By: Ronald J. Ryan, CFA, Chairman, Ryan ALM, Inc.

In this post, Ron continues with his series on identifying solutions to various pension-related problems. This one addresses the issue of asset allocation being driven exclusively from an asset perspective.

Most, if not all asset allocation models are focused on achieving a total return target or hurdle rate… commonly called the ROA (return on assets). This ROA target return is derived from a weighting of the forecasted index benchmark returns for each asset class except for bonds which uses the yield of the index benchmark. These forecasts are generally based on some historical average (i.e. last 20 years or longer) with slight adjustments based on recent observations. As a result, it is common that most pensions have the same or similar ROA. 

This ROA exercise ignores the funded status. It is certainly obvious that a 60% funded plan should have a much higher ROA than a 90% plan. But the balancing item is contributions. If the 60% funded plan would pay more in contributions than the 90% plan (% wise) then it can have a lower ROA. I guess the question is what comes first. And the answer is the ROA with contributions as a byproduct of that ROA target. The actuarial math is whatever the assets don’t fund… contributions will fund.

If the true objective of a pension is to secure and fully fund benefits and expenses (B+E) in a cost-efficient manner with prudent risk, then you would think that liabilities (B+E) would be the focus of asset allocation. NO, liabilities are usually missing in the asset allocation process. Pensions are supposed to be an asset/liability management (ALM) process not a total return process. Ryan ALM recommends the following asset allocation process:

Calculate the cost to fully fund (defease) the B+E of retired lives for the next 10 years chronologically using a cash flow matching (CFM) process with investment grade bonds. CFM will secure and fully fund the retired lives liabilities for the next 10 years. Then calculate the ROA needed to fully fund the residual B+E with the current level of contributions. This is calculated through an asset exhaustion test (AET) which is a GASB requirement as a test of solvency. The difference is GASB requires it on the current estimated ROA before you do this ALM process. Ryan ALM can create this calculated ROA through our AET model. If the calculated ROA is too high, then either you reduce the allocation to the CFM or increase contributions or a little bit of both. If the calculated ROA is low, then increasing the allocation to CFM is appropriate. Running AET iterations can produce the desired or most comfortable asset allocation answer.  

Cash flow matching (CFM) will provide the liquidity and certainty needed to fully fund B+E in a cost-efficient manner with prudent risk. The Ryan ALM model (Liability Beta Portfolio™ or LBP) will reduce funding costs by about 2% per year or roughly 20% for 1-10 years of liabilities. We will use corporate bonds skewed to A/BBB+ issues. According to S&P, investment grade defaults have averaged 0.18% of the IG universe annual for the past 40-years. Fortunately, Ryan ALM has never experienced a bond default in its 21-year history (knock wood).

Assets are a team of liquidity assets (bonds) and growth assets (stocks, etc.) to beat the liability opponent. They should work together in asset allocation to achieve the true pension objective.

For more info on cash flow matching, please contact Russ Kamp, CEO at  rkamp@ryanalm.com