Maverick risk now dominates fiduciary risk in its impact on the asset mix policy decision. This prevailing risk to plan sponsors means the possibility of underperforming other funds, rather than the possibility of an outcome that has a direct adverse financial impact on a pension plan's stakeholders (e.g., an unexpected jump in the plan contribution rate, or worse, an inability to meet future pension obligations).
Maverick risk is hard to ignore
Many fiduciaries are currently focusing on maverick risk rather than fiduciary risk in assessing what the asset mix policy should be. Why? It probably has a great deal to do with the fact that the great equity bull market has made the fiduciary risk perspective too hypothetical to receive the serious consideration of many fiduciaries. It gets tiresome (even embarrassing) to keep hedging against bad investment outcomes which never seem to happen.
If your equity weighting has been below that of your peers, your relative fund return has suffered. This just adds to the embarrassment. Nobody likes to be in the fourth quartile. Now add some rearview mirror watching to this dynamic.
In other words, if realized equity returns have been high, should we not raise future equity return prospects too? So if market action has raised our fund's equity-debt ratio by five points (say from 60-40 to 65-35), maybe we should simply raise our new policy weights by five points, too, rather than rebalance.
Fiduciaries should have a clear understanding of how risk and reward is apportioned between their various "pension deal" stakeholders, especially sponsors and participants. Their sole focus should be to find the asset mix policy that provides the best balance between financial "gain and pain" for those stakeholders.
Fiduciary risk or maverick risk?
Unfortunately, such a textbook-clear decision context seldom exists. "Pension deals" are often fuzzy rather than clear. The incentive for fiduciaries to do the right thing is often further weakened by the way results are measured and reported.
If fiduciaries believe, for example, there are strong sanctions attached to "fourth quartile" results relative to other funds, the typical asset mix policy of other funds will drive asset mix policy.
Understanding why maverick risk has become so dominant is one thing, condoning it is quite another. Only the fiduciary risk context will withstand the test of time as the right one for fiduciaries to focus on.
Pension fund investing should not be about winning (or not losing) horse races. Instead, it should always be about maximizing long-term net fund return for stakeholders, subject to clear risk constraints. This is not just a legal obligation. It is a moral obligation, too.
The pension plan balance sheet can be exposed to two kinds of risks: asset mix policy risk and implementation risk. One of the critical responsibilities of a fund's governing and managing fiduciaries is to understand and control both kinds of risks.
Exercising this responsibility requires periodically assessing (a) what the potential payoff is for undertaking each of the two kinds of risks, and (b) what the maximum allowable balance sheet exposure for each kind of risk is to be. I have been encouraging fiduciaries to adopt the "value at risk" framework to facilitate this responsibility.
A risk-control framework
The essence of the value at risk, or VAR, framework for controlling risk on defined benefit balance sheets is estimating how much money is at risk, and thinking through the minimum return that needs to be earned to justify that risk. Making estimates and judgments isn't easy. Estimates will never be more than just estimates; human judgment is always subject to debate before the fact, and second-guessing after the fact.
It is nevertheless a powerful finding that the average pension fund undertakes an implementation VAR equal to about 4% of fund assets each year, implying that this amount of money will be lost once every 20 years. What cost of risk capital charge should be applied to the estimated VAR?
To get the conversation going, we at my firm have been suggesting 13%. This seems to us to be the required rate of net return on a portfolio of high-risk investments today. The implication is that at the total fund level, the policy implementation process must produce a net value added of at least 50 basis points per year (i.e., 0.13 x 4% for the average fund).
Can we use this framework to create a platform for more informed asset mix policy decisions as well?
A series of 'shorter terms'
What is the relevant time horizon for addressing the asset mix policy question? It is only partially right to argue that because pension liabilities are very long, the investment time horizon should be very long, too. True, the benefit of earning excess return only materializes over long investment horizons as it compounds. Also true, temporarily depressed securities prices offer attractive buying opportunities for long-term investors. But that is not the whole story. A fund gets to the long term only by surviving shorter term setbacks.
In other words, the long term is the accumulation of a series of shorter terms. Within each of these shorter terms, fiduciaries must do all they can to ensure that (a) the pension plan remains solvent, and (b) that the stakeholders underwriting possible asset shortfalls understand what their exposure is, and that they are able and willing to make good on them through increased contribution rates. It is our experience that these considerations make three to five years a good investment horizon to decide on the proper asset mix policy for most pension plans.
Even though the current 2% equity-bond risk premium is low by historical standards, it is still worth earning if it reflects lower risk "new world" realities. After all, every 1% of additional long-term return on assets can reduce pension expense by 10% to 20%, depending on the duration of pension liabilities. Earning an added 1% long term can mean the difference between affordability and unaffordability of a target pension benefit.
But we have also noted the fiduciary risk context requires stress-testing the pension plan balance sheet for financial exposure over the relevant investment horizon. This is often done by simulating many possible outcomes through statistical analysis. A bold alternative to such a "black box" approach is to explicitly create the most plausible VAR scenario for the 1998-2002 investment period. Actually building such a bad-outcome scenario means taking ownership of it, and therefore taking it much more seriously.
Assessing the risk-reward implications of candidate asset mix policies such as a 60-40 equity-to-debt asset mix in defined benefit pension plans requires a fiduciary risk context.
Which asset mix now best balances long term gains versus potential 1998-2002 pain? Is it still 60-40? Or is now the time to take a "road less traveled"? This is the essential fiduciary risk question facing pension fund fiduciaries in 1998. Sponsors should avoid letting maverick risk influence that question.