Calling time

Let’s face it, the retail prices index has had its day.

First calculated all the way back in 1947, the venerable retail prices index (or RPI) was, for years, the main way we measured inflation in the UK. It is used to measure the way that prices increase from month to month and year to year. It is embedded in all sorts of contracts and customs, from the interest the Government pays on index-linked gilts, to the prices you pay for train tickets. It is written into the legal terms of many UK pension schemes.

However, the RPI stopped being a national statistic in 2013 and it is now receiving only limited ongoing maintenance. This year, Mark Carney himself outlined flaws in the RPI and said he thought those with the authority to do so should pick a date – perhaps seven or eight years in the future – after which RPI would not be published at all.

Does the RPI still work, and what should pension schemes do about it? 

No price index is perfect, but RPI has a number of particular quirks.

Prices can fall as well as rise

RPI can give some rather curious messages when prices go down. 

If the cost of my basket of goods goes from ￡100 this month to ￡110 next month then back to ￡100 again, it doesn’t take an economist to tell me that I’m paying the same now as I did two months ago. The RPI, though, would suggest that I’m paying less.

Housing

RPI includes mortgage interest. There is good reason for that – after all, for lots of people in the UK the mortgage is one of the biggest things they spend money on, if not the biggest; however, it has some odd consequences. 

Last year the Bank of England increased interest rates to slow down the rise in prices. The increase in interest rates increased mortgage payments, which increased “prices” as measured by the RPI – exactly the opposite of what was intended[1].  Back in 2009, RPI was consistently negative because cuts in interest rates slashed mortgage payments, even though prices more broadly continued to rise.

Of course, very few pensioners have mortgages[2]. So why do pension schemes (for instance) link their members’ incomes to the vagaries of the mortgage market?

From bias free of every kind?

Because of the way it is put together, RPI can systematically overestimate increases in prices compared to more modern and robust measures. This happens in areas, such as clothing, where the RPI is looking at lots of different items, prices of which can move in very different ways.

Because of this, RPI has been consistently telling us that prices have increased 0.5% to 1% faster each year than the Consumer Price Index (CPI) – the inflation measure used around the globe. Ironically, that gap widened back in 2010 when ONS improved its data gathering on clothing prices. As data collection for CPI is improved, it is possible that the upward bias of RPI could increase further.

Yesterday’s index, telling yesterday’s news

Perhaps the biggest problem, though, is the fact that the RPI is no longer being maintained. 

Some of the quirks of the RPI could have been mitigated through improvements to the way it is put together. As it is, though, that will not happen. As time moves on, the RPI will give an ever less realistic picture of how prices are really changing.

A legal hodgepodge

So what does this mean for pension schemes? 

Pension schemes in the UK are required to provide annual pension increases, linked to inflation at a rate selected by the Secretary of State. Historically, increases were based on the RPI, but in 2011 the CPI became the index of choice.

However, scheme rules have a wide range of different approaches to defining the increases they would give. Some referred only to legislative requirements, so switched to CPI automatically.

Many more, however, wrote “RPI” into their benefits in some way, shape or form. There is a hodgepodge of different ways that schemes chose to do that – which has real consequences for schemes and their members.

Some, such as Arcadia and QinetiQ, have found that their benefits don’t really need to follow RPI after all: it turned out that their legal documents were written in such a way that their trustees and sponsors could choose to use CPI instead of RPI.

Employers of others, such as Thales, Barnardo’s and (it seems) BT thought their legal documents might let them choose to use CPI, but the Courts decided they had to stick with RPI.

Each of these cases depended on the precise wording of the legal documents, wording that was chosen many decades ago, long before anyone ever imagined we might be discussing whether RPI continues to be up to the task. The impact for schemes can be huge – potentially as much as 10% of the overall cost of benefits.

Just because you can, doesn’t mean you should

Of course, just because a scheme can change from RPI to CPI pension increases doesn’t mean it is right to do so. CPI may be considered a better measure, but it will give members lower pensions and can be viewed as understating how much people’s cost of living increases. 

For some schemes – particularly those that are weakly funded or have weak employers – it may be right for trustees to make that switch, to give them a better chance of providing pensions to all their members. 

For better funded schemes it may be a more difficult call; however, trustees of these schemes should ask themselves whether RPI – as it drifts further from reality – remains appropriate, and whether it really represents what they, and members, expect. Trustees need to consider both the cost of maintaining RPI linkage, and the benefit in switching – for instance, a reduction in the cost of benefits could be used to support a reduction in investment risk, which might be to the advantage of all involved.

There are very few financial instruments that provide CPI-based income. Schemes using RPI can use index-linked gilts to manage their inflation risks. Schemes using CPI can buy index-linked gilts, but will always know that the income they receive on those gilts may not tie up with the benefits they have to pay out.

Boilers and other not-so-broken things

I recently had my boiler replaced – for reasons I can’t quite explain, in January. The boiler worked – sort of; however, it was old; it was inefficient; it was noisy; it was costly to run; it was no longer fully supported; spare parts were no longer stocked, if they were available at all. 

I chose to replace the boiler, not because it was broken today, but because it was increasingly likely to break and to cause major problems when it did. In some senses, the RPI might be the same. For all its flaws, it still, in the eyes of some, works – sort of.

However, as time goes on, with the limited maintenance it is now receiving, it is increasingly likely that the RPI will break irreparably. Even before that happens, more and more problems and unnecessary costs will be created as the RPI becomes ever more detached from reality.

To protect themselves, and their members, pension schemes need to be able to choose the inflation measure they think is right. Can it be right that employers and trustees are prevented from doing what is right for their members as a whole, based on decisions taken all those years ago, when circumstances were different?

To be able to make that choice, pension schemes need help from Government. Only Government has the power to fix the situation, to abandon RPI for ever, or to cut through the mess of legal documentation so that trustees and employers can choose what is right for their scheme and their members. Government needs to act now – waiting for the boiler to break is likely to be far more difficult.

[1] This is a large part of why the Bank of England, before it adopted CPI inflation, used the RPI excluding mortgage interest (RPIX) in its interest rate decisions.

[2] English Housing Survey 2016-17 indicates that out of 6.6m households with “reference person” aged over 65, 4.8m own their home outright, 1.1m are social renters, 0.4m are private renters and only 0.3m are buying with a mortgage.