The Risky Business of Pensions

William Sheffler, SDCERS Board Member – May 2007

For most members the health of their retirement plan is measured by its investment returns. For many years the SDCERS board touted its investment success as evidence of the plan’s soundness. This practice continued while MPI and MPII were approved. In fact, as we now know, there is a lot to managing a plan besides investment performance.

There are several types of risks that a retirement plan, like the city pension, must bear. They include:

1) Investment Risk

2) Actuarial Assumption Risk

3) Plan sponsor risk

4) Asset / Liability matching risk

Investment risk: Up to 80% of this plan’s benefits are paid out of investment earnings. Therefore, it deservedly is the focus of the board’s attention. Many public plans, including SDCERS, are limited by law as to what types of investments they can make. In our case, equities (stocks) can not exceed 70% of the plan assets. For many years, this plan was required to invest principally in bonds or savings type instruments. This limitation reduced the rate of return, and forced the city (and its employees) to contribute more to the fund, since its earnings were lower. The justification for these restrictions was that public money should not be subject to losses, like those we see in the stock market. There also other historical justifications for that policy.

Actuarial Risk: The actuary makes several assumptions when estimating the annual cost of the plan, and its financial condition. Investment return is just one of them. Others include future inflation, mortality, disability, employee turnover, and expected age at retirement and future salary increases. To the extent that these assumptions are incorrect the plan cost may have greater or fewer liabilities. Those deviations also get reflected in annual costs. Every three years the actuary does a comprehensive study to determine the accuracy of each assumption. Changes to the assumptions are made as a result of his findings.

To illustrate, let’s look at the salary increase assumption. The actuary’s salary increase assumption is roughly X% per year. If salaries are unchanged during a valuation year, the plan will have a gain. That’s actuary talk which means liabilities are less than expected. Also consider a 3 year labor contract that keeps salaries level for two years and then in the third year, requires that salaries rise 7%. The plan would have substantial gains for each of the first two years, and a loss in the third year. These types of variations reinforce the need to follow plan trends for several years, and not individual annual results.

Plan sponsor risk: This means the possibility that for some reason the employer can not properly fund the plan. That can happen when the employer’s income drops, but their payroll costs do not. It can also happen when an employer delays funding their plan so long that most benefits must be paid from contributions, and there is not sufficient time to accumulate investment earning to make up the difference between what the plan sponsor contributes and what it must pay out. The value of the fund drops, because more money is going out than coming in, and even less investment income is added to the fund. After a few years, the plan enters a “death spiral’ as more and more money must be put in to the plan, and less and less investment income is added. Eventually, the plan becomes unaffordable and either the sponsor or the plan goes bankrupt. This was the danger of MPI and MPII. It is also a potent argument against skipping annual contributions, or ‘temporarily’ reducing them.

Asset/Liability Matching Risk: This problem is much more subtle, but potentially quite dangerous. This means that the plan investments and its benefits costs should change in an identical fashion in response to economic trends. Our plan benefits are very sensitive to inflation. Generally speaking, if inflation is high, salaries will increase faster than usual. Plan liabilities will also rapidly rise. To keep the plan properly funded, or to avoid having unfunded liabilities multiply, plan assets need to keep up with the inflationary trend. The investments which best achieve that result are common stocks. Many private sector union plans provide benefits that are a fixed monthly amount for each year of service. Their liabilities would not be sensitive to inflation, therefore a greater allocation to bonds, or fixed income would be called for. In some extreme cases, well funded plans exactly match their investment income to their benefit payments. Thus they “immunize” themselves from risk due to inflation, or salary increases.

As an actuary, I feel that the SDCERS board needs to address the asset / liability matching risk. It has not been done in the history of the plan. It is important for us now because of the prior practice of focusing investment concerns on gaining high rates of return, with minimum chance of loss. That itself can cause us to be complacent. This risk could ‘blindside’ the plan finances. That would happen when we had plan liabilities increasing faster than assets, even when the investment return exceeded the actuarial assumption. It’s the result of our benefit obligations not modeling the investment allocation.

In addition to directly studying this dynamic, the Board now has new tools, brought to us by the actuary Cheiron. These tools allow us to simulate changing investment assets and pension liabilities over several years. Those simulations include literally thousands different combinations of conditions. This is generally referred to as ‘stress’ testing. It is intended to reveal those particular combinations of events that could damage the financial status of the plan. It also generates for us the probability or odds of those conditions arising. The chief application of that method is to evaluate and identify catastrophic events. This kind of testing most likely would have identified the weakness of the MP-I and MP-II agreements.

There are of course, other things that can go wrong with a pension plan. This article highlights some of them. It’s the job of your board of administration to identify and minimize as many of those dangers as possible.

[Editor’s note: This article was discussed at the May 2007 SDCERS Board meeting. Mr. Sheffler was asking if the SDCERS' actuary would calculate the capacities of the City to make annual contributions to SDCERS in future years. Mr. Sheffler provided this article to us for the CSDREA newsletter and/or the web site.]