Original planning page, not a reprint. Yields are the U.S. Treasury real par curve for November 3, 2026. Costs are the present value of the real spending stream on that curve. Not a broker quote. Not tax, legal, or personalized advice. Check the order ticket and a Social Security estimate before you buy.
Related: Retire at 60 · Tax timeline 55–75 · Guardrails · Social Security COLA lag
Build 20 years. Shape it. Do not build a flat 30, and do not stop at 5.
A TIPS ladder is a spending floor, not the portfolio. Fat rungs until Social Security starts. Thin rungs after that, equal to essential spending minus Social Security, in today's dollars. Social Security already has a COLA. Do not inflation-gross it a second time.
Buy the ladder in the traditional IRA at the broker. TreasuryDirect is the other door: individual or entity accounts, auction bids, $100 minimum. A bank, broker, or dealer is how an IRA buys Treasuries. TreasuryDirect is not an IRA custodian.
Phantom income is the reason the IRA wins. In a taxable account, Treasury says federal tax is due each year on interest, and that a change in principal during the year may affect your taxes. The IRS makes you report the inflation bump to principal as original issue discount even though you did not receive the cash. Inside a traditional IRA that annual tax does not hit. Distributions are taxable in the year you receive them. Maturity inside the IRA is not the taxable event. The withdrawal is.
State tax cuts the other way. Treasury interest is free of state and local income tax when you hold the bond yourself. An IRA withdrawal is generally just IRA income. If you move to a state that taxes retirement distributions, the state exemption does not come through the IRA. Do not park a big ladder in taxable just to chase that exemption unless a tax pro runs the numbers. The federal phantom-income hit usually dominates.
A Roth IRA also works, and qualified withdrawals are tax-free. Roth space is scarce. Prefer stocks there. Default home for the ladder is the traditional IRA.
RMDs start at 73 under the current IRS FAQ, including traditional IRAs, even if you are already retired. A ladder that is most of the traditional IRA will eventually be forced to pay out more than the maturing rung. Keep other IRA assets so the RMD can come from somewhere else, or do not build a 30-year ladder that the IRS will tear up. Confirm your own required beginning date on IRS.gov. The FAQ states age 73. It does not spell out a later increase.
Treasury sells TIPS for 5, 10, or 30 years. The coupon is fixed at auction and is never less than 0.125%. Interest is paid every six months on the inflation-adjusted principal, so the dollar coupon moves. At maturity you get the higher of adjusted principal or original par. You do not get less than original par.
Sort by real yield, not coupon. A 0.125% coupon can still yield 2.5% real if you buy it below par.
Auctions will not fill a ladder. New bonds land on a set calendar. 5-year: April and October, reopenings June and December. 10-year: January and July, reopenings March, May, September, and November. 30-year: February, reopening August. The in-between years are older bonds on the secondary market. That is where almost every rung comes from.
A TIPS fund (TIP, SCHP, VTIP, and the like) is not a ladder. No maturity date. Those funds lost money in 2022 when real yields rose, which is exactly the year a retiree does not want the "safe" sleeve down. A rung you hold to the year you plan to spend does not have that problem. The statement value can still bounce. Ignore it.
Four planning runs on the November 3, 2026 real curve: 5-year 2.35%, 7-year 2.49%, 10-year 2.65%, 20-year 2.94%, 30-year 3.09%. Each year's spending is discounted at the yield for that horizon. That present value is what a fairly priced ladder costs to deliver the spending (coupons plus maturing principal arranged to hit the target). A par-bond check at a flat 2.65% matched the annuity math to the dollar. These are ceilings for planning, not the number on tomorrow's ticket.
Real dollars per year, held the whole term. Multiply if your gap is in between. $20,000 a year is twice the $10,000 column.
| Length | $10k / year | $40k / year | $60k / year | Dollars of ladder per $1 of annual spending |
|---|---|---|---|---|
| 10 years | $88,000 | $350,000 | $525,000 | 8.8× |
| 15 years | $123,000 | $491,000 | $737,000 | 12.3× |
| 20 years | $153,000 | $611,000 | $917,000 | 15.3× |
| 30 years | $200,000 | $799,000 | $1,198,000 | 20.0× |
Thirty years of $40,000 is about $800,000. That is a lot of IRA to freeze if Social Security will cover a chunk of that $40,000. Length is a spending decision, not a "longer is safer" trophy.
Same pre-Social Security gap. The only change is whether later rungs pretend Social Security does not exist.
| Shape | IRA cash to lock it |
|---|---|
| Bridge only: $40k × 10 years | $350,000 |
| Social Security covers a lot: $40k × 10, then $5k × 10 | $383,000 |
| Social Security covers some: $40k × 10, then $15k × 10 | $448,000 |
| Same, longer thin tail: $40k × 10, then $15k × 20 | $518,000 |
| Bigger gap: $60k × 10, then $25k × 10 | $688,000 |
| Flat $40k × 20 (funds Social Security's job twice) | $611,000 |
| Flat $40k × 30 | $799,000 |
The default shape ($40k then $15k, 20 years) is about $448,000. The flat 20-year version of the same $40k is about $611,000. Shaping saves roughly $163,000 because you stop buying income Social Security will already pay. Extending the thin tail another 10 years costs about $70,000 more, not another $200,000. That is the only 30-year idea worth a look, and only if you want the small residual gap locked.
Same default shape: $40k for 10 years, then $15k for 10. You have not bought yet. The whole curve shifts before you do.
| Real yields vs Nov 3, 2026 | Cost to lock the same spending | Vs buying today |
|---|---|---|
| 1.00% lower | $483,000 | +$35,000 |
| 0.50% lower | $465,000 | +$17,000 |
| Today | $448,000 | — |
| 0.50% higher | $432,000 | −$16,000 |
| 1.00% higher | $417,000 | −$31,000 |
Waiting for a better yield is a bet, not a plan. A 1% drop in real yields before you buy makes the same floor about $35,000 more expensive. After you own the rungs, a yield move changes the statement value and does not change the real spending if you hold to maturity. That is the whole point of the ladder. Do not sell it because the screen went red.
Level real income, whole term, same curve. Useful if you already know how much of the IRA you are willing to freeze.
| IRA slice | 10-year floor | 15-year floor | 20-year floor | 30-year floor |
|---|---|---|---|---|
| $250,000 | $28,600 / yr | $20,400 / yr | $16,400 / yr | $12,500 / yr |
| $500,000 | $57,100 / yr | $40,700 / yr | $32,700 / yr | $25,000 / yr |
| $750,000 | $85,700 / yr | $61,100 / yr | $49,100 / yr | $37,600 / yr |
$500,000 buys about $41,000 a year for 15 years, or about $33,000 a year for 20. It does not buy a $60,000 lifestyle for 30 years. If the gap is bigger than the slice, the ladder covers essentials and the guardrail sleeve covers the rest. Do not stretch the ladder until the stock sleeve is a rounding error.
Illustration, not a forecast and not a Monte Carlo. $1,000,000. Need $40,000 real per year for 10 years. Ladder case spends $350,000 on a 10-year bridge and leaves $650,000 in stocks, untouched. No-ladder cases withdraw $40,000 at the start of each year, then apply the return. Bonds in the 60/40 earn 0% real, which is generous to the no-ladder case.
| Path | Ladder + untouched stocks | No floor, 100% stocks | No floor, 60/40 |
|---|---|---|---|
| Good: stocks +6% real every year | $1,164,000 left | $1,232,000 left | $936,000 left |
| Grind: stocks −4% real every year | $432,000 left | $343,000 left | $434,000 left |
| Crash then repair: −35%, then +5% for 9 years | $655,000 left | $505,000 left | $571,000 left |
Read it straight. In a good decade, skipping the ladder and staying in stocks wins, by about $68,000 on this setup. You paid for insurance you did not use. In a grinding down decade, the ladder and a 60/40 finish in the same neighborhood, and both beat selling stocks every year. In a year-one crash, the ladder finishes about $150,000 ahead of all-stock and about $84,000 ahead of 60/40, because you did not sell the recovery. That is the job. It is not free upside.
Added Nov 4, 2026. Planning math on that day's real curve. Confirm the CUSIP yield before you wire a rung.