RPI reform: substituting one pension scheme lottery for another

28th July, 2020

  • On 16 April, the government announced that, in light of Covid-19, the deadline for responses to the consultation on reform of the Retail Prices Index (RPI) Methodology would be extended from 22 April 2020 to 21 August 2020.

    So, we will probably need to wait a bit longer before we find out whether RPI, first introduced in 1956, will be reformed and, if so, when.

    In the meantime, this provides an opportunity to take stock on the consultation and the prospective impact that reform of RPI could have on defined benefit (DB) pension schemes and their members.

    By way of reminder, increases to pensions in payment are still linked, in some shape or form, to RPI in around three-quarters of defined benefit (DB) schemes. Consequently if, as proposed, RPI is replaced by the Consumer Prices Index (CPI) then, given that there is a historic difference, on average, of 1% between RPI and CPI / CPI-H, a 65-year-old pensioner, with a currently RPI linked pension, could suffer a lifetime reduction in pension income of 10-15%.

    The reform of RPI could, therefore, have a material impact on DB pension schemes. Further, it should not be assumed that the impact is simply good news for employers (reduction in liabilities) and bad news for members (reduction in pension benefits). Pension scheme assets and their liabilities need to be considered; many schemes have already hedged inflation risk and will not welcome any change.

    Before discussing the holistic impact, it might be useful to set the scene by considering why reform of RPI is being proposed in the first place.

    A very poor measure of general inflation

    The National Statistician (principal adviser to the UK Statistics Authority) has discouraged the use of the RPI, deeming it “a very poor measure of general inflation, at times greatly overestimating and at other times underestimating changes in prices and how these changes are experienced”.

    Nevertheless, it remains widely used.

    One major use of the RPI by government is in its issuance of index-linked gilts, to adjust their coupon payments and the repayment of the principal. There is substantial demand for these from DB schemes seeking to match RPI-linked liabilities.

    CPI, introduced in 1997, is a much more recent measure of inflation, which does not suffer from the same shortcomings as the RPI. However, the CPI does not include a measure of owner-occupiers’ housing costs (the cost of living in and maintaining one’s own home) and so the CPIH was introduced in 2013 to address this.

    CPIH has been the lead measure of inflation for the Office of National Statistics (ONS) since March 2017.

    The RPI and the CPIH ‘wedge’

    Since 2010, the measured rate of RPI annual inflation has been on average one percentage point per annum above the CPIH. The effect of the different formulae that the RPI uses accounts for around 0.7 percentage points of this difference and can predominantly be attributed to clothing prices, which have accounted for 0.4 percentage points of the formula effect since 2010.

    A much fuller explanation of the ‘wedge’ is provided in the consultation. This also describes the framework for inflation measures, where the UK Statistics Authority (‘the Authority’) is responsible for official statistics but there are also roles for the Bank of England and for the government.

    The recommendations

    Taking into account consultations, reviews and inquiries over the past eight years, two recommendations are made:

    1. that the publication of the RPI should cease; or
    2. taking into account that outright abolition of RPI would require legislation, in the interim, the shortcomings of the RPI should be addressed by bringing into it the methods and data sources of CPIH.

    The consultation also considers whether implementation should be at a date other than 2030 (when, under the legal framework, there would no longer be a government role in the process for changes to RPI) and, if so, at what date between 2025 and 2030 (the government not wanting change before 2025).

    Finally, the consultation covers the transition for implementation of the recommendations.

    How the RPI will change

    The Authority remains minded to address the shortcomings of the RPI by bringing into it the methods and data sources from the CPIH. Once fully implemented, the annual and monthly growth in the CPIH and the RPI will be equal.

    When the above happens, based on recent experience, the RPI annual measured rate of inflation will be lower, on average, by 1 percentage point per annum. However, that can vary over time as illustrated in the ONS time series for inflation and price indices.

     https://www.ons.gov.uk/economy/inflationandpriceindices#timeseries

    The Authority is seeking views on how to move from the current RPI and the new approach using the methods and data sources of the CPIH. A ‘chain link’ is preferred and the consultation includes a worked example of how CPIH methods and data sources will be introduced into the RPI.

    Following the transition, whilst the RPI and CPIH index values will not actually match, the growth in the two indexes will, from the implementation date, be identical.

    What does it all mean for pension scheme sponsors, members and trustees?

    Well, that depends. 

    For members who have benefits linked to CPI or do not receive increases, there will be no change. However, where members have all or part of their benefits either in deferment or payment linked to RPI then they will see a change in how their pension increases in future.  Based on historic evidence, where a pension is increased with RPI, future increases might be around 1% lower depending on the impact of any ‘caps’ and ‘floors’. 

    Most pension schemes have benefits that increase at different rates as a result of changing legislation and so this change may be limited to a portion of a member’s benefits.  The impact of any caps to any pension increases will also potentially reduce the impact of this change.  If RPI calculated before the change would have been 3.5% and under the new methodology is 2.8% then, where pension increases are capped at 2.5%, there would be no change to the increase applied.

    In any event, we expect many members will feel like this is a reduction in their pension promise and telling members that the current calculation in RPI is overstated, due to an error in the formula, will provide little comfort.

    For pension scheme sponsors the impact will also depend on what measure of inflation is used when calculating pension increases and how much inflation hedging is in place. Most assets that hedge inflation are hedging RPI as there are very limited assets that are linked to CPI. 

    The following comments assume that no compensation is paid to holders of RPI linked assets.

    Where the pension increases are linked to RPI and there is RPI inflation hedging in place then, assuming 100% of the assets are hedged, the funding impact should be negligible.  Where schemes are under-hedged then their funding level should improve.

    Where pension increases are linked to CPI, but the scheme has inflation hedging that is linked to RPI, then the funding level will reduce and the more the scheme is hedged the greater the reduction.

    Therefore, in the same way where there has been a lottery in whether pension scheme increases are linked to RPI or CPI, there will be a lottery in the impact of any change to the calculation to RPI for scheme sponsors.

    Trustees will be tasked with explaining the impact to members where their pension increases are affected and liaising with sponsors on the funding impact of any change on the pension scheme.  Some trustees may review their investment strategy in light of this expected change, but there is still uncertainty around timing and there are very limited alternative assets that will provide any form of inflation hedging based on CPI or CPIH.

    Unintended consequences?

    The next question is whether this change will have a knock-on effect on statutory revaluation.  It seems non-sensical to retain two different measures of inflation when calculating pension increases.  So, will the government move CPI linked pension increases to CPIH?  This will certainly simplify pension scheme administration and may be easier to explain to members. In addition, since CPI and CPIH have been very similar, it is less likely to have a material impact to the level of benefits being paid.

    It will also mean that trustees are able to hedge the inflation risk in their scheme more accurately as there will be a single measure of inflation used for both pension increases and hedging assets which are typically index-linked gilts.

    Will Covid-19 affect timing?

    The government currently pays coupons and the redemption proceeds on index linked gilts based on RPI increases. Were RPI to change so that the amount the government pays out reduces then this will have a positive impact on government finances.  Over recent months, the government has been dealing with the Covid-19 pandemic and this has resulted in considerable government borrowing which will need to be paid back at some point.  Given this, one potentially attractive way to save money would be to bring forward the date of the change in RPI to 2025 (or even earlier) rather than waiting until 2030. 

    A further potential perceived advantage could be that this does not involve raising money from the general public directly but reducing the amount paid out by the government.  

    More pension increase litigation?

    And what of trustees? If CPI-H replaces RPI then will some trustees feel moved to ask for changes to be made to scheme benefits to compensate pensioners (such as the 65-year old mentioned above)? Or will we witness potentially acrimonious litigation between trustees and employers, along the lines of the very well reported British Airways pension increase cases?

    It can’t be ruled out but the prospect of more litigation is hopefully remote. When statutory revaluation and pension increases were changed from RPI to CPI by the government in 2011, we saw few trustees look to amend benefits or use discretionary powers to pay greater increases.  The British Airways case was unique and trustees will remember that their duty is to pay benefits in line with the scheme rules.  Where trustees have the discretion to pay further pension increases, they will need to consider carefully the funding position of the scheme and the employer covenant.

    Watch this space

    Changes in pensions indexation looks set to become one of the key pensions issues of this decade.

    Although similar in terms of impact on members to the change in statutory revaluation and pension increases when they were changed from RPI to CPI, this change has a far wider impact.

    This is because the change will be felt by more than just pension scheme members. It will impact holders of Index Link Gilts, which includes pension schemes as well as the many other parties to contracts that are linked to RPI, such as over the counter derivatives, student loans, infrastructure contracts, PFI loans as well as real estate contracts.

    Therefore, we would expect a much greater focus on this change, although this might be tempered due to the need of the government to raise money due to the extraordinary spending resulting from the pandemic.

    Share article:
    • Published byJudith Fish

      Judith is a qualified actuary with over 20 years post qualification experience She is an Accredited Professional Trustee, based in the London office, who works on a range of different schemes. She believes that good governance and collaborative working results in...

  • Get in touch with us

    Call us on 028 9041 2018 or fill out the form below and someone will get back to you.