Strategic Asset Allocation: Liability Relative

On 31 December 2016, Bank A had a defined benefit pension plan asset value of USD 18,239,000,000. That's exciting news for retirees! The corporation, however, has a different viewpoint. These assets within the defined benefit pension plan must be properly managed for the benefit of employees. So the future value drives the asset-management process. Then what's the focus of the bank's asset managers?

No, actually. That's not the primary focus because excess growth can incorporate excess risk.

That's right! The bank's defined benefit pension plan has a single focus: meeting the future liability needs of the participants. That means that the best asset-allocation strategy is the liability-relative approach. The liability-relative approach is a smart strategy for pension funds because it aligns the interests of employees and the sponsoring employer.

That's not it. The bank's defined pension plan won't have multiple goals; it has one primary focus.

For example, employees participating in the bank's plan will receive benefits based on factors related to their work and compensation. But notice what's not included. What's one aspect of the defined benefit pension fund that doesn't impact the calculation of benefits paid to employees?

Clearly, yes! The rate of return on the plan isn't a work-related factor in the benefit calculation. From the employee's perspective, the benefits don't change based on the assets within the fund. So there's no emphasis on growing the fund above the necessary present value of liabilities. It's all about having the necessary funds available for retirees. It's a similar thought process for the employer. Taking on additional risk to grow the asset base consistently above the present value of liabilities will help lower future contributions to the plan but can also have the opposite effect. What's the downside risk to seeking excess returns?

No. That's a work-related factor that helps determine benefits.

That's not it. That's definitely a work-related factor that will contribute to the benefit calculation.

Not necessarily. Higher volatility isn't a bad thing if it actually leads to higher returns, so it's not always a drawback.

That's not it. Most plans don't allow for early withdrawals, so that's not a negative.

That's right. Plan sponsors may have to contribute more to the fund, which is a definite negative. So plan sponsors usually want the asset allocation to focus on either maintaining the asset base relative to the liability or growing the asset base to catch up to the liability. That's the primary focus of the liability-relative approach. So if you identify multiple portfolios that fulfill the bank's risk tolerance, VaR limit, time horizon, and liquidity needs, you'll need to evaluate the asset allocation in detail to determine which portfolio best serves the sponsor and beneficiaries.

For example, under the asset-only approach, a mean–variance optimization will produce a portfolio on the efficient frontier, so it will have the highest Sharpe ratio between portfolios that meet risk requirements. Now, compare that to a liability-relative approach, which will recommend a liability-hedged portfolio to match the interest rate exposure of the present value of the plan's liabilities, plus some additional risk exposure depending on the funding status of the plan. Given the differences between the two approaches, which allocation do you think would have a larger exposure to equities?

Excellent! The asset-only approach is going to use equities to grow the asset base per a specific level of risk. So equity exposure is usually higher. But that's not the focus of the liability-relative approach. Remember the goal is to match liabilities, so in most cases the liability-relative asset allocation won't lie on the efficient frontier developed from mean variance optimization.

No, actually. The liability approach isn't focused on maximizing return per level of risk.

No. The goals of each approach are different.

That's a key difference between the asset-only and liability-relative approaches. Another key difference is the usage of fixed income versus equities. As illustrated, equity exposure isn't the primary asset allocation focus in liability-relative allocation, yet equities shouldn't be forgotten. That's because they can be used to grow an underfunded pension back to fully funded status. This strategy is called __liability glide path__, which is a technique where the plan sponsor identifies the proportion of liability-hedging fixed income and return-seeking equity and the duration of the liability hedge as the funded status and contributions change. So as surplus risk is decreased, the path of the asset allocation changes. What asset class exposure do you think increases over time?

No. Surplus risk is decreasing, not increasing.

That's right! As surplus risk decreases, the volatility of the funded status also decreases. So more fixed-income exposure can be used to align the fund with a liability-hedged portfolio. This also increases the confidence in cash flow planning as the volatility of contributions will decline. Speaking of the liability-hedged portfolio, given its focus on liability matching, the asset allocation will have specific risks like interest rate sensitivity, inflation, and credit risk. So risk-factor modeling is important when using the liability-relative approach.

That's not it. Investments need to earn a return to keep up with inflation.

To summarize: [[summary]]

Growing the asset base

Meeting the future liability needs

Achieving short-term liquidity to pay benefits and long-term objectives of growth

Rate of return on the fund

Years spent contributing to the fund

Amount of salary earned at the company

More volatility in the fund

A higher number of withdrawals

Higher required contributions in the future

Asset only

Liability relative

Similar allocation for both approaches

Equities

Fixed income

Cash and equivalents

Continue

Continue

Continue

Continue