Eight claims about indexed annuity returns, tested one at a time

"You get the upside of the market with none of the downside." If you have sat through a presentation about an indexed annuity, you have probably heard a sentence close to that one. It is built from true parts, which is exactly what makes it hard to evaluate. The parts are true in a narrow sense; the sentence as a whole promises more than the contract delivers.

This post takes eight claims that circulate in brochures, seminars and sales conversations, states each in its strongest form, and tests it against how an indexed annuity contract actually works. The verdicts are not meant to condemn the product, which suits some savers well. They are meant to show where a claim is accurate, where it is stretched and what to ask. All the figures below are hypothetical, chosen for easy arithmetic, and no company's real terms.

If you want to practice the underlying arithmetic yourself, see work out an indexed annuity credit by hand with your own contract terms. For the product's overall design, the guide to fixed indexed annuities: caps, participation rates and trade-offs covers it.

The eight claims at a glance

Claim 1: "You can't lose money"

The claim, strongest form: your principal is protected; if the market falls, you lose nothing.

The test. The first part is accurate in a specific sense. A fixed indexed annuity has a floor, usually 0%, so a falling index does not reduce your account through the credit formula. That is the core of the product, and it is real.

The claim breaks down when "lose money" is read more widely than "lose to the index."

• Leaving early. Say you have $100,000 of account value in year two, a free withdrawal of 10% and a charge of 7% in that year. A full surrender takes $90,000 × 7% = $6,300 in charges, so you receive $93,700 before tax, less than your account value and possibly less than the premium. If the contract has a market value adjustment, it could lower the figure further or raise it. See annuity surrender charges and free withdrawals.
• Statutory minimums. The law sets a minimum value the contract must keep, but in the early years that minimum can be lower than the full premium. Read the guaranteed values on the data page; the idea is explained in the minimum guaranteed value in a fixed annuity: what the floor protects.
• Charges. A rider fee comes out of the account whether or not the index credit was positive.
• Taxes. Withdrawals of gains are ordinary income, and gains taken before 59½ generally carry a 10% additional tax unless an exception applies.
• Insurer failure. The guarantee is only as strong as the company behind it, with state guaranty association limits as the backstop.
• Inflation. A year of 0% credit in a year of rising prices is a loss of purchasing power.

Verdict: accurate about index declines inside the contract's rules, and misleading as a general statement. A fairer version is "the index cannot reduce your credited value, but leaving early, fees and the insurer's strength still matter."

Claim 2: "You get the market's gains"

The claim, strongest form: you participate in the market's rise and sit out its falls.

The test. Participation is limited by design. The cap, the participation rate or the spread (explained in cap rate and participation rate) trims every gain. The company pays for the floor partly by keeping part of the upside, so the question is how the trade works out across different market paths, and it can go either way.

Take two hypothetical four-year paths for a price index, and a contract with a 7% annual cap and a 0% floor, annual point-to-point.

Path A: +20%, +15%, −10%, +12%. The index compounds to 1.20 × 1.15 × 0.90 × 1.12 = 1.391, or about +39.1%. The contract credits 7%, 7%, 0% and 7%, which compounds to 1.07³ = 1.225, or +22.5%. The index is ahead by about 16.6 percentage points.

Path B: −10%, −20%, +15%, +10%. The index compounds to 0.90 × 0.80 × 1.15 × 1.10 = 0.911, a loss of about 8.9%. The contract credits 0%, 0%, 7% and 7%, which gives 1.07² = 1.145, or +14.5%. The contract is ahead by about 23.4 points.

Neither path is a forecast. Together they show what the structure does: it gives up the best years and avoids the worst. A saver who cares most about avoiding the second path will value the trade; a saver who expects mostly good years will not. Contracts that credit a share of the gain (participation) or subtract a spread give different shapes; the table in the worksheet post compares them.

Verdict: "market-like gains" is stretched. A more accurate statement is "a limited share of index gains, in exchange for a floor on losses."

Claim 3: "Your gains are locked in every year"

The claim, strongest form: each year's credit becomes part of your base and can't be taken back by a later decline.

The test. On the whole this is true for annual crediting designs, and it is one of the product's real features. Once the credit is added on the contract anniversary, later index declines do not reduce it, except through charges, withdrawals or an adjustment on exit.

The qualifier is timing. Credits are generally determined on the anniversary, not continuously. Say the index rises 10% in the first eleven months of a contract year and then falls 8% in the twelfth. The index finishes the year at 1.10 × 0.92 = 1.012, up 1.2%, so the credit under annual point-to-point crediting is 1.2%, not 6% or 10%. The mid-year peak is never measured. If you surrender or take a large withdrawal before the anniversary, many contracts credit nothing for the partial year, or a reduced amount. The contract's rule for interim withdrawals and for the death benefit on a non-anniversary date is in the crediting section.

Verdict: accurate after the anniversary, and a question mark before it. Ask what is credited on a mid-year exit.

Claim 4: "It guarantees a fixed growth rate, like 7% a year"

The claim, strongest form: the contract grows at a guaranteed rate whatever the market does.

The test. This claim is nearly always about a benefit base on an income rider, and not about the money you can withdraw in a lump sum. A rider may state that the income base grows by a roll-up rate each year while you defer, to determine how large the lifetime income will be. The account value is separate and may be much smaller.

Say a hypothetical rider rolls a $100,000 income base at 7% a year for ten years: $100,000 × 1.07¹⁰ = $196,715. If the lifetime withdrawal percentage at the age you start is 5%, the guaranteed income is $196,715 × 5% = about $9,836 a year. That is a real, useful guarantee for someone who wants income. But the account value may be $130,000 at the end of the ten years, and if you surrender, you receive the account value less charges, not $196,715. A rider also has an annual charge, usually on the base; if the charge is 1% of a $196,715 base, that is about $1,967 a year, which is a higher share of a smaller account value.

The distinction is the subject of benefit base versus account value: the number income riders really use, and the rider mechanics are in guaranteed lifetime withdrawal benefit riders.

Verdict: accurate about the income base on a rider, and misleading when presented as the growth of your money. Ask: "Is that the account value, or the number the income is calculated from? Can I withdraw it?"

Claim 5: "There are no fees"

The claim, strongest form: you pay nothing; the company is paid from its own profit.

The test. A basic fixed indexed annuity usually has no listed annual fee, and that part is true. The company's costs and profit are covered in other places.

• The gains you give up. Caps, participation rates and spreads are how the product is priced. A saver who gets 6% when the index gained 9% has paid a 3-point cost that never appears on a statement.
• Commissions. They are paid by the insurer, and they influence the terms. The guide on annuity fees, explained explains how.
• Surrender charges. A long schedule is a cost if you need the money.
• Rider charges. Optional riders are billed separately, and often on the base. A 1% charge on a $150,000 benefit base is $1,500, which is 1.25% of a $120,000 account value.
• Bonus recapture. A premium bonus can come with a longer schedule, lower caps or a vesting rule.

Verdict: accurate about listed fees in the base contract, and incomplete. A better question is "what do I give up in exchange for the protection, and what does each optional feature cost?"

Claim 6: "It tracks the index's return"

The claim, strongest form: you earn what the stock index earned, within the limits.

The test. The crediting formula uses the index's price level in most contracts, so it ignores dividends paid by the companies in the index. For broad stock indexes, dividends have historically made up a meaningful share of long-run total returns, so a price-only measure reads lower than a total-return measure over the same years. If the cap is the binding limit, the omission matters less, because the credit stops at the cap anyway; if the participation rate or spread is the main limit, it matters more.

Other index choices also change the picture. A proprietary or volatility-controlled index, built by a bank, may allow a higher participation rate or cap because it is designed to move less, but it may also lag in strong markets. Many such indexes have short or hypothetical histories. The contract describes the index, and the index's rules document describes how it is built. Ask for both.

Verdict: "tracks the index" is loose. Confirm whether the index is price-only or total-return, whether it is a standard or proprietary one, and what its track record actually is.

Claim 7: "Historical results show what you'd have earned"

The claim, strongest form: a chart shows that over the past twenty years, the contract would have averaged a stated percentage.

The test. Past index data is real, but a back-test applies today's caps and rates to it. The company did not actually offer those rates over those years, and it can reset them in the future. A chart may also pick a favorable start date or period. Treat any such figure as an illustration of how the formula behaves, with a footnote saying so. A labeled illustration is a legitimate teaching device, as in Claim 2 above; the problem arises when it is presented as an expectation.

Questions worth asking: which years are included, what cap or rate was assumed, whether the cap would have been the same each year, whether fees and withdrawals are reflected, and whether the same illustration is shown with the guaranteed values alongside the non-guaranteed ones.

Verdict: a back-test is a hypothetical, whatever the chart looks like. Ask for the guaranteed column next to it.

Claim 8: "The cap you see is the cap you keep"

The claim, strongest form: the rate in the proposal applies for as long as you hold the contract.

The test. In most contracts the company declares caps, participation rates and spreads for each crediting period, often yearly, within minimums or maximums stated in the contract. The proposal shows the current figures. The contract shows the limits of what can change. The gap between the two is the room the company has to adjust.

Say the first-year cap is 6% and the company renews it at 4.5% in a later year. In a year with a 9% index gain, the credit falls from $6,000 to $4,500 on a $100,000 base, a difference of $1,500 for that year. The surrender schedule still runs, so the owner cannot easily leave. A contract that guarantees its rate for multiple years, or fixes the minimum cap at a number close to the current one, would reduce the risk. The post on declared rates and non-guaranteed elements in annuity contracts sorts every term into guaranteed or changeable.

Verdict: rarely accurate beyond the first period. Ask for the contract's minimum cap, participation rate and maximum spread, and how much notice the company gives of a change.

How to use the verdicts

A claim in the strongest form is a marketing summary. The questions that follow from the eight tests are practical.

1. Where in the contract does it say that? A promise that exists only in a brochure is not a contract term.
2. Is this guaranteed, or can the company change it?
3. What happens in a bad exit year? Ask for the dollar figure in year two and year five.
4. What is the index, and does it include dividends?
5. Which numbers does an illustration assume, and what do the guaranteed columns show?
6. How is the person recommending the product paid, and would the answer change if I chose a different product?

Sellers of annuities are bound by state suitability and best-interest standards, covered in best interest rules for annuity sales. A seller who answers these questions plainly and in writing is behaving the way the rules expect. If the answers are evasive, the state insurance department takes complaints, and a fee-only planner who is not paid by the sale can review the contract with you.

None of this means an indexed annuity is a poor choice. For a cautious saver who values a floor, will hold the contract through its surrender period and understands that gains are limited, it can be a reasonable fit. The aim of testing the claims is to buy the product that is actually in the contract and not the one in the pitch.

Key takeaways

• The 0% floor protects the credited value from index declines, but not from surrender charges, rider fees, taxes, inflation or insurer failure.
• Caps, participation rates and spreads give up part of the upside by design, and the result against the index depends on the path of the market.
• Credits are generally measured on the anniversary, so a mid-year exit may earn little or nothing for that year.
• A guaranteed roll-up usually applies to an income base that is not a lump sum you can withdraw.
• Check whether the index excludes dividends, treat back-tests as hypothetical, and expect caps and rates to change on renewal within the contract's limits.