LONDON (Reuters) - Britain’s top statistician unexpectedly decided against major changes to the country’s longest-running inflation index on Thursday, rejecting a move that could have significantly cut government borrowing costs.
The decision to keep the Retail Price Index in its current form is a major boost for older Britons who have company pensions linked to RPI, as well as to holders of Britain’s inflation-linked government bonds.
Changes to RPI had been expected to bring it closer to the lower Consumer Price Index measure of inflation used by the Bank of England, and prices for 10-year index-linked gilts soared on Thursday’s news, taking yields to a record low.
RPI was developed after World War Two, but in recent years it has increasingly diverged from CPI and for months Britain’s Office for National Statistics had been consulting economists, fund managers and other users of the data on its proposals.
Most economists had believed that the ONS was set on change, but in a statement Jil Matheson, who supervises the ONS and is the government’s chief statistical adviser, said she had decided to keep RPI and create an alternative parallel index.
“There is significant value to users in maintaining the continuity of the existing RPI’s long time series without major change, so that it may continue to be used for long-term indexation and for index-linked gilts and bonds,” Matheson said.
There is a big question mark over whether the new index, known as RPIJ, will find favour. The government said new index-linked bonds would continue to be linked to RPI, not RPIJ. C P I , which is based on the standard European Union measure of inflation, is Britain’s most widely used alternative.
“The RPIJ may find no takers and exist in obscurity. At the moment, we do not plan to forecast it,” said Nomura economist Philip Rush.
The ONS plans to have the RPIJ measure in operation for March data onwards.
David Dyer, a portfolio manager at AXA Investment Managers, said he welcomed the decision, and that it was unexpected.
“It is a surprise ... though of course we were arguing for this on the basis that RPI is a very long-standing index used by many parties in striking long-term financial contracts. It was not just fund managers who were arguing against the change, but also other investors,” he told Reuters.
“At a time when savers have been hit, this would have hurt them even more. But on the other hand, when the government is cutting the deficit, this would have been a handy saving for the chancellor to have,” he added.
Some economists had estimated cash-strapped Chancellor George Osborne would have saved up to 3 billion pounds a year in interest payments if RPI had been changed, out of annual debt servicing costs of 47 billion.
Moreover, many experts - including BoE Governor Mervyn King, who commented on the issue in November - believe the techniques used to calculate RPI are outdated, and see little statistical justification for keeping them.
“I think it’s remarkable that they have said that the current formula does not meet international standards, and then continue to use it for a number of key functions,” said Sam Hill, a bond strategist at Royal Bank of Canada.
The ‘Carli’ averaging method used in RPI - which will be dropped in RPIJ - is not used in CPI nor by most other countries, and has been criticised by the International Monetary Fund for giving an upward bias to RPI inflation.
Its continued use will contribute to RPI inflation averaging more than a percentage point above CPI in future, with the bulk of that due to the statistical methods used rather than differences between the baskets of goods used for CPI and RPI.
However, an overwhelming majority of respondents to the ONS consultation opposed a change to RPI, citing the need for long-term consistency in the series.
The ONS said this was a legitimate objection, even if it implied the continued use of outdated statistical techniques. The consultation did not require respondents to say if they would financially benefit from opposing change to the index.
Annual RPI inflation was 3.0 percent in November, compared to 2.7 percent for CPI, reflecting a recent narrowing in the gap between the two measures due to differences in how changes in insurance and mortgage interest payments are accounted for.
Additional reporting by Patrick Graham; Editing by Stephen Nisbet