We’ll need to be slightly careful about this, no?

The Governor of the Bank of England has urged the Government to abandon the use of the retail prices index (RPI) as a measure of inflation, especially in the issuance of government bonds.

The intervention by Mark Carney could have profound implications for both savers and borrowers because RPI tends to run at 0.7 percentage points higher than the consumer prices index (CPI), which is increasingly considered to be a more reliable measure of inflation. Nevertheless, RPI is still widely used in government contracts.

“It would be helpful to have just one public-facing measure of cost of living for consumers,” Mr Carney told the House of Lords’ economic affairs committee. “At the moment we have the RPI, which most people acknowledge is of no merit, and CPI, which virtually everyone recognises and is the target in our remit.

“At some point it would be good to consolidate to focus on one [measure of inflation]”.

The change would mean that savers who invest in inflation-linked gilts would likely earn a lower return. With £311bn of such index-linked gilts in ­issuance, a 0.7 percentage point change in interest could save the Government in the region of £2.2bn per year.

There is no one measure of “inflation” so discussing which one we should use makes sense.

But there’s a significant difference between issuing new bonds which use the CPI and changing the old bonds in issue to use it.

Actually, I think it would be a great idea to issue the new ones using CPI. Then we can see, from he difference in market prices, what people think about it…..

18 comments on “We’ll need to be slightly careful about this, no?

  1. Given that the basket of goods and services RPI and CPI are based on are so comprehensively fiddled with I think they are largely irrelevant to most people

  2. Government is a lagging indicator. The fact that they’re switching to CPI years after everyone said they should, means that they’ve picked the bottom of the market (see also Gordon Brown’s gold sale, or the MoD’s housing sale-and-leaseback). We can confidently expect that CPI will henceforth be higher than RPI.

    And what about the civil servants who calculate the CPI – are their pensions based on CPI or RPI? Incentives matter…

  3. The way they cover housing costs is key for most people. Doesn’t CPI use an invented imputed housing cost and RPI just ignores housing? How long before Mark Carnage leaves the country?

  4. “savers who invest in inflation-linked gilts would likely earn a lower return”: bollocks. Really bad economics. A child at his father’s knee could knock that one on its head.

  5. “could save the Government in the region of £2.2bn per year”

    That’s nonsense. If lenders were willing to accept a lower interest rate, the government could have lowered the “plus” part of the current RPI+ interest rate on its gilts. (or are they still at RPI-, in which case increase the minus)

    The lenders’ expected interest rate will be roughly the same in each case.

    The government might save money in the future, depending on how the two indices move against each other, but equally they might end up paying more if the relative movement is the other way. That’s not a saving, that’s a risk premium.

  6. Diogenes said:
    “The way they cover housing costs is key for most people. Doesn’t CPI use an invented imputed housing cost and RPI just ignores housing?”

    I thought RPI does include house prices, but CPI doesn’t?

    However it’s complicated because there’s also RPI-X which doesn’t include house prices (X for excluding housing), and CPI-H which, as you say, uses imputed rental values for housing.

    Hasn’t the government recently switched to CPI-H (presumably because they hope rents have peaked, so falling rents will reduce CPI)?

    Depends which versions Carney is proposing using.

  7. Didn’t they change the BoE inflation target measure from RPI to CPI but the kept the target at 2%?

  8. Thanks for the correction, Richard!

    Dearieme, if one bond is following an index that says that inflation is 2.5% and another index is saying 2.3%,do you think you would earn the same return on each bond?

  9. ” if one bond is following an index that says that inflation is 2.5% and another index is saying 2.3%,do you think you would earn the same return on each bond?”

    And if you are selling two bonds, one giving a 2.3% coupon and one at 2.5%, do you expect the price to be the same?

    If so, I have bonds I’d like to sell you

  10. “do you think you would earn the same return”

    Short answer to this specific question is “yes” – the yield would be the same: the trading price of the asset would adjust to make it so. That is very much the point.

  11. I’d give the BoE a separate inflation target for Housing Costs ( cost of square footage of living space per person ). 0% seems reasonable to me for an established technology.
    They’d have to be allowed to issue the planning permits, which would be fun.

  12. @ Diogenes
    The *expected* return will be the same if the market is dominated by pension funds. The actual returns may differ if the difference between reported values of the indices changes in an unanticipated fashion, or if the Chancellor of the Exchequer has the law changed to permit early redemption of one,but what the Pedant-General says is true under all normal conditions.
    If the market is dominated by taxable persons then the return on each will reflect a compromise between the tax positioon of the various investors and one may have a lower return for basic rate taxpayers balancing a higher return for higher higher and the other a higher return for basic rate taxpayers balanced by a lower return for higher rate taxpayers. But that’s nit-picking.

  13. john77, when I was involved in contractual negotiations in the early 2000s, and indexation came under consideration, the question of whether to use RPI or CPI often cropped up. I think that people often used to opt for RPI on the basis that it was always higher than CPI, because that was what they heard from the media. Om being shown the fact that they moved in ostensibly random ways against each other, sometimes CPI higher than RPI but mostly RPI higher, they paused for thought. This is an honest question, and no I do not want to buy a bridge, would UK gilts inevitably tend to the same yield if linked to inflation indices compiled on different bases? If they did, would that mean that government bonds in other jurisdictions tend to the same yields?

    I guess this is asking whether the bond markets ignore the niceties of index construction in order to minimise potential arbitrage between different indices?

  14. Diogenes said:
    “Dearieme, if one bond is following an index that says that inflation is 2.5% and another index is saying 2.3%,do you think you would earn the same return on each bond?”

    The bond interest rate is index +X%, or index – X%.

    If one bond is index +0.5%, and the index is expected to average 2.5% over the life of the bond, that’s an expected 3%. If the other bond tracks an index that’s expected to be 2.3% over the life of the bond, it will almost certainly be priced at index + 0.7%.

    And if it isn’t, investors won’t buy it at £100 for a £100 bond; they’ll only buy it at slightly less so that the expected return comes out at 3% (in our example) of whatever they pay (that’s how big bonds are sold – the government doesn’t issue them at face value, but at whatever the market will bear, more or less than face value depending on whether the market’s opinion of the interest rate offered is high or low).

    The only things I can think of that might make a difference are:

    1) if one index is more volatile than the other, investors will demand a higher expected return on the more volatile one (a risk premium, because it’s less certain what you’re getting).

    2) If investors have a particular desire to track one index rather than the other, they might accept a lower expected return on the one they want. This is where Carney might have a point – if a lot of pension funds have future obligations that depend on CPI, they might want some CPI-linked bonds to invest in, so that the investment returns match their future cost increases. For them, a CPI-bond is lower risk, because their payment obligations are also going up by CPI, so they would accept a lower risk premium. I don’t know how much that is the case; I think the current situation is that some pension funds have switched to CPI but many are still on RPI, either because they just haven’t changed or because their scheme rules are drafted more tightly than others so they can’t change.

  15. Anecdotally, the only things I am paying more for in absolute £ terms compared to last year are things that go up every year by RPI, such as the ground rent on my flat and train tickets.

    My energy prices have gone down, and my grocery spending has remained the same (and I do notice when packages get smaller but remain the same price).

    Technology – the only thing that has gone up in price is graphics cards 😀

Leave a Reply

Name and email are required. Your email address will not be published.

This site uses Akismet to reduce spam. Learn how your comment data is processed.