UK Economic Outlook: RIP to RPI
The precise formula used to calculate inflation and what inflation measure to use may seem incredibly esoteric but it has far reaching and significant real world impacts not least for pensioners and holders of index linked gilts. There are two main measures of inflation in the UK the Retail Price Index (RPI) and the Consumer Price index (CPI) whilst the different indices include different things (RPI for example includes mortgage interest payments) the main and most consistent difference between the two is due to differences in the formulas used to calculate the final number. In one infamous example the price of certain items of women’s clothing increased by 400% between 2010 and 2017 according to the RPI and by 20% according to CPIall based on the same underlying data.
It should be noted that there is no ‘correct answer’ to what the true representation of inflation is. Both sets of statistics are constructs designed to replicate imperfectly something that is inherently unobservable. However a judgement can be taken and the judgement of the Consumer Prices Advisory Committee (CPAC) is that “the formula used in RPI should not be used in any index aiming to achieve a good estimate of changes in consumer prices” and further that it “is not suitable for use”. As such the Office for National Statistics removed it is as an official statistic in 2013. And yet it continues to be used.
Which brings us, in a somewhat circuitous manner and via some arcane statistical miscellanea to what happened to index-linked gilt yields at the start of September. Sajid Javid announced that RPI would be reformed, over the next 10 years, in line with the CPIH index that uses the CPI formulas but like RPI and unlike the main CPI index takes account of housing costs and so is very similar in terms of content to RPI. Given that this would reduce the measured level of RPI going forward it would also reduce cashflows to holders of index linked gilts as such prices fell and 30-year yields rose by 0.2%. The differences between CPIH and RPI can be seen in Figure 1.
The reform of RPI has real and significant impact for pension investors but the impact will depend on the individual schemes’ liabilities (the degree to which they are exposed to RPI rather than CPI), investment strategy and funding position. At one extreme a fully funded scheme that is mostly exposed to RPI and has hedged all its inflation risk will not be impacted, the gains from having to pay a lower level of inflation to its members will be offset by losses on its index linked gilt portfolio. At the other extreme a poorly funded scheme that has hedged a minimal amount of its liabilities will be significantly better off as the gains from lower payouts to members will not be offset by losses on the hedge portfolio. Most schemes with RPI based liabilities should end up better off this is because most schemes are less than 100% funded even if they are fully hedged on a funded basis. For example a scheme that is 90% funded and has hedged 100% of its funded liabilities still has 10% of its liabilities that are unhedged as such it would register a gain on 100% of its liabilities but a loss on only 90% of its assets.
Of course, many schemes already formally switched their indexation measure to CPI following the CPAC decision in 2013; this latest news affects those who were unable to do so as their scheme rules prevented them doing so (their rules may have specifically referenced RPI rather than a more loosely defined inflation measure).
One less interesting tidbit to people who don’t spend their lives in the minutiae of LDI: if RPI gets closer to CPI, then we can eventually dispense with one more simplifying assumption when building an LDI benchmark (less ‘basis risk’ as we remove the RPI-CPI wedge assumption) – which should mean benchmarks become more reliable.
This story is going to develop slowly but has significant implications. Kempen will keep you informed of any new developments.