David Bennett
Worrying ripples from the US pensions industry are beginning to stoke an important debate in the UK.
The problem first really surfaced in May last year, when the SEC accused municipal pension funds of ‘playing numbers games’ to understate unfunded liabilities. The SEC argued sponsors had been forecasting asset performance based on ‘over-optimistic’ return assumptions, leading funds to take on too much risk in a search for high-yielding investments. Since then, a series of billion dollar shortfall announcements have followed.
But let’s be clear: this is not an isolated problem; UK pension funds need to heed the lessons of the US if they are to avoid the inherent problems caused by overoptimistic assumptions.
At Redington, one of our core investment principles is the premise that using prudent expected return assumptions incorporates a margin of safety leading to better outcomes. It is crucially important to understand that for certain asset classes, expected returns are difficult to calculate. There are no right or wrong answers.
Bond-like assets, which have what we describe as contractual cash flows, are relatively easy – especially when held to maturity. The dominant variables are the default rate – the probability of default in a given period – and the recovery rate – the amount of your investment you get back, often after a long lag, in the event of default.
At the other end of the spectrum, equity returns are notoriously difficult to accurately predict. There is rich data available for certain markets, which can be used to observe historical ranges for the equity risk premium (the excess return of equities above the risk-free rate).
Because of the very high historical correlation between global equity markets (ex Japan), it is tempting to infer conclusions about global developed market equity returns from the rich, historical US data set. However, many allowances need to be made when trying to extrapolate from US data-based analysis.
Our research suggests market timing – which is incredibly hard – dominates the 10-year holding period return, with a small number of entry and exit points having a surprisingly big impact on excess returns. Even the most credible and widely followed valuation models provide little certainty.
So why do we prefer to use prudent expected returns? An obvious point is that the choice of expected return does not impact the actual return the asset class delivers. But what it can do is drive asset allocation. In the event returns exceed expectations, strategy can be adjusted to lock in the unexpectedly strong position. On the other hand, over-optimistic expected returns can leave schemes significantly behind their path to full funding and in a position where the original funding objectives are exceptionally challenging. In conjunction with a risk budget, prudent assumptions can help deliver more optimal asset allocations.
In this context, how can prudent assumptions lead to better outcomes? Well, a simple example can illustrate this concept (with the caveat that that there is a lot of devil in the detail): Suppose a portfolio is structured with a strategic asset allocation to deliver a given target expected return. If I increase the expected return assumption on, for example, equities from 2% over gilts to 4% over gilts, I can halve the amount of equities in the portfolio. If the actual return is 4%, the portfolio with the lower expected return assumption will outperform as it has twice as much equities.
Hopefully, the UK pension industry can learn from the experiences across the Atlantic. As Winston Churchill reportedly said: “Those who fail to learn from history are doomed to repeat it.”
David Bennett is head of investment consulting at Redington
Comments