Public Pension Funds: Asset Allocation Strategies

by Karl C. Mergenthaler, CFA and Helen Zhang
J.P. Morgan Investment Analytics & Consulting
karl.c.mergenthaler@jpmorgan.com , helen.x.zhang@jpmorgan.com

The U.S. public pension system is facing daunting challenges. The current economic malaise is creating significant financial stress for state and local governments. In this environment, an estimated 80% of state and local workers are covered by a Defined Benefit (DB) plan, most of whom rely solely on their DB plan for retirement savings. Yet, most public pension plans are under-funded.

In order to meet their long-term obligations, the average expected return for large public pensions is 8%. However, if we apply reasonable capital market assumptions to the average asset allocation of these plans, it is clear that these goals will not be met easily. In our view, plan sponsors will need to enhance their performance through superior asset allocation.

In this article, we discuss and analyze the current asset allocation of public pension plans. Furthermore, we provide a brief summary of Modern Portfolio Theory, as well as several new approaches to the asset allocation dilemma. In this way, we provide a context for decision-makers to refine and manage their asset allocations in these challenging times.

The Expected Return Gap

In Exhibit 1, we summarize the typical asset allocation of a public pension plan, as well as the expected return for each asset class.

Exhibit 1 — Expected Return Gap

As the chart indicates, we estimate that the weighted average expected return of a typical public pension plan is approximately 7%, or 100 basis points shy of the median expected return of 8%.

Comparison of Corporate vs. Public Plans

In general, large corporate and public pension plans are broadly diversified. The asset allocation is impacted by factors including the size of the plan, funded status, management resources and other considerations. In Exhibit 2, we summarize the typical asset allocation for large corporate and public pension plans.

Exhibit 2 — Average Asset Allocation

Source: J.P. Morgan Investment Analytics & Consulting estimates

As shown, U.S. public pension plans tend to have higher equity exposure than corporate plans, particularly with respect to international equities. Over the past two decades, both corporate and public Defined Benefit pension plans have increased their exposure to international and emerging market equity.  Within fixed income, some large corporate plans have moved to a Liability Driven Investing strategic allocation, including high exposure to long duration fixed income. Both public and corporate plans have increased their exposure to alternative assets over the past two decades and are likely to continue this trend going forward.

Importantly, the size of the plan tends to have a significant impact on asset allocation. Larger plans tend to have greater exposure to alternatives. In many cases, large plans are better equipped to handle the illiquidity of alternative assets, leveraging resources such as experienced staff with expertise in selecting and monitoring private equity and hedge fund managers.

Asset Allocation Options

The financial crisis of 2008 and 2009 forced pension plan sponsors to re-consider their approach to asset allocation. Investors are looking for new and creative ways to set an allocation with the goal of meeting return expectations.

Modern Portfolio Theory: Modern Portfolio Theory, or “MPT,” is a time-tested and classic approach to portfolio construction. MPT was first outlined in 1952 by Harry Markowitz in his paper “Portfolio Selection.” Investors use a mean variance optimizer to build portfolios along an “Efficient Frontier.”  Of course, the recent credit crisis exposed the flaws of MPT and questioned its ability to provide diversification and reduce investment risk.

Although MPT did not fail, the usefulness of the model was tested during the past two years. Markets were far from efficient, correlations all grew close to 1.0, and many securities became illiquid. The departure from efficient markets compromised MPT’s advertised diversification benefits and provided less shelter to investors in their time of need.

Economic Regimes: Many institutional investors are considering the concept of building portfolios based on expected performance of asset classes in various macro-economic “regimes.” In this approach, practitioners consider the future direction of macro-economic factors such as inflation and growth. Accordingly, investors can attempt to establish their asset allocation and take tactical positions based on their projections about future economic conditions.

In a similar fashion, investors may take into account other economic factors, including unemployment rates, industrial production, construction spending, Fed Funds rate, manufacturing, and housing starts. The expected performance of each asset class in the various macro-economic regimes must be considered. Furthermore, investors must consider the expected correlations of asset classes to one another in each regime and the likelihood of future economic regimes to set asset allocations accordingly. In our view, investors must have an understanding of behavioral relationships, insight into economic cycles, causes of the turning points, and the ability to forecast correlations of assets.

Risk Parity:  The Risk Parity approach is based on the concept that true diversification can be achieved if each major asset class provides an equal contribution to risk. Advocates of this approach argue that the risk contribution of the traditional portfolio is skewed towards equities, and therefore fails to provide appropriate downside protection.  The methodology seeks to decrease the potential losses from individual asset classes, and has a return pattern that deviates from the norm.

In Exhibit 3, we illustrate the approximate asset allocation of a portfolio under the risk parity model, using J.P. Morgan Long-Term Capital Market Assumptions.¹

Exhibit 3 — Risk Parity Portfolio

As shown in Exhibit 3, one major characteristic of the risk parity portfolio is a heavy weighing towards fixed income, since the asset class tends to have less risk and lower correlations to other asset classes.

In order to generate adequate expected returns, many risk parity advocates suggest that pension plans employ leverage. However, the cost of borrowing may become a significant factor in the process, as both the higher allocation to fixed income and the use of leverage makes the portfolio more susceptible to shifts in interest rates. If rates begin to climb, leverage will become increasingly expensive while the market value of fixed income assets decline, greatly lowering the return enhancement incentive for employing a leveraged portfolio. In the event of a financial crisis, leverage could potentially amplify losses and cause extremely harmful consequences.

Liability Driven Investing: In recent years, corporate pension plans have seen a resurgence of liability driven investing, or “LDI.” In general, the concept of LDI suggests that asset allocation should consider the liability profile of the plan with the primary objective of lowering volatility in the funded status.   In general, LDI strategies tend to translate into portfolios with a significant allocation to long duration fixed income securities and an overlay of interest rate swaps to increase the duration of the overall portfolio. This methodology considers the economic behavior of liabilities and should result in portfolios with a more stable funded status.

In our view, there are several reasons that an LDI approach may not be appropriate for public pension plans. Importantly, public plans do not have to contend with the same accounting and regulatory constraints that corporate plans face. Moreover, since many public plans are currently under-funded, adoption of an LDI strategy would effectively lock-in the deficit and force the sponsors to make a contribution in the near term.

Black Litterman Model:  The Black Litterman Model is a technique that uses many of the concepts of Modern Portfolio Theory, and incorporates data related to actual investment dollars allocated to various asset classes. Please see Andrew Robertson’s article on Portfolio Optimization for a discussion of the Black Litterman Model.

Conclusion

In our view, asset allocation is both an art and a science. Institutional investors must balance experience and analytics to create the optimal portfolio. The recent financial crisis has brought unique challenges, and while Modern Portfolio Theory is constantly being questioned, newer allocation theories are also developing in the effort to find more effective solutions. More comprehensive data allows a better understanding of the cyclic nature of asset returns and how they interact with leading economic indicators, giving way to the regimes approach. Other frameworks such as Risk Parity and LDI suggest different solutions for addressing risk contribution and funded status by increasing fixed income allocations.

In our opinion, asset allocation is critical in generating the required returns of public pensions. Investors must carefully consider their long term strategic allocations, and while each new method provides an interesting perspective of managing the portfolio, they also introduce new return patterns and risks that require careful examination. The advent of more sophisticated financial modeling tools allows institutional investors to better assess varying allocation strategies as managers strive to constantly question and refine their assumptions and techniques.

To view the next article, Endowment & Foundation Spending Policies: One Size Does Not Fit All, click here.