A Little Learning Is A Dangerous Thing

“A little learning is a dangerous thing
Drink deep, or taste not the Pierian spring:
There shallow draughts intoxicate the brain,
And drinking largely sobers us again”

This is an extract from Alexander Pope’s famous book-length poem written some 300 years ago, “An Essay on Criticism”. The first line above is often rephrased as “A little bit of knowledge is a dangerous thing”. This has much truth in investments today.

This publication of the Canvas provides a real life example of a pension fund that believed it used LDI (Liability Driven Investment) techniques to de-risk itself. Unfortunately, a strategy that was intuitively appealing did not de-risk it at all!

On the 31st May 2010, the actuary of the “Knowledgeable Pension Fund” informs its trustees that there is exactly the required quantum of assets to meet the on-going pensioner obligations (allowing for future increases in line with 100% of inflation). He also states that he is valuing the pensioner liabilities using a discount rate of 3% (the prevailing real yield on the R197 government-issued inflation-linked bond).

As the Fund is targeting pension increases in line with 100% of inflation, the Fund deduces that it requires a return of “Inflation+3%” per annum to stay solvent and provide for a full inflation-linked pension increase. The Trustees conclude that an “absolute return” strategy is the perfect investment as the investment manager’s benchmark is a return related to inflation. Moreover, it is possible to select from six manager offerings targeting the required “Inflation + 3%” return.

After a full manager due diligence, they decide to allocate their assets to two absolute return managers targeting “Inflation+3%”.

The Trustees review the solvency of the Fund one year later. For the year ending 31 May 2011, inflation was 4.22%. The Fund thus required their managers to achieve a return of 7.22% (inflation + 3%). Both appointed managers outperformed the benchmark and delivered returns of 9.25% and 8.06% respectively (a combined return of 8.66%, or otherwise put, some 1.4% more than what they had targeted). However, the actuary reveals that the Fund has fallen into deficit and that pension increases this year need to be reduced.

How did this happen?

The pensioner liabilities have a value that is determined with reference to both prevailing inflation and changes in real yields. It is easy to appreciate that the value of the liabilities should increase by the realised inflation over the period concerned (after all, pensioners would like to retain the purchasing power of their pensions and hence funds award pension increases related to inflation). However, liabilities are also “valued” with reference to what the “market” believes the price of such liabilities should be. The actuary therefore determines their value by considering the price of financial instruments with similar characteristics.

Pensioner liabilities and inflation-linked bonds have very similar characteristics: they both make predetermined payments at future points in time and these payments increase in line with inflation. In our example, the actuary uses the prevailing yield on the R197 government-issued inflation-linked bond to determine the liability value as he believes it is a good proxy for the characteristics of the Fund’s liabilities. The yield on this bond can thus be used to determine a market value for the Fund’s liabilities.

It stands to reason that if the price of this inflation-linked bond increases, then so too would the value of the Fund’s obligations and vice versa.

As at 31st May 2011, the prevailing yield on the R197 inflation-linked bond had fallen to 2.52% which meant that the value of the pensioner liability increased by 11.48% over the year! (more than 4% higher than the 7.22% the Fund believed it had needed).

The graph below contrasts the rolling 12-month “Inflation+3%” target return against that which the Fund actually required to remain solvent and provide increases in line with inflation. It can thus be seen that CPI+3% is well below the Actual Return Required and is clearly not an appropriate benchmark to be targeting. This is not to say that going forward it will continue to underperform. The opposite could hold true. The key point is that the benchmark or targeted return of Inflation+3% is not moving in line with changes in the Fund’s liabilities.

How could this have been prevented?

The Trustees used the wrong benchmark in this instance – they believed they needed to target a return of “Inflation+3%” whereas in fact they needed to outperform the return of the “R197 Inflation-linked bond” instead. These are two dramatically different benchmarks which result in very different investment strategies. Complicating it further, the liabilities cannot practically be replicated by a single inflation-linked bond like the R197. This is because liabilities are generally longer dated that the R197, and also have a different payoff profile. Whilst the proceeds are inflation linked, the profile of proceeds are different to the profile of pension payments. Pension funds thus need a liability driven investment manager in order to determine the optimal portfolio of inflation-linked bonds required.

Conclusion

Trustees evaluate their investment managers by considering whether they outperform their benchmarks. Many of the benchmarks given to these managers have no relation to the liabilities and thus outperformance should provide very little comfort to the fund. In our example, we gave a “real life” scenario of a fund that believed it had the appropriate benchmark and its managers outperformed this benchmark. Disappointingly, and to their surprise, they still ended up in a deficit position. Deriving an investment strategy that actually de-risks pension funds is a complex process. The possible risks of a chosen investment strategy should be appreciated before implementation. This doesn’t mean that trustees should be experts in investments. Rather, they should have a basic understanding of the core issues and select investment managers with a demonstrable track record in achieving such benchmarks. After all, as Bob Edwards (an American talk show host) recently remarked: “A little learning is a dangerous thing but a lot of ignorance is just as bad”.