Annuity/Life Insurance Arbitrage: Turning Your Good Health into Income, Estate Efficiency, and Charitable Leverage
The concept behind annuity/life insurance arbitrage is remarkably straightforward. A life insurance company and an annuity company are both pricing the same event—the duration of one person's life—but they approach that event from opposite economic directions. The life insurer assumes an obligation that becomes payable when the insured dies, while the issuer of a life-only immediate annuity assumes an obligation that continues for as long as the annuitant remains alive. If the two companies price that mortality risk differently enough, an economic spread can arise. The client can potentially capture that spread by purchasing both contracts and treating them as a single financial transaction.
Consider a simple example. An individual commits $1 million to a single-premium immediate annuity, or SPIA, and simultaneously arranges for $1 million of permanent life insurance. The SPIA provides guaranteed lifetime income. A portion of that income is used to pay the life insurance premium, and the client retains the remaining cash flow. When the individual dies, the life-only SPIA terminates, while the life insurance pays $1 million. If, after income taxes, the annuity provides $140,000 annually and the cost of maintaining the $1 million life insurance policy is $60,000, the client retains $80,000 per year. Economically, the client has committed $1 million, receives an 8% annual net cash flow, and has arranged for $1 million to be restored at death.
The critical analytical point is that the annuity and the life insurance cannot sensibly be evaluated as independent investments. They were deliberately assembled to create a single economic position. The original analysis of this strategy correctly identified the underlying opportunity as the different mortality assumptions that can be embedded in annuity and life insurance pricing on the same individual. Once the products are combined, the relevant question is not what the SPIA earns by itself or what the life insurance costs by itself. The relevant question is what the combined transaction produces after taxes and the cost of maintaining the capital-replacement insurance.
Why the Arbitrage Can Exist
Life insurance and life annuities occupy opposite sides of mortality risk. Other things being equal, a life insurance carrier benefits economically when an insured lives longer because payment of the death claim is postponed. A carrier issuing a pure life-only SPIA experiences the reverse economics because it must continue making payments for as long as the annuitant survives. In a theoretically perfect market, these opposing exposures might be priced so consistently that there would be little or no spread to capture. Actual insurance markets are not that uniform.
Life insurance underwriting can be intensely individualized. Medical records, laboratory results, medications, cardiovascular history, tobacco use, build, family history, avocations, and numerous other factors can influence an underwriting classification and therefore the cost of insurance. A traditional commercial SPIA generally does not reward exceptionally favorable individual health with the same underwriting granularity. Its payout is primarily determined by age, contract design, prevailing economic conditions, the carrier's mortality experience, and its pricing assumptions.
This difference becomes particularly interesting for an older individual who is healthier than the average person of the same chronological age. A life insurance carrier may recognize favorable longevity and offer a correspondingly favorable premium, while a commercial annuity carrier may still offer a relatively high payout associated with the client's chronological age. Carrier-to-carrier differences in mortality experience, capital requirements, investment assumptions, competitive positioning, and product design can widen the discrepancy further. The result is what I regard as a form of mortality arbitrage: one market is willing to pay the client attractively for remaining alive while another market is willing to insure the client's death at a sufficiently low cost to leave a substantial spread.
The Integrated Economics
One criticism frequently directed at immediate annuities is that their payout rate should not be confused with an investment rate of return because part of each payment represents the return of the purchaser's own principal. That observation is perfectly valid when the SPIA is analyzed in isolation, but it becomes considerably less useful when the annuity is deliberately combined with an equivalent amount of life insurance.
If an individual commits $1 million to a SPIA, receives lifetime cash flow, and does nothing to replace the capital at death, the distinction between investment earnings and liquidation of principal is obviously important. The individual is economically consuming the original capital over his or her lifetime. In the arbitrage transaction, however, the life insurance performs the capital-replacement function. The original million dollars is exchanged for lifetime income, and part of that income is deliberately used to maintain another contract that restores $1 million when the annuity obligation ends.
Using the earlier example, if the after-tax annuity cash flow is $140,000 and the guaranteed cost of maintaining the $1 million death benefit is $60,000, the client's economic spread is $80,000 annually. Relative to the original $1 million commitment, that is an 8% annual cash-flow yield. At death, the annuity terminates and the life insurance pays the $1 million death benefit. Whether a portion of the gross annuity payment represented interest, mortality credits, or return of basis certainly matters for taxation, but it does not negate the economics of the integrated position.
This same reasoning applies to embedded carrier expenses. The annuity carrier has acquisition costs, compensation, reserves, administrative expenses, mortality assumptions, investment expenses, and a profit margin. The life insurer has mortality charges, acquisition expenses, commissions, reserves, administrative costs, and its own profit margin. Those expenses matter because they affect the prices the two companies offer. Once the annuity payout and the required life insurance premium have been contractually established, however, those embedded costs have already entered the transaction. Deducting them again from the resulting spread would amount to counting the same economic costs twice.
The professional analysis should therefore concentrate on the numbers that actually change the client's outcome. Those include the after-tax annuity payment, the premium required to maintain the promised death benefit, any external legal or administrative expenses, financing costs if leverage is used, and any nonguaranteed contractual assumptions that could alter those amounts in the future. How the carriers internally manufacture their respective prices is secondary once the contractual economics are known.
FTER and the Tax Efficiency of the Annuity
The federal income-tax treatment of the SPIA can materially improve the cash-flow economics. For purposes of this analysis, I use the term Federal Tax Exclusion Ratio, or FTER, as shorthand for the exclusion ratio applicable under Internal Revenue Code §72. FTER is a descriptive term rather than the formal statutory name; the Code and IRS guidance refer to the concept simply as the exclusion ratio.
Under §72, payments from a qualifying nonqualified annuity are generally divided between taxable income and recovery of the owner's investment in the contract. The exclusion ratio is based on the relationship between the investment in the contract and the expected return under the contract. IRS Publication 939 describes the calculation as dividing the investment in the contract by the expected return to determine the exclusion percentage.
Assume, for example, that a $1 million SPIA provides $150,000 of annual lifetime income and that the applicable FTER results in 60% of the qualifying payment being treated as recovery of investment during the basis-recovery period. The client receives the entire $150,000 in cash, but under this simplified example only $60,000 is currently taxable while $90,000 represents excluded recovery of basis. The significance to the arbitrage is obvious. The life insurance company does not care about the tax character of the dollars used to pay its premium. The entire cash payment is available to fund the policy even though only part of the annuity distribution may initially produce current taxable income.
The FTER therefore increases the amount of spendable after-tax cash available from the SPIA during the basis-recovery period. It does not convert return of principal into investment earnings, nor does it need to. Its value lies in improving the tax efficiency of the cash flow that funds the other side of the arbitrage.
The exclusion is not unlimited. For annuities with post-1986 starting dates, the aggregate amount excluded cannot exceed the owner's unrecovered investment in the contract. Once basis has been completely recovered, subsequent payments generally become fully taxable. A serious arbitrage analysis should therefore project the after-tax spread both during the FTER period and after basis has been exhausted. If the annuitant lives substantially beyond the actuarial recovery period, the annual after-tax spread can decline even though the cumulative economic benefit of continuing to receive lifetime annuity payments continues to increase.
A less frequently discussed tax consequence also deserves attention. Section 72 provides a deduction in certain circumstances when annuity payments cease at death before the annuitant has recovered the entire investment in the contract. Thus, where a life-only annuitant dies earlier than the actuarial assumption and unrecovered basis remains, the tax law can provide relief for that unrecovered investment. This further illustrates why the tax consequences should be modeled across different longevity outcomes rather than reduced to a single life-expectancy projection.
The Estate-Tax Dimension
For a high-net-worth client, the estate-planning consequences can be even more important than the annual cash-flow spread. Assume the client begins with $1 million of conventional investment assets. If those assets remain owned at death, their date-of-death value generally remains part of the client's gross estate. The pure life-only SPIA behaves very differently because the client has exchanged the capital for a contractual lifetime payment right. At death, assuming there is no refund feature, period certain, or other continuing death benefit, that payment right terminates.
This is an important feature, not an incidental consequence. The original $1 million does not remain in a conventional account awaiting transfer at death. The asset's economic benefit has been consumed through the lifetime payments. The life-only SPIA therefore can compress the estate because the contractual asset that generated the income disappears with the annuitant's death.
The life insurance operates in precisely the opposite direction. Section 101 generally excludes qualifying life insurance proceeds paid by reason of death from the beneficiary's gross income. For estate-tax purposes, §2042 generally includes life insurance proceeds when they are receivable by the insured's estate or when the insured possesses incidents of ownership in the policy at death. Treasury regulations describe incidents of ownership broadly and include powers such as changing beneficiaries, surrendering or assigning the policy, pledging it for a loan, or borrowing against it.
If the insurance is properly structured from inception outside the insured's estate, such as through an appropriately designed irrevocable life insurance trust, and the insured retains no incidents of ownership, the death benefit can potentially remain outside the gross estate. Existing policies transferred out of the insured's ownership require additional attention because §2035 can cause certain transfers or relinquishments involving life insurance rights within three years of death to be pulled back into the estate-tax calculation.
The combined result is particularly compelling. The SPIA, an asset economically associated with the insured during life, terminates at death. The replacement capital supplied by the life insurance can arise outside the insured's taxable estate. The transaction therefore does more than replace principal. It potentially changes the tax location of the replacement principal.
If the alternative is retaining $1 million of investment assets inside a taxable estate, the correct comparison is not between $1 million of investments and $1 million of insurance on a pre-tax basis. It is between the amount the family ultimately receives from the conventional asset after applicable estate taxes and the amount it receives from properly structured insurance outside the estate. For sufficiently large estates, that difference can materially exceed the value of the annual mortality spread itself.
Why the Form of the SPIA Matters
The estate-compression argument depends upon using the appropriate annuity design. A life-only SPIA without a refund or period-certain feature maximizes the mortality component because the carrier's obligation ends when the annuitant dies. If the contract contains a cash-refund benefit, installment-refund provision, or guaranteed payment period, value may continue after death and pass to a beneficiary. That may be desirable in ordinary annuity planning, but it changes the economics of this particular strategy.
Where separately owned life insurance has already been selected to replace the capital at death, purchasing extensive death protection within the annuity itself may also be inefficient. The client would effectively be paying the annuity carrier to preserve some of the principal after death while simultaneously paying a life insurer to accomplish the same capital-replacement objective. The arbitrage becomes conceptually cleaner when each market is allowed to perform the function for which it is being selected: the life-only commercial annuity is used to maximize lifetime income, while the life insurance provides the legacy.
The Risks That Actually Matter
Treating the two products as an integrated transaction does not make the strategy risk-free. It does, however, help distinguish genuine economic risks from accounting details that have already been reflected in product pricing. Liquidity is the most obvious limitation because the life insurance replaces the original capital at death, not during life. A client who may need to recover the $1 million principal several years after implementation should not exchange it for an irrevocable life-only income stream merely because the calculated spread is attractive.
The quality of the insurance guarantee is equally important. If the assumed premium required to maintain the $1 million death benefit depends upon nonguaranteed crediting rates, favorable index performance, or future policy values that may not materialize, the apparent spread has not actually been locked in. A policy with a higher contractual premium but a stronger lifetime guarantee can therefore be superior to a policy with a more attractive illustrated premium. The purpose of the insurance leg is capital replacement, and the credibility of the entire arbitrage depends on that replacement actually occurring.
Carrier solvency and concentration are also real considerations because both sides of the transaction depend upon insurance-company claims-paying ability. Inflation presents another limitation because a fixed $1 million death benefit twenty years from now replaces $1 million nominally, not the purchasing power originally represented by $1 million. The same problem applies to fixed annuity payments. Finally, the economics are highly sensitive to life insurance underwriting. If the client receives a poor underwriting classification, the increased premium can reduce or completely eliminate the mortality spread.
Those are legitimate objections because they can alter what the client actually receives. They are fundamentally different from pointing to commissions or carrier expenses that have already been incorporated into the contractual payout and premium.
Financing the Transaction
The arbitrage can also be leveraged by financing some or all of the capital used to acquire the SPIA. In that case the annuity cash flow must support income taxes, the life insurance premium, borrowing costs, and any required principal repayment or collateral expenses. If the underlying mortality spread materially exceeds the financing cost, leverage can produce an unusually attractive return on the client's actual capital committed to the transaction.
Financing also changes the nature of the risk. Floating borrowing rates, collateral requirements, lender covenants, renewal terms, and potential personal or trust guarantees become important. Historical discussions of this strategy frequently used LIBOR-based financing, but contemporary transactions must use the actual contractual benchmark and lender spread rather than obsolete LIBOR assumptions. The income-tax treatment of borrowing must also be modeled conservatively because §264 restricts deductions for interest associated with certain indebtedness involving life insurance and annuity contracts. Financing should therefore enhance a transaction that already works economically rather than being used to create an apparent arbitrage that disappears if interest rates or collateral requirements change.
The Charitable Application
The same mortality-arbitrage principle can be extended to charitable planning, and this may be one of its most useful applications. Instead of using life insurance to replace capital for descendants, the donor can purchase and own the commercial life-only SPIA while a qualified charitable organization separately applies for, owns, and is beneficiary of life insurance on the donor, assuming applicable state insurable-interest law and the charity's gift-acceptance policies permit the arrangement.
The ownership distinction is important. Merely naming a charity as beneficiary of a life insurance policy that remains owned by the donor does not convert the donor's personal premium payments into currently deductible charitable contributions. The cleaner structure is for the charity to own the policy and receive the economic benefit, while the donor makes completed charitable contributions to the qualified organization. Subject to §170, applicable percentage limitations, substantiation requirements, and other tax rules, contributions of money to qualified charitable organizations can be deductible. The charity can then use those contributed funds to maintain the insurance it owns.
This creates an unusually attractive interaction with the SPIA. The donor receives guaranteed lifetime income from a commercially priced annuity, and the FTER causes a portion of qualifying annuity payments during the recovery period to be treated as return of basis rather than current taxable income. The donor then contributes a portion of the cash flow to the charity and may obtain a charitable income-tax deduction for that contribution. The charity uses the funds to maintain life insurance on the donor, and at the donor's death the charity receives the insurance benefit.
The donor therefore does not have to choose between lifetime income and a substantial charitable legacy. The commercial annuity provides the income, while the charity-owned life insurance provides the legacy. The tax law can add efficiency on both sides because the annuity benefits from the §72 exclusion-ratio treatment while qualifying contributions to the charity may produce deductions under §170.
Why a Commercial SPIA Can Be More Efficient Than a Charitable Gift Annuity
The appropriate comparison is with a charitable gift annuity, or CGA, which is a valuable but economically different instrument. In a CGA, the donor makes an irrevocable transfer to a charity in exchange for the charity's contractual promise to make lifetime payments to one or more annuitants. Part of the initial transfer represents the actuarial value of the annuity obligation and part represents the charitable gift.
A charitable gift annuity is intentionally designed to leave a meaningful amount for charity after the annuity obligation terminates. The American Council on Gift Annuities states that its suggested maximum rates are generally designed to produce a target residuum for charity equal to approximately 50% of the original funds contributed, and its current methodology also incorporates assumptions concerning investment returns, expenses, and mortality. This is not a weakness in a charitable gift annuity; it is an essential part of its charitable purpose. It does, however, mean that the CGA is not designed solely to maximize the donor's lifetime income.
A commercial SPIA is priced for a different objective. The insurance carrier is competing in the commercial annuity market and has no obligation to preserve a charitable residuum. Particularly with a pure life-only annuity, the carrier can devote the economics of the contract to producing the highest commercially supportable lifetime payment. For the right age and pricing environment, that can make a commercial SPIA materially more efficient for the donor's income objective than a charitable gift annuity.
The charitable arbitrage therefore separates functions that a CGA combines. The commercial insurer is asked to maximize guaranteed lifetime income, while the life insurer is asked to create the charitable death benefit. The charity owns the insurance and receives the eventual proceeds, while the donor retains the commercial annuity and makes charitable contributions during life to support the policy.
Consider a simplified illustration in which a charitably inclined donor has $1 million available. Instead of transferring the entire $1 million to a charity in exchange for a CGA, the donor uses the capital to purchase a commercial life-only SPIA. Assume the annuity produces $150,000 annually and, after application of the FTER and applicable income taxes, the donor has $130,000 of spendable cash. If the donor contributes $50,000 annually to a qualified charity and that amount is sufficient for the charity to maintain a $1 million life insurance policy it owns on the donor, the donor retains $80,000 annually while also potentially receiving the charitable deduction attributable to the qualifying $50,000 contribution. At death, the commercial SPIA terminates and the charity receives the $1 million insurance benefit.
The appropriate comparison with a CGA is therefore not merely between two published payout percentages. It is between the complete economic outcomes. The analysis should compare the donor's after-tax lifetime income, the tax treatment of the annuity under the FTER, the charitable deductions generated during life, and the amount ultimately received by the charity. A charitable gift annuity may remain preferable where simplicity, lack of medical underwriting, an immediate charitable gift component, or other planning considerations dominate. Where the donor is older, favorably insurable, wants substantial guaranteed income, and also intends to make a significant charitable gift at death, however, the commercial SPIA and charity-owned life insurance combination deserves serious comparison.
The Broader Planning Principle
What makes annuity/life insurance arbitrage interesting is not any particular insurance product. Its value lies in separating economic functions and allowing different markets to price each function independently. The annuity carrier prices longevity and provides lifetime income. The life insurer prices mortality and provides capital at death. An ILIT can position that capital outside the insured's estate for descendants, while a charity can own the insurance when the intended legacy is philanthropic. Federal income-tax law treats the annuity cash flow, charitable contributions, and life insurance proceeds differently, creating additional opportunities for tax efficiency when the structure is properly designed.
For the family wealth-transfer client, the SPIA can generate lifetime income while its disappearance at death reduces the amount of property remaining in the estate, and properly structured life insurance can replace the capital outside the estate. For the charitable client, the same SPIA can generate guaranteed lifetime income while charitable contributions fund charity-owned insurance that ultimately produces the philanthropic legacy. In both cases, the structure uses one financial institution to maximize the value of remaining alive and another to maximize the value delivered at death.
The correct comparison is therefore never simply whether a stock portfolio could earn more than the annuity or whether a charitable gift annuity has a lower stated payout than a commercial SPIA. The comparison must be between complete economic outcomes. An alternative strategy must be evaluated on the income it provides during life, the taxes imposed on that income, the liquidity retained by the client, the value ultimately available at death, and the income and estate taxes imposed before that value reaches the intended family or charitable beneficiary.
Conclusion
Annuity/life insurance arbitrage is ultimately an exercise in exploiting differences in the way markets and the tax system value longevity, mortality, income, and transfers at death. A client commits capital to a life-only SPIA and receives guaranteed lifetime income. The FTER can reduce the current taxation of part of that income during the basis-recovery period. A portion of the annuity cash flow maintains life insurance, and the client retains the remaining spread. At death, the life-only annuity terminates, while the life insurance replaces the capital that was committed to the annuity.
When the insurance is properly owned outside the insured's estate, the transaction can simultaneously reduce the value remaining inside the taxable estate and create replacement capital outside it. When the charitable version is used, the donor retains the commercial lifetime annuity, makes potentially deductible contributions to a charity that owns the life insurance, and the charity ultimately receives the death benefit. Because a commercial SPIA is designed to maximize commercially supportable lifetime income rather than preserve a charitable residuum, this structure can, for the appropriate donor, provide greater lifetime income than a conventional charitable gift annuity while still producing a substantial and contractually defined charitable legacy.
The strategy is not appropriate merely because an annuity payout looks attractive, nor should it be rejected simply because part of the annuity payment represents recovery of basis. The meaningful inquiry is whether the complete transaction produces an attractive after-tax lifetime cash flow, whether the life insurance reliably replaces the committed capital, whether the client can tolerate the loss of liquidity, and whether the amount ultimately reaching the family or charity compares favorably with realistic alternatives.
When those conditions are satisfied, annuity/life insurance arbitrage is not simply the purchase of two insurance products. It is an integrated piece of financial, tax, estate, and charitable planning in which the different economic treatment of life and death is deliberately used to create a result that neither contract could produce as efficiently on its own.