Pension Fund Risk Management Prioritizes Benefit Payment Over Volatility

Deep Dive

Howard Marks recounts an experience attending a meeting with the board and senior staff of a U.S. state pension fund. He was invited to listen and offer feedback on a consultant's report based on a board member survey. The consultant's session impressed Marks due to its focus on what he considers the most important investment topics, often aligning with his own perspectives.

The discussion then shifts to attitudes toward risk, beginning with a two-by-two matrix presented by the consultant. This matrix maps a plan's ability to bear risk, defined by its financial health and that of its sponsor, against its willingness to bear risk, or its attitude toward taking on risk and withstanding potential losses. The four cells of this matrix are described as "capitalizing on" (high ability, high willingness), "defensive" (high ability, low willingness), "protective" (low ability, low willingness), and "naive" (low ability, high willingness). The consultant's survey indicated the board possessed a moderate willingness to accept risk despite an above-average ability to bear it, stemming from the plan's solid funding status and the state's strong economic performance. Marks found this an organized and intelligent approach to risk-bearing, recalling a similar situation when he chaired the University of Pennsylvania's investment committee.

Moving on from the matrix, the consultant explored other facets of the board's risk attitude. One hundred percent of board members agreed that exposure to risk is necessary to meet the plan's objectives. The consensus was that they would regret adopting an aggressive strategy and experiencing a market collapse more than being conservative and missing out on gains. The board expressed a strong preference for normal market participation risks over those associated with innovative but opaque approaches. All board members recognized that true diversification implies the presence of some underperforming assets within a portfolio at all times. Marks considered these views reasonable, noting the board's explicit preference for conservatism despite potential underperformance in strong markets, a reality they had observed over the preceding two bullish years. They accepted risk as unavoidable and understood that caution limits returns.

Next, the discussion turned to setting objectives. The consultant presented the board members' ranking of objectives: first, determining the correct asset allocation; second, hiring outperforming managers; third, beating the assumed rate of return; fourth, increasing risk at the right time; and fifth, outperforming peers. Marks was impressed that beating peers was ranked last, and the board strongly disagreed with the notion that it is acceptable to lose money as long as one does better than competitors. He cautioned against viewing investing solely as a competition, emphasizing that for a pension plan, success means being able to pay benefits and minimize costs to the sponsor. He also noted that a plan's ability to pay benefits does not automatically signify good management, as it could be due to favorable market conditions rather than skill, referencing Nassim Nicholas Taleb's concept in "Fooled by Randomness."

On the subject of volatility, the board members ranked the Sharpe ratio last among six performance metrics. They considered avoiding volatility in sponsor contributions less of a priority than the ability to pay benefits or attain full funding status. Most members sought a balance between stable contributions and pursuing high returns, though some prioritized contribution stability. Marks acknowledged this as a challenging question for boards but expressed his view that investors often pay too much attention to volatility, suggesting that the risk of permanent loss is more critical. He argued that volatility is often a concern due to situational, institutional, political, or career-related factors rather than an intrinsic investment risk, citing examples like daily-priced mutual funds versus sovereign wealth funds. He recognized, however, that for certain investors like pension plans and university endowments, volatility can be a material real-world risk due to its impact on contributions and operating budgets.

The consultant addressed the choice of investment approach, with the board agreeing that the portfolio should be built to prepare for all environments rather than relying on market timing. A substantial majority expressed comfort with using leverage on 15% to 20% of the plan's assets, which Marks deemed reasonable for a well-funded plan sponsored by a financially strong employer. He cautioned, however, about the risk of lenders pulling leverage and the importance of not holding assets with low returns below the borrowing cost. A slimmer majority supported allocating 25% of the portfolio to illiquid assets, though some felt surrendering flexibility was not justified by a higher promised return. Slightly more members favored focusing exclusively on expected returns net of fees, while a few thought fee minimization should be a goal in itself, a point Marks considered difficult as high fees are certain while performance is not guaranteed.

Regarding assessing performance, the consultant reported that the board members considered achieving the actuarial assumption the most important standard, with beating the policy benchmark and having managers beat their benchmarks as secondary. Outperforming peers and popular indices like the S&P 500 were deemed relatively unimportant. Marks agreed with these priorities, stating that the most crucial aspect is whether the plan achieves the actuarial rate of return necessary to meet future benefits. However, he pointed out the challenge of assessing performance over shorter periods for purposes like annual reviews, promotions, and personnel retention, as achieving the actuarial assumption can be irrelevant in a strong or weak market year. He argued that in the short to intermediate term, performance assessment must be relative, comparing results to peers or policy benchmarks, while acknowledging the deficiencies of each standard.

Marks elaborated on the appropriate period for performance assessment, suggesting that a fixed number of years is not ideal. The assessment period must be long enough to allow influencing factors to even out, for unusual events to dissipate, and to encompass both bullish and bearish environments, essentially covering a full market cycle. He explained that assessing performance over only good times might favor risk-prone portfolios, while beating benchmarks in declining markets might indicate risk aversion rather than superior risk-adjusted returns. Finally, he suggested the board consider the level of personnel turnover, as it can indicate issues with hiring, performance assessment, or management practices, although some turnover is understandable due to compensation limitations in the public sector.

In conclusion, Marks expressed his strong approval of the board's approach, noting their willingness to accept less than maximum risk, their preference for avoiding market declines, their limited concern for peer rankings, their minimal interest in volatility-adjusted metrics, and their focus on assessing the investment team and portfolio. He felt the board and consultants were asking the right questions and reaching reasonable conclusions, finding the session highly informative.