CAIA® pensions: why DB and DC plans invest differently

On this page

A defined benefit plan invests to meet promised retirement payments, so the sponsor must consider assets alongside the value and timing of liabilities. A defined contribution plan places contributions into individual accounts whose eventual benefits depend on account outcomes. The investment objective changes because the person or institution bearing investment and longevity risk changes; a shared pension label does not make the portfolios interchangeable.

Begin with the promise

In a defined benefit, or DB, arrangement, the plan formula defines a retirement benefit. Future payments are obligations to be financed. The sponsor and trustees therefore ask whether invested assets can meet those payments under uncertain returns, interest rates and life spans. In a defined contribution, or DC, arrangement, the contribution is defined while the final benefit is uncertain. The participant's account rises or falls with contributions, market performance, expenses and withdrawals. CFA Institute's institutional-investor curriculum describes this distinction and identifies funded status, sponsor strength and workforce characteristics as DB risk considerations.

An exam vignette may call both structures pension funds, then ask for an investment objective. Identify the promised quantity before selecting an asset mix. If the vignette provides a benefit formula and future cash payments, think in terms of a liability-backed pool. If it provides an individual account with contributions but no guaranteed retirement payment, think in terms of a member's retirement resources and investment choices. Some real arrangements share risks, so read the actual promise rather than forcing every plan into a pure textbook category. The OECD's pension-system overview notes that hybrid forms exist.

The DB objective is measured against liabilities

A DB fund can show a positive investment return while its funding position worsens. The reason is that liabilities also change. Lower discount rates can increase the present value of future benefits, even if the asset portfolio earns money. Longer expected payment periods can have a similar effect. A manager who is judged only against a market index may therefore appear successful when the sponsor's ability to meet promises has weakened. The more relevant question is how assets behave relative to the cash flows they are meant to cover.

The funded ratio is the value of plan assets divided by the present value of plan liabilities. It is a snapshot, not a forecast. Consider an invented plan with £90 million of assets and £100 million of measured liabilities. Its ratio is 90%, and the difference is £10 million. If assets grow to £95 million while a discount-rate change lifts measured liabilities to £110 million, the ratio becomes about 86%. The asset gain did not close the gap because the obligation grew faster. These figures are illustrative and say nothing about an actual scheme or the contributions required under any jurisdiction's rules.

This is why liability-driven investing is a meaningful phrase in DB analysis. Assets such as bonds may be selected partly for how their value and cash flows respond to the factors that move liabilities. That does not mean a bond portfolio eliminates all risk. Benefit payments need liquidity, inflation provisions may matter, and market values can change for reasons unrelated to the liability discount rate. Return-seeking assets may still have a role, but the objective is not simply to maximise expected return. It is to support promised payments at a cost and risk that the sponsor can bear.

Sponsor strength changes the judgement. A financially robust sponsor can generally absorb an unexpected contribution more readily than a weak one. Yet an underfunded scheme with a weak sponsor is not automatically a licence to chase high returns; a further loss could worsen an already fragile position. CFA Institute's 2026 institutional-investor reading lists funded status and sponsor financial strength separately because both matter. They are inputs to risk capacity, not a mechanical asset-allocation formula.

The DC objective is personal and time-dependent

A DC plan has no single promised payment to match across all members. Each person's account must support a future use, but participants differ in age, other assets, retirement plans and ability to tolerate loss. A default investment arrangement can offer a practical starting point, often changing its risk exposure over time, yet the default cannot turn market returns into a guaranteed pension. A near-retirement participant may be especially exposed to a market loss just before withdrawals begin; a younger participant may face a long period in which insufficient growth or contributions erode the adequacy of eventual income.

The right measure for a DC participant is not a DB funded ratio. The participant asks whether saving and investment outcomes are on track for a spending objective under uncertain returns and lifespan. Even that estimate depends on assumptions about future contributions, inflation, retirement timing and drawdown. A plan sponsor must design and oversee choices with members in mind, but it does not usually make up every member's investment loss. The OECD's work on investment risk and pensions describes how DC investment risk reaches retirement income and how other parts of a retirement system may affect the final outcome.

Imagine two invented employees whose employers each contribute £5,000 this year. For a DB member, that employer cash flow helps finance a promise defined by the plan; the contribution alone does not set the member's benefit. For a DC member, the same £5,000 is added to an account and becomes part of that person's invested balance. If investments fall by 10% immediately, the simplified DC balance attributable to that payment becomes £4,500 before fees or further contributions. The DB plan's assets may also decline, but the member's formula benefit does not automatically become £4,500. The sponsor faces a funding problem instead. This thought experiment isolates who initially bears the market shock; actual plan terms and law remain relevant.

Longevity changes the objective again

Investment risk is only part of retirement uncertainty. A DB scheme promising lifetime payments must plan for beneficiaries living longer than assumed. Its obligation can grow even without a market move. A DC member drawing from an account must manage the chance that withdrawals last longer than expected. An annuity or other arrangement may transfer some of that risk, but the vignette must say that one exists. Do not infer a lifetime guarantee merely from the word pension. The OECD's analysis of pension risk sharing distinguishes how different arrangements distribute investment and longevity risks.

The time horizon also differs from the simple phrase long term. A DB scheme may have obligations extending far into the future, but it still needs cash to pay today's retirees. A mature plan with many payments due soon has different liquidity needs from a young plan with more active contributors. For DC, an individual approaching retirement does not necessarily stop investing on the retirement date; withdrawals may run over many years. The candidate should distinguish an investment horizon, a payment schedule and the capacity to absorb interim loss. Treating all three as the same date leads to weak portfolio reasoning.

Where US pension insurance fits

The Pension Benefit Guaranty Corporation describes coverage for many private-sector US DB plans, subject to rules and legal limits, and says DC plans are not insured by it. Its coverage guidance is useful context for the curriculum distinction, though PBGC does not make every DB promise unlimited or every public pension eligible. Insurance affects the consequences of sponsor and plan failure for covered participants. It does not remove the sponsor's need to manage funding or excuse an analyst from considering the plan's actual obligations.

When answering a CAIA case, state the jurisdiction. A US private DB plan may have a PBGC backstop; a plan elsewhere operates under another legal and insurance framework. A DC account has no PBGC guarantee simply because its employer also sponsors a DB plan. Nor should the candidate treat an insured DB promise as a market-traded asset with no residual risk. Coverage conditions, benefit limits and plan status matter. The safe conceptual point is narrower: the DB structure contains a promised benefit and associated sponsor-level funding risk, while the pure DC structure leaves the investment outcome with the member.

A decision checklist for exam cases

Ask what is promised, who must pay if investments disappoint, and who bears the risk of a longer retirement. Then compare assets with the correct target: DB assets against present and future benefit obligations, or DC assets against each participant's future spending needs. For DB, check funded ratio, liability sensitivity, sponsor capacity and payment liquidity. For DC, check contribution adequacy, member horizon, default design and drawdown uncertainty. Avoid prescribing a fixed allocation from the plan label alone; the vignette must supply the facts that justify a specific decision.

The most common error is to see a high expected return as proof that a DB policy is suitable. An apparently attractive return may come with asset losses precisely when funding or sponsor finances are strained. A second error is to quote a funded ratio for a DC account as though there were one common promised liability. A third is to assume that pension insurance guarantees DC market returns. The CAIA programme page gives the wider study context, and Omni's how-it-works page explains its learning format. For this concept, the efficient revision test is to follow the risk from the portfolio to the party who must ultimately absorb it.