THE SHIFT IN LGPS ACTUARIAL DISCOUNT RATES
Over the past ten years, the actuarial firms advising the LGPS have largely moved their discount rate methodology (i.e. how they assess the present value of future liabilities) away from the gilts yield curve and towards either a market-based or an inflation-based number. There is nothing new with this. Over time methods have changed to reflect commercial pressures, developments in financial mathematics and computing, and practice in areas such as life insurance. However, one important principle has always been to keep a strong link between the valuations of assets and those of liabilities.
In the decades immediately prior to 2016 most, but not all, of the LGPS actuaries used the gilts curve to value liabilities, as the private sector still does, albeit with some substantial fudging to make it acceptable to stakeholders. That was partly a hark back to the days when bonds constituted the majority of the assets, and partly the 1995 introduction of the Government’s Minimum Funding Requirement.
The latter was intended as a robust and conservative methodology and is still used in the private sector. But within the LGPS it was fudged from the beginning, and actuaries have now moved a long way away from it, though it remains as a footnote to the triennial reports, and it is used when employers leave the Scheme.
Actuarial advisers have made the most recent shift because bond yields at below 1% have almost certainly been distorted by government monetary and regulatory policies over the past 25 years. Not only have interest rates been close to zero for the past decade, but tax and regulations have forced insurance companies and private sector pension funds to invest in bonds. Within the LGPS commercial pressures have virtually forced actuaries to move away from the gilts curve.
Looking at the asset side of the equation, readers may think that it is obvious that market values should be used. However, within the actuarial profession there has long been a view that market prices are too volatile to be used. In 1921 one comment was: ‘not inclined to give much credit to present market values because they were more or less fictitious, being the values which a man who was compelled to realise was prepared to take, and …it was extremely unwise to write everything down to that level.’ Another actuary asked what had happened on the 19th October 1987[*] to make everything worth 25% less.
The real issue with using potentially volatile market values for assets was the difficulty of making them consistent with the methodology used for liabilities. It was easier when pension funds were substantially invested in government bonds, but since the 1920s the equity and real estate weightings gradually grew. The solution in early days was to use the lower of market or book values. Actuaries then moved to a discounted value of the expected cashflows deriving from investments. In a simpler world when equities were listed and generally paid dividends, that reduced market volatility. In today’s world when an increasing percentage of assets are held privately and future cashflows are less certain it requires more subjective assumptions.
There is a clear divide between the private sector pension funds which largely still use gilts yields as the basis for their discount rates, and the LGPS, which broadly doesn’t. That is defensible, given the very different portfolios they are invested in and the need for consistency between liability and asset valuation. However, it does mean that an actuarial valuation will not necessarily perform one of its main functions: providing a reasonable assessment to stakeholders of how likely the pension fund is to be in a position to pay pensions.
In the private sector, where the discount rate is artificially low, liabilities are probably overstated. Consultants and fiduciary managers have largely invested assets in low yielding gilts and corporate bonds. There is a consequent onus on employers to inject greater contributions than they may in the long term actually need to. That can lead to other undesirable consequences, such as companies going into pre-pack administration to get rid of their pension liabilities.
In the LGPS, liabilities may be understated depending on the methodology used. If inflation goes up to 5%+ or markets fall with the consequence of a higher dividend yield, discount rates would rise further and the understating might increase. In practice, I am sure there would be a ‘fudge’ under the fig leaf of prudence, much as there was ‘smoothing’ when gilt yields fell from 5% to 1%, to create a discount rate which reflected the actuary’s subjective view of what was appropriate.
This takes us full circle to the role of actuarial advisers in the 19th and early 20th centuries, when their role was precisely this: to provide an expert view on what constituted an appropriate valuation of the pension fund’s liabilities and assets.
The question today is whether that will suit the Government’s agenda. It will be interesting to read the Section 13 valuation of the LGPS done on a standardised basis by the Government Actuary. It must be due out any moment now. How will they choose to value liabilities?
[*] Date of Black Monday and a 25% fall in market values over two days. I could make the response that the utter failure of the BP rights issue made the market aware that market values were too high - effectively the emperor’s nakedness was exposed.