The buffer costs its cap, and nobody prices it that way
RILAs sell downside protection for six years, but the cap on gains is the real premium.
By The Ledger · · 5 min read
A registered index-linked annuity promises to cushion the first 10 to 30 percent of market losses over a six-year term. In exchange, it caps how much of the market's gain you get to keep. The insurer that writes the contract sets that cap — and can reset it every term.
Ten-year Treasury yields sat near 4.4 percent in November 2025, the highest sustained level in roughly two decades. That single fact is doing most of the work behind the RILA sales pitch: bonds look less like free stability and more like a rate bet, so insurers are marketing a third bucket — part stock, part shield — to investors who no longer trust the old 60/40 split to behave.
The retiree with $200,000 and six years to wait
Picture an investor at 58, four years from a planned retirement, with $200,000 earmarked for the "safe" side of a portfolio. In a RILA with 20 percent downside protection and a 9 percent upside cap over a six-year term, a market that falls 15 percent costs that investor nothing — the insurer absorbs it. A market that rises 25 percent over the same term pays the investor roughly 9 percent a year, capped, not 25 percent. The insurer keeps the difference. That gap between what the index actually returns and what the contract pays out is the business model, not a footnote to it.
Surrender charges typically run through the first five to seven years, and pulling money out early can cost several percentage points of the original investment. Withdrawals from nonqualified contracts are taxed as ordinary income on the earnings portion, and money taken out before age 59½ can trigger a 10 percent federal penalty. None of that shows up in the pitch about downside protection. It shows up in the contract's fine print, which is where insurers put the numbers that don't sell as well as the ones on the brochure.
Bonds used to do this job for a fee near zero
The 60/40 portfolio became retirement orthodoxy in the 1990s because bonds did two things reliably: they paid a coupon, and they moved opposite stocks often enough to smooth a portfolio's ride. That relationship held through most of the 2000s and 2010s, a period when the Federal Reserve kept policy rates near zero and bond yields followed. A 10-year Treasury yielding 1.5 percent, as it did through much of 2020 and 2021, offered stability but almost no income — asking bonds to be both ballast and paycheck, and mostly failing at the second job.
The Fed's rate increases from 2022 onward pushed yields back toward 4 to 5 percent, which should have restored the bond's old role. Instead it exposed a newer problem: bond prices fell sharply as yields rose, meaning the "safe" half of many 2022 portfolios lost value at the same time stocks did. That correlation breakdown — bonds and stocks falling together — is the specific memory driving interest in RILAs now. Insurers are selling protection against a repeat of exactly that year, dressed as a structural upgrade rather than a hedge against one bad correlation regime.
The insurer prices the cap to cover its own hedge, not your goals
A RILA's upside cap isn't arbitrary. Insurers buy options to fund the downside protection they're promising, and options get more expensive when market volatility rises. When volatility is high, insurers set lower caps to keep the product profitable for themselves. That means the contract offering the most protection often arrives with the least room to gain — precisely when an investor might want both. The trade-off isn't disclosed as a formula anywhere in the marketing language; it shows up only when the next term's cap is set, sometimes years after the original sale.
There's a second cost that rarely appears in the pitch: opportunity. Over the six-year term ending in a strong bull run, an investor holding an uncapped S&P 500 index fund keeps the whole gain, minus an expense ratio typically under 0.1 percent a year. The RILA holder in the same market caps out at whatever rate the insurer set at the start of the term — often in the high single digits to low teens annually — and pays an internal cost structure that isn't itemized as a fee the way a mutual fund's expense ratio is. Comparing the two only after the term ends means comparing a known number against one that was fixed in advance and never adjusted for how the market actually performed.
Who actually needs the buffer, and who's paying for insurance against a scenario that may not repeat
The honest case for a RILA is narrow: an investor within five to seven years of needing the money, unwilling to hold bonds at current yields but also unwilling to ride out a 2008-style drawdown in stocks. For that investor, paying for a capped, protected middle ground can make arithmetic sense, particularly inside a tax-deferred account where the ordinary-income treatment on withdrawal matters less. The case gets weaker for anyone with a decade or more of horizon, where historical data shows uncapped equity exposure has outperformed capped strategies across most rolling periods — precisely because caps remove the market's best years along with its worst ones, and the best years happen more often than the worst.
The open question the industry hasn't priced yet
Nobody has data yet on how RILAs perform across a full multi-decade retirement, because the product is barely 15 years old and most contracts run in six-year terms. The back-tested comparisons advisers show clients are simulations, not track records. Whether the buffer is worth its cap depends on the sequence of markets an individual retiree lives through — a sequence nobody can know in advance, which is the same uncertainty bonds were originally supposed to solve.
Is a RILA the same as a bond? No — a RILA links returns to an index and caps gains, while a bond pays a fixed coupon regardless of market moves.
Can you lose money in a RILA? Yes — protection typically covers only 10 to 30 percent of losses; a bigger downturn or an early withdrawal can still cost principal.