RPI: rumours of death exaggerated

by George Hatjoullis

The UK RPI was disowned as a national price index some time ago. Yet the Office of National Statistics still calculate it and report it. Rumours of its death seem somewhat exaggerated. Why should we care? First, the RPI is still the reference on a number of important contracts. Second, it is consistently and systematically higher than the official price index (with brief exceptions), the UK CPI. The year on year change in the CPI reported August 16, 2016 is 0.6%. The RPI rose 1.9% over the same period. In a world of zero to negative interest rates this is not a trivial difference. If you have a contract that increases a payment or a value by the RPI, you have something valuable (or painful).

National Savings Index Linked Savings Certificates

You cannot invest in these anymore but if you already have them you can roll them over when they mature. Moreover, you can extend the maturity where relevant. So if you have a 3 year maturing you can extend to 5 years. These certificates pay RPI +.05%, so this month it is 1.95%. How many default free investments offer you such a return at the moment? If you have some you know about it. However, resist the temptation to cash them in and consider extending maturity when possible. They are presently offering you 1.35% above the CPI. Strictly speaking this is a default free real return of 1.35%. Not too shabby in this low-interest rate environment.

Index Linked Gilts

Index Linked Gilts are government debt linked to the RPI. The return on, and value of, ILGs is more complicated than just the RPI but it is an important component. The index is of course reflected in the price you will pay for these instruments so there is no necessary advantage to investing, unlike NSI certificates, unless the market is surprised by the difference between the CPI and RPI. However, it is worth noting that the index accrual is linked to the RPI and not the CPI (I am not aware of any CPI linked issues) and this means the principal and coupon uplift will be higher than the CPI would require. This imposes a much higher debt cost on the state which is ironic because the RPI is not a national statistic.


Pensions in payment and deferred pensions often have an index linked component. The RPI was once the index of choice and the legacy is still widespread. The maximum amount of uplift is often capped at, say, 5%, but given inflation levels this is not much of a constraint at the moment. This means you are strictly speaking being given real pension increases. Nice.

State Pensions

Now be prepared to be amazed. If you are a man born before 6 April 1951 or a woman born before 6 April 1953 you have a very nice deal. The pension uplift uses the CPI and not the RPI but it also uses an index of wages and 2.5%. It will go up by the higher of the 3 criteria. It will go up by at least 2.5%, which is incredibly generous given current inflation and interest rate levels. Moreover, average earnings typically rise faster than all inflation indices (though not invariably) so not much risk under any scenario. Very generous.

If you are born after these dates things look less clear. The assumption being made is that the same generous indexation process will apply though I have found no direct confirmation of this assumption (which is why it is an assumption). Given the amazing generosity of the existing index arrangements I would keep a wary eye on announcements relating to new pension indexation process. Indexation has material implications for government social expenditure so it is a sensitive area.

The continuing use of the RPI in so many contracts is unlikely to last forever given it is no longer deemed a national statistic. However, while it does it can offer some advantages.