UK Economic Outlook: The £85bn statistical adjustment
For decades, the Government’s favoured measure of inflation was the Retail Price Index (RPI) which has been in use since 1947. However, it turns out that RPI is a highly flawed measure that produces some very odd results. The Bank of England stopped using RPI in 2003. Then, in 2013, the UK Statistics Authority itself determined that RPI was not fit for purpose and even declassified it as a national statistic.
Nevertheless, RPI continued to be used in most contracts relating to inflation and so continued to be published. Most notably, for our purposes, it is still used in the calculation of a lot of inflation-linked pension liabilities, as well as for index-linked gilts and inflation swaps, which pension schemes use to hedge those liabilities. As far as pensions go, RPI is very much essential, even if it is flawed.
The most recent chapter of the saga was initiated by a report published in 2019 by the House of Lords Economic Affairs Committee. The report concluded, in simple terms, that the Government should pick a measure of inflation that wasn’t a breach of statutory duties on statistics, and jolly well stick with it. And so, as governments often tend to do when faced with a tricky situation, a consultation was launched.
The conclusion of the consultation finally came last week and the result was pretty much the only option left open. Being unable to ditch RPI, but also unable to use it in its present form, the decision was made to keep it in name, but totally change the calculation. It would now become an entirely different inflation measure which produces much more sensible results. Specifically, from 2030, RPI will be calculated in exactly the same way as an existing index called the “Consumer Prices Index including owner occupiers’ housing costs”, or CPIH to its friends. CPIH is based firmly on the headline CPI index that the Bank of England now uses for inflation targeting, but adjusted to capture the costs of owning a house.
As stated above, whilst this may seem arcane, it has significant real world effects. RPI is structurally around 1.0% per year higher than CPIH. So most pensioners will see a gradual but ultimately very significant change in their incomes. And likewise, holders of index-linked gilts will now get 1.0% per year lower cashflows under the new definition of RPI. The PPF estimated a loss in value of £85 billion for the holders of index-linked gilts — which means mainly pension schemes and insurers. And some other estimates are even higher.
As you might have guessed, this has had a massive impact on the index-linked gilt market. Historically, long-dated market-based measures of inflation have traded at 0.5% over short-dated measures. The chart shows how this gap has evolved over the last five years: it has gradually moved down, as the market took on board that 1.0% per year or so difference between RPI and CPIH. It moved first in early 2016 when a reform was initially proposed, and then again at the start of 2019 when the House of Lords report was published.
This has meant that the long-dated index-linked gilts held by pension schemes are now cheaper with the RPI reform than they otherwise would have been. To an extent, this has been disguised by the fact that the price of index-linked gilts has been rising over this period due to other factors, so any impact has been hard to spot. However, had the reform not taken place, prices would have risen by even more and long-dated index-linked gilt yields could have hit minus 3.0% per year, instead of the minus 2.0% per year they currently trade at!
The impacts on pension schemes and will vary a lot from case by case. As an illustration three examples have been provided.
- A scheme that had hedged 100% of its RPI inflation-linked liabilities would see no change to its funding position. The reduction in cashflows in from lower index-linked gilts and RPI swaps would match the reduction in cashflows out from lower payments to pensioners.
- A scheme with RPI-linked liabilities that were less than fully hedged, should see a net benefit. The scheme will pay a lower amount on all its RPI-linked liabilities, but the reduction in cashflows on its RPI-linked assets will be more limited.
- A scheme that had mostly CPI liabilities and hedged them with index-linked gilts or RPI swaps would be worse off. The cashflows of the liabilities will not change, whilst the cash flow of the assets held will be lower going forward. The PPF itself is an extreme example of this.
What has been particularly interesting though is the market reaction after the publication of the consultation outcome. Many participants thought the change was only partly priced into the market, around 80% was often mentioned. They expected to see index-linked gilt yields rise quite significantly on the announcement, as the uncertainty was lifted. But in the end the opposite happened. Index-linked yields have fallen relative to nominal yields and thus long-term market-implied RPI inflation levels have actually risen, which is odd as the reform of itself reduces future RPI!
It is very hard to know exactly why the market reacted like that, but there are some interesting theories.
Firstly that the impact was more priced than expected so there was little potential upward move. Indeed, some have argued that the market had been pricing in some probability of an even stronger reform, with the new RPI calculation applying from 2025, rather than 2030.
Secondly, that the removal of the uncertainty that had built up around the consultation was removed allowing investors to purchase index-linked gilts with more confidence about the new regime. Put another way, the announcement released pent-up demand for hedging.
Finally, and perhaps most interestingly, that some investors seem to have been tactically positioned for index-linked gilt yields to rise. Then, when this failed to occur, they needed to buy back the securities they had previously sold, contributing to price rises.
So, what next? Well, 2030 is still ten years away, and politics is a fickle game: some commentators have suggested that a future government might reverse the reform to placate angry pensioners. It’s always possible, but we think it unlikely. The impact of this is a very slow burn, with 2030 being only the start of a gradual erosion. If a future government wanted to give away £85 billion, there are other ways to spend it that would be much more effective in winning votes.
More immediately, pension schemes will need to work out how this impacts their funding position, positively or negatively. Whilst asset prices already reflected the reform, even before the final announcement, the impact on the liabilities will require actuarial calculations. For some sponsors this could generate a step up in contributions. Ironically, some of those hardest hit will be the most prudent schemes that fully hedged their liabilities.
The market reaction also raises investment strategy questions. Now that the future of RPI is clear, inflation hedging already looks really rather expensive relative to, say, the Bank of England’s long term target for CPI of 2.0% per year. And there is unrelenting demand for more inflation exposure from pension schemes as they de-risk, mature, consolidate, and buy out. Unless the Treasury takes steps to have the DMO significantly step up the amount of inflation-linked issuance, this situation will only get worse. And that will leave trustees between the devil of over-priced inflation-linked assets, and the deep blue sea of a large unhedged risk exposure. Not a nice place to be.
Download our white paper
The £85bn statistical adjustmentDownload
This document of Kempen Capital Management UK Ltd (KCM) is for information purposes only. The information in this document is incomplete without the verbal explanation given by an employee of KCM. KCM is licensed as an investment manager and authorised to provide investment services, and, as such, is subject to supervision by the Financial Conduct Authority. No part of this presentation may be used without prior permission from KCM.