In trying to understand how the Bank of England responds to economic news and ideas, it is important to know how its forecast changes every three months to understand the big picture of the Bank’s thinking on the economy. But this is impossible because the fan charts are a bad graphical representation of the Bank’s view. I have told the Bank many times why the fan charts are rubbish, but not until now written down the reasons.
Here are the Bank’s growth (green) and inflation (red) big pictures published from November (left) and February (right).
What it is possible to deduce from these charts:
- Not much. The Bank thinks year-on-year growth will be between +6 per cent and -1 per cent by the end of 2012. Inflation will be between +4.5 per cent and -0.5 per cent. The ranges are too wide to make any sensible inferences as to the Bank’s thinking. If that is what the Bank wants us really to take away from the charts, it might get a nasty surprise that we don’t think its input is very valuable at a time of budgetary stringency.
- Since, that is not what the Bank thinks, and Mervyn King, Bank governor, had a lot to say about the economy yesterday, what else can we glean? It appears the Bank has become gloomier about growth and thinks medium-term inflation will be lower but short-term inflation will be higher. This inference is sensible and comes from looking at the darkest parts of the charts, representing the Bank’s central forecast.
- The MPC has become very dovish compared with November. Remember that small changes in the inflation forecast can make big differences for interest rates. The rule of thumb many current and former Monetary Policy Committee members have used is that the Bank tends to move rates by 0.25 percentage points for every 0.1 percentage point movement in the inflation central forecast. Mervyn King, Bank governor, hates this mechanistic reduction, but it has proved a good predictor of the Bank’s reactions in his time as governor.
- The dovishness is probably greater because the February picture is based on the assumption of lower interest rates than the November picture. This is becasue the Bank conditions its forecasts on market interest rates which have fallen since November. There is a greater monetary push now in the forecast but the outcome is weaker.
- That’s about it. What you cannot do is look at the different shading and work out the probability the MPC places on different outcomes, something that would be indeed useful. Why not? Because it is wrong, for example, to look at the growth chart and see one shaded area below zero for 2012 and deduce that there the Bank sees a 5 per cent probability of the economy still contracting year-on-year two to three years hence. Why is this wrong even though there is an equal probability in each shading band? Because the distribution is shaded on a pairwise basis, so all you know is there is a 10 per cent chance the economy in 2012 will be contracting or that it will be growing between 5.5 and 6 per cent. You cannot know the split between the two. Useless.
The one thing that you absolutely cannot deduce from the big pictures is what Charlie Bean, Bank deputy governor, said yesterday was the true big picture:
“The big picture here is that, if you like, the average picture across growth - across the whole of the distribution – is about the same now as it was in November and that’s equally true for the inflation projection.”
The only way I could get close to this “real” big picture was to blow up the fan charts on my computer, measure the width of the distribution for every quarter of the forecast to get an estimate of the average and plug the results into a spreadsheet using the Bank’s “backcast” and the most recently published GDP figures.
To his credit, Mervyn King, Bank governor, always insists everyone should look at the full probability distribution of the charts and not just the central forecast, shaded in darkest green. At the press conference yesterday he said:
“I think it’s just very, very important – particularly in this situation - for you to explain to your viewers and your readers that forecasts are not numbers which say this will happen. They are judgements about the relative probabilities of various outcomes occurring. And of course there is always a possibility of these outturns. Many things can happen.”
But I defy anyone to look at what is published by the Bank, and arrive at the view the Bank wanted to give. It is impossible and the fan charts are simply deficient at giving the big picture. It is for these reasons that we have decided not to publish the fan charts in the FT anymore. We do not want to mislead our readers even if the Bank is happy for the inflation report to mislead its audience.
I understand the Bank governor is the sticking point within the Bank, the person insisting the Bank cannot publish more details about its forecast (means, medians, modes, sensible points on the probability distribution) for a week after publication. That position is terribly small-minded.
Again, to the governor’s credit, he has in the past relented from a stupid position with no ill effects. Until May 2007, he refused to let the Bank release a transcript of the inflation report press conference. Then the FT started publishing one, since there was a demand and it aided transparency. His opposition crumbled soon after. Let’s hope the governor can show he is as grown-up now, in recognising the deficiencies of the Bank’s inflation report fan charts, as he was in 2007 with the transcript.
