, every year, for thirty years. At the 2% real interest rate they assumed, the answer was “a little less than $22.40,” which made the highest withdrawal rate anyone could guarantee about 4.46%.

That 2% was an assumption. Today it is a quote. On September 11, 2026, the Treasury real yield curve priced the same thirty-year stream at $19.91 per dollar of annual spending, a guaranteed rate of 5.02%. On November 9, 2021, it priced the identical stream at $34.41, a guaranteed rate of 2.91%.

Which leaves an awkward question for anyone who has absorbed the standard advice. If you are not supposed to time the market, why should today’s TIPS yield change how much of your retirement you choose to make certain?

The Price of the Same Retirement Fell 42%

The chart below prices one fixed thing. The stream never changes: $50,000 a year, adjusted for inflation, for thirty years, the kind of spending that covers property taxes and groceries whether or not the stock market cooperates. Only what the market charges for it moves.

At the November 2021 trough in real yields, that stream cost $1,720,569. On September 11, 2026, it cost $995,642. A household that wanted to guarantee exactly the same retirement needed $724,927 less capital to do it, having done nothing and saved nothing in between.

One caveat belongs here rather than in a footnote, because it is the kind that gets left out. Treasury only began publishing a thirty-year real yield in February 2010, so “the cheapest in the series” is a claim about sixteen years, not about history. On the tenors that did exist in 2008, real yields ran higher than they do now: the ten-year reached 3.15% on November 21, 2008, against 2.60% today. Campbell, Shiller and Viceira examined that episode and concluded much of it was a liquidity dislocation rather than a change in the underlying price of patience. Lehman’s TIPS collateral was being liquidated into a market where dealers would not take the other side, commodity-overlay investors were forced sellers, and TIPS asset swap spreads “increased from their normal level of about −35 basis points to about +100 basis points” while nominal Treasury spreads sat still. A quoted real yield is a market price, and market prices sometimes contain things that have nothing to do with you.

The Test Is Whether the Decision Contains a Prediction

Market timing means acting on a view about where a price is going. “Stocks look expensive, I will wait” only pays off if stocks subsequently fall. The claim is about the future, and it can be wrong in a way you can measure.

Buying an asset that matches a liability asks less of you. The only input is what it costs, today, to make one specific future expense stop depending on markets at all. Where the thirty-year real yield goes next never enters the calculation, so if real yields rise again next year you have lost nothing that mattered to the decision. You already own the groceries.

That distinction has a formal home. Campbell and Viceira showed that what counts as a riskless asset depends entirely on what the investor is trying to fund. In their model, an investor who cares about a stable stream of real consumption should hold long-term inflation-indexed bonds, which in the limit behave like an indexed perpetuity; as they put it in “Who Should Buy Long-Term Bonds?”, conservative investors “hold assets to hedge the risk that real interest rates will decline.” Treasury bills, perfectly stable over any three months, leave that same investor fully exposed.

Campbell, Shiller and Viceira pushed the point further. The volatility of inflation-indexed bonds, they argued, “far from being a drawback, demonstrates the value of inflation-indexed bonds for conservative long-term investors.” A TIPS ladder can swing around on your brokerage screen while being the safest thing you own, measured against the spending it was bought to fund. Robert Merton made the same argument for households in Harvard Business Review: measured in the units that matter, which is how much income a pile of money can be converted into, T-bills are “very risky, nearly as volatile as the stock market.”

In 2022, Pensions Lost 18.6% and Got Better Funded

Corporate pensions ran this experiment in public, at scale, and the results are in their audited filings. Milliman’s study of the 100 largest U.S. corporate plans found that in calendar 2022 those plans returned −18.6%, the worst investment year in the study’s history. Their funded ratio improved anyway, from 96.3% to 99.3%. The discount rate used to value what they owe went from 2.73% to 5.18%, and the projected benefit obligation fell from $1.85 trillion to $1.33 trillion. The liability repriced harder than the assets did.

That is the mechanism a retirement plan built only around portfolio returns cannot see. A pension actuary marks both sides of the balance sheet every quarter. A household that tracks a portfolio balance and multiplies by 4% is marking one side and holding the other fixed at an assumption, which is why 2022 felt like an unrelieved disaster to individual investors and showed up as an improvement in pension filings.

The analogy stops before the prescription, though, and this is worth stating plainly because the vendor literature tends not to. Corporate plans de-risk as funded status improves partly because a sponsor gets very little benefit from surplus. Excess assets are difficult to withdraw and are taxed punitively when they are, while the sponsor owns the entire downside of a deficit. That asymmetry is a feature of ERISA and the tax code, and a household does not face it. You keep your surplus. Pensions demonstrate that liabilities have prices; they do not demonstrate that you should buy insurance the moment you can afford it.

If You Already Own the Ladder, Nothing Got Cheaper

Here is where enthusiasm about high real yields usually goes wrong. Suppose you were the household that bought the thirty-year ladder at the November 2021 peak, paying $1,720,569 for $50,000 a year. Five years later you have collected five payments and twenty-five remain. Priced at today’s curve, those twenty-five payments are worth $890,585. At the curve you bought them on, the same twenty-five payments were worth $1,422,012. Your position is marked 37% lower.

And it makes no difference to you. The $50,000 a year still arrives, still indexed to CPI, still backed by the Treasury. The mark-to-market loss is an accounting shadow cast by a liability you already defeased. Selling would realize it; holding the thing you bought does not.

So the repricing is available only to the household that has not bought yet, which is a much narrower group than “retirees.” Take a hypothetical portfolio of $1.2 million, held flat in real terms, against that same thirty-year $50,000 floor. In November 2021 it covered 0.70x of the cost. In September 2026 it covers 1.21x. Nothing was earned. The denominator moved. For what those coverage levels imply about how much to guarantee, and the point below which buying a partial floor makes a plan worse rather than better, see how much of your retirement should be a guaranteed income floor.

Where the Line Into Timing Sits

Three behaviors get bundled together under “responding to yields,” and only one of them requires a forecast.

The middle row is dynamic asset allocation, and under a broad enough academic definition someone could call any state-dependent allocation change timing. It does not carry the thing that makes timing dangerous, which is a prediction that has to come true. The third row does, and it is the trap that price-aware liability matching sets for the people most attracted to it. A retiree who can fully secure essential spending today and waits for a better print has put an achievable goal at risk on a view about interest rates. The thirty-year real yield fell 196 basis points between its November 2018 high and its December 2021 low. It can do that again.

Bond managers formalized the answer forty years ago. Martin Leibowitz and Alfred Weinberger’s contingent immunization let a portfolio be managed actively while maintaining what they called a safety net: a minimum acceptable return, the present value of which is recomputed continuously at current market rates. The gap between assets and that required amount is the cushion. When the cushion reaches zero, active management stops and the portfolio is immunized that day. The trigger is a comparison between what you hold and what the thing you want costs at that moment, against a threshold set in advance. Leibowitz applied the same machinery to dedicated bond portfolios for pensions in 1986, where cash-flow matching arranges coupons and principal to land on the liability schedule directly.

Scott and Watson’s floor-leverage rule carries the same idea into retirement with an annual ratchet: gains in the risky sleeve are harvested each year and spent on additional guaranteed income. Their specific implementation, which pairs a large spending guarantee with a small 3x leveraged equity sleeve, is not something to copy without understanding how daily-reset leveraged funds behave. The mechanic underneath it is sound and general. Decide the rule before the market gives you the opportunity, so that acting on it is bookkeeping rather than a fresh decision each time you look at a yield.

Modern academic work has moved in the same direction. Mantilla-García, Martellini, García-Huitrón and Martínez-Carrasco argue in the Journal of Banking & Finance that defined-contribution plans should be measured by a funding ratio whose denominator is “the present value of the total retirement income achievable,” and report that allocation rules built on that metric dominate conventional target-date strategies on retirement outcomes. David Blanchett reaches a compatible conclusion in the Financial Analysts Journal, finding that a funded-ratio-driven strategy yields different guidance from the probability-of-success framing most retirement software still uses.

The Case for Not Buying a Floor at All

The strongest objection is that the whole exercise is optional. Barton Waring and Laurence Siegel’s annually recalculated virtual annuity solves the same problem without buying anything. Each year you spend what a freshly purchased annuity would pay, priced at that year’s interest rates on your current portfolio over your remaining years. Spending fluctuates with markets, and you can never run out of money. Notice that this rule is also price-conditioned, which makes Waring and Siegel allies on the general principle and opponents on the specific remedy: they would say a retiree willing to flex consumption never needs to buy the floor, at any price.

Moshe Milevsky and Virginia Young showed that the option to delay annuitizing has real value well into a retiree’s seventies or eighties, depending on risk aversion. That result is about annuities specifically, where much of the value of waiting comes from mortality credits you cannot collect early, so it transfers only partly to a TIPS ladder you can sell tomorrow. It is still a reminder that “secure it now” is not the default answer.

Zvi Bodie has argued for decades that the safe asset is TIPS and that you should build the floor first, full stop, regardless of what it costs. Michael Kitces argues the opposite corner, that a household funded ratio marked at current market rates swings around too much to steer by and ignores human capital and mortality pooling. Both objections land. The useful response to Kitces is that a volatile reading is still a real one, and that the alternative most people use, a fixed 4%, is not a stable measurement so much as an unchanging assumption.

What the Price Does Not Tell You

• Cheaper in absolute terms is not cheaper relative to stocks. A higher risk-free real rate raises the bar that every risky asset has to clear, and it plausibly raises expected returns across the board. The floor got cheaper in dollars; whether it got more attractive than equities depends on what happened to the risk premium, which nobody can observe. See the risk-free rate as your hurdle rate.
• CPI-U is not your inflation. TIPS index to the national urban consumer basket. The BLS experimental index for older Americans rose 3.3% a year from December 1982 to December 2007 against 3.1% for CPI-U, driven mostly by medical care and shelter. A ladder that is exactly right in CPI-U terms can drift against a retiree’s actual costs.
• A thirty-year ladder is not longevity insurance. It pays until the last rung matures and then stops, whether or not you have. Only an annuity pools mortality risk. Sexauer, Peskin and Cassidy’s decumulation benchmark handles this by spending 88% of capital on a twenty-year TIPS ladder and the remaining 12% on a deferred life annuity.
• The tax treatment is awkward in a taxable account. Inflation adjustments to principal are taxed in the year they accrue, years before you receive them. See how to build a TIPS ladder for the account-location ordering and the 2037 to 2039 maturity gap.
• Your liability may not be as fixed as the model. Siegel and Waring’s dual duration work shows that a liability only partly linked to inflation is best hedged with a blend of TIPS and nominal bonds rather than TIPS alone. Household spending is partly discretionary and partly deferrable, which makes the true liability softer than a level thirty-year stream.

Bottom Line

Observing the price of certainty is a different act from predicting it. The market will currently sell a thirty-year stream of real income for about 42% less capital than it charged at the 2021 trough, and that is a number you can look up rather than a call you have to make. Whether you should act on it depends on how much of your spending is already covered at those prices, not on whether 3% real feels like a good level. The moment the decision starts depending on where yields go next, you have swapped a price you can see for a forecast you cannot check.

Key Takeaways

• Real yields set the capital cost of guaranteed retirement income. The same $50,000 a year of inflation-adjusted spending for thirty years cost $1,720,569 in November 2021 and $995,642 on September 11, 2026.
• Responding to a current price requires no forecast. Market timing pays off only if a prediction comes true. Buying an asset that defeases a liability pays off the moment you own it, whatever yields do afterward.
• Both sides of a plan are priced. The 100 largest U.S. corporate pensions lost 18.6% on assets in 2022 and still improved from 96.3% to 99.3% funded, because their liabilities repriced harder.
• Rising yields are not a windfall if you are already hedged. A household that bought the ladder at the 2021 peak is marked 37% lower on its remaining payments and receives exactly the income it purchased. The gain accrues only to buyers who have not bought.
• Coverage is the decision variable, not the yield. Two retirees looking at the same 3% real yield should reach different conclusions if one can secure essential spending at that price and the other cannot.
• Precommitment is what keeps this from becoming timing. Contingent immunization has worked since 1982 by fixing the trigger in advance and recomputing it at current market rates, never by waiting for a better number.

Frequently Asked Questions

Is buying TIPS when real yields are high market timing?

Not by the usual definition. Market timing requires a forecast of future prices or returns, and it fails when the forecast is wrong. Buying an inflation-indexed bond that matures when you need the money depends only on today’s price and on what you need the money for. The decision does become timing if you delay a purchase you could afford because you expect yields to rise further.

Should I wait for TIPS yields to go higher before building a ladder?

That question contains an interest rate forecast, which is the one ingredient liability matching lets you avoid. The thirty-year real yield fell 196 basis points between November 2018 and December 2021, so waiting has a real cost if you are wrong. If you cannot yet cover the spending you care about at current prices, the answer is usually to keep growth assets working rather than to wait for a better yield; if you can cover it, waiting risks an outcome you have already secured.

Do higher real yields make retirees better off?

They make future guaranteed income cheaper to buy, which helps anyone who has not bought it yet. Retirees already holding duration-matched TIPS see their holdings marked down by roughly the same amount their liability fell, so their real position barely changes. Retirees holding equities and cash and considering a floor are the ones the repricing benefits.

What real yield is high enough to justify locking in income?

No single level works, which is why a threshold rule is the wrong shape for this decision. A retiree with $900,000 and $100,000 a year of portfolio spending and a retiree with $4 million and $60,000 of spending face identical yields and completely different situations. The question worth asking is what fraction of the spending you care about you can secure at today’s prices.

How much does it cost to guarantee $50,000 a year for thirty years?

Priced off the Treasury real yield curve on September 11, 2026, about $995,642, which is a level payout rate of 5.02%. That figure is a curve valuation before bid-ask spreads, commissions and the practical problem of the 2037 to 2039 maturity gap, and it moves every day the curve moves. It also assumes the ladder simply ends after thirty years.

Does this contradict the advice to ignore market forecasts?

It sharpens it. The advice worth keeping is to avoid decisions that require you to predict prices. A yield you can read off a Treasury page is not a prediction, and neither is a written policy that says what you will do when coverage of your essential spending crosses a level you chose in advance.

Related Guides

• How Much of Your Retirement Should Be a Guaranteed Income Floor? for the coverage levels at which flooring helps, and the level below which a partial floor makes a plan worse.
• How to Build a TIPS Ladder for the mechanics, the phantom-income tax problem, and a cost calculator that prices a ladder off the live curve.
• Asset-Liability Matching for the duration framework underneath all of this.
• The Risk-Free Rate Is Your Hurdle Rate for what higher real yields do to every other asset you own.
• Understanding Your CEFR Score for the funded-ratio measurement this argument runs on.