18 February, 2020

As pretty much everyone now knows, the Retail Prices Index (RPI) is in poor health as a statistical measure of inflation. Thus, over the past decade or so, its use has begun to be phased out in favour of the more robust CPI measure, which treats housing costs differently, and uses a better averaging formula. Business rates, for example, have been uprated by CPI rather than RPI since 2018 – which is good news for landlords because CPI typically runs at around 50 to 100 basis points lower than RPI. However, RPI’s influence on prices in everyday life is still everywhere, including real estate rents. Recent analysis by my colleague Dominic Smith shows, for example, that of the sale-and-leaseback assets whose valuations contribute to CBRE’s Long Income Index, 76% are RPI-linked, while only 13% are CPI linked – and RPI’s share is even higher for other forms of long income (which are becoming more popular).

RPI’s problems have been known to statisticians for some time, and indeed the UK Statistics Authority has been arguing that it should be discontinued, or at least modified to address some of its deficiencies so that it behaves more like CPI. But changing RPI is not merely a matter of statistical neatness. If you are a pensioner who has bought an annuity whose income is inflation-linked to RPI, you do not suddenly want your income to inflate by less than you paid for it to inflate, even if the things you will buy with that income are (in the real world) inflating at the lower rate. And the same goes for landlords buying index-linked real estate rents.

However, an RPI-to-CPI switch is more or less what is proposed. From 2030, the Statistics Authority will have the right to make major changes to RPI to make it behave more like CPI, or indeed to stop publishing it altogether. Before 2030, the Chancellor is entitled to veto major changes, because there are some RPI-linked gilts in the bond market which would be affected (by 2030 the relevant gilts will all have matured).

Ten years might be regarded as enough time for the market to adjust, and there is some evidence (the effects of Brexit on inflation confuse the picture) that the market is already pricing long-dated RPI-based instruments to account for the fact that RPI indexing will probably be worth less after 2030. We’ve also heard that contract writers are hedging their bets by including clauses which allow the contract to jump horses to CPI (plus a margin) if RPI stops being published or is materially changed.

However, the Chancellor recently surprised the market by appearing to suggest that he might be willing to make these changes in 2025 – five years earlier than the market was expecting. He has proposed a consultation on whether the date should be brought forward, and also on how RPI might be aligned to CPI. The consultation document will be published alongside the forthcoming 2020 Budget.