UK Gilt yields and discount rates – Does the historical link still apply?

Last summer, I discussed the BHS Pension Scheme deficit.  

At the time, a figure of £571 million (on a buy-out basis) or £250 – £500 million (First Report of the Work and Pensions Committee and Fourth Report of the Business, Innovation and Skills Committee of Session 2016 – 17 dated 20 July 2016) was being talked about and we now see that a settlement with Sir Philip Green has been reached at £363 million. 

To my mind, the discount rate applied is the key assumption that affects the calculation and the relevant regulations[1] contain the following:

“… the rates of interest used to discount future payments of benefits must be chosen prudently, taking into account either or both –

(i) the yield on assets held by the scheme to fund future benefits and the anticipated future investment returns, and

(ii) the market redemption yields on government or other high-quality bonds…

Notwithstanding  the fact that schemes can (and do) invest in equities, the discount rate appears to have become correlated with a return on Gilts plus a premium to reflect the additional return expected from the riskier class of asset in which the scheme actually invests.  This is referred to as the “Gilts plus” approach by practitioners.

The Pensions Regulator has published Scheme funding statistics[2] that show that from September 2005 to September 2014 the discount rate is around 1% per annum above the nominal Gilt yield[3].  The statistics also show[4] that the real discount rate has become significantly more prudent over the same period. 

An alternative approach to deriving the discount rate is the “inflation plus” or “CPI plus” approach. This is based on the Fisher equation, a hypothesis that the total return on an asset is made up of an element that represents inflation together with an element reflecting the real return.  This is analogous to the “Gilts plus” approach with inflation replacing the return on Gilts and a premium being added.

Until 2011 “Gilts plus” and “inflation plus” were strongly correlated, however from the end of 2010 they began to diverge.  This divergence results principally from the current negative real Gilt yield as opposed to the long term positive yield observed since 1900.