Conditioning and the Bank of England’s credibility problem

30 June 2023/No Comments
By Nick Dunbar

Like hair, a loss of credibility is hard to reverse once it happens. And the Bank of England has a credibility problem, worsened by not a hair product but because of the way it conditions its forecasts.

On 23 May, governor Andrew Bailey and other Monetary Policy Committee members conceded to the Treasury Select Committee that their inflation forecasts had been wrong, as a result of issues in the pricing of food. At the same time, the Bank’s chief economist Huw Pill conceded that its forecasting models were based on only the last 30 years of data, excluding the inflationary 1970s and 80s.

What in May had been a crisis of food costs, in June became a housing affordability crisis, when the Bank imposed a 50-basis point rise in base rates as a fightback in its credibility wars. In a sign of the struggle it faces, Pill’s predecessor as chief economist, Andy Haldane, has now openly called for the Bank to pause its rate hikes.

While the UK’s sclerotic productivity and febrile politics are not the Bank of England’s remit, much of the Bank’s problems stem from what is called conditioning. Rather than trying to predict future interest rates along with its forecasts of inflation, GDP and unemployment, the Bank says its forecasts are ‘conditioned’ on rates implied by the market.

Forecasting with the swap market

But where does the Bank obtain these market-implied forecasts and how good are they? They come from the forward sterling overnight interest-rate swap (OIS) curve, which has replaced the now-abolished LIBOR curve. This curve is constructed by taking the slope of the spot OIS curve, which for a given maturity is the fixed rate received in exchange for a stream of floating SONIA payments.

If you go long a 13-month swap and short a 12-month swap, then in effect you are betting on the 1-month rate in a year’s time. If enough people make these bets, then the swap market functions as a betting market that ‘predicts’ future rates. At least, that’s according to the textbooks.

In practice, a range of counterparties – banks, asset managers, hedge funds, corporate borrowers and pension funds – trade in spot interest rate swap contracts every day, mainly for hedging purposes. Only a minority will trade in order to express a view on the forward OIS curve. So the Bank’s claim that this curve represents a considered market view on future UK monetary policy is shaky.

Forecast time horizon Six months One year Two years
This visualisation shows the forward sterling overnight index swap (OIS) rate for a chosen time horizon, along with the spot OIS rate. For each date on the horizontal axis, the spot OIS rate is plotted with the forward OIS rate (or market forecast) offset by the time period. So for a time horizon of one year, the 12 month forward OIS rate is displayed, allowing us to compare it to the actual OIS rate at the end of the time horizon.
Data source: Bank of England

To see how good a predictor it is, we have created a visualisation. Pick a time horizon – say one year. Over time we compare the one-year forward OIS rate with the actual OIS rate a year later. Most of the time, the ‘market prediction’ isn’t that good, particularly in times when monetary policy is changing.

Related Articles