On Friday, Ben Bernanke published his report into forecasting and communication at the Bank of England. It called for a revamp of the central bank’s main economic model and, having written extensively about the subject, I was disappointed.
This was a Nobel Prize-winning former Federal Reserve chair donning the mantle of a management consultant, recommending changes that executives already wanted to make, not asking searching questions and avoiding controversial recommendations that Bernanke clearly believed were necessary.
The first four of 12 recommendations relate to internal plumbing at the BoE as a response to “significant shortcomings” in its economic modelling. There is nothing objectionable about them and they will help improve the storage of data and quality of economic modelling.
The key unaddressed question here was how on earth did the BoE’s management and governance arrangements allow its modelling to get into such a bad mess? This is something that parliament should probe further to prevent it happening again, since the BoE court of directors clearly did not deploy its governance function adequately. Over to you, Treasury select committee and Lords economic affairs committee.
The fifth recommendation is extraordinary. It calls for the BoE to examine forecast errors carefully, “particularly errors that are not due to unanticipated shocks to the standard conditioning variables”. The problem is that it expects all this should be done behind closed doors. It is not difficult to do it in public. The Office for Budget Responsibility does. I do the same in this newsletter below.
Governor Andrew Bailey hates to “do hindsight”, as he said again on Friday. This makes him look shifty. I don’t think he is, but unless you undertake full forecast evaluation in public and show you have learnt lessons, people will not accept your partial account that forecast errors were due to Russia’s invasion of Ukraine or data revisions.
Recommendation number six again demonstrated the dysfunction of the BoE and contained my favourite chart of the report. It showed that the central bank employs many PhD researchers in the core areas of modelling, the majority of whom spend little or no time on the core function of the central bank. How has the governance of the BoE not noticed that it is employing expensive economists who seem to be using the bank as a vehicle for academic research without the hassle of teaching students? Again, something for parliament to pick up.
The remaining recommendations were public-facing and essentially called on the BoE to disown its own central forecast, insert some ad hoc scenarios and not much more.
Bernanke clearly wanted the BoE to address the key problem that the forecast is predicated on assumptions that the Monetary Policy Committee does not believe and therefore can give strange results that are impossible to explain publicly. “Following the practice of some of its peers, the Bank might at some point consider replacing the market-determined rate path used in the economic forecast with a rate forecast by the Bank itself,” he wrote.
The solution of disowning the current central forecast and adding in some scenarios will lead journalists to ask: “What is the scenario for interest rates that would stabilise inflation at your target of 2 per cent?” It is a good question and needs to be asked repeatedly.
But Bernanke instead dodged the issue, saying a “change would be highly consequential and this report recommends leaving decisions on this issue to future deliberations”.
By Bernanke’s own reasoning, this makes his report inconsequential. I agree.
US consumer price shock
The US consumer price inflation data last week came as a shock. For the third consecutive release, monthly core CPI inflation was 0.4 per cent, rounded to one decimal place. This was well above expectations and destroyed the notion that January’s data was affected by seasonal adjustment difficulties and February’s was a blip.
Financial market expectations that there would be no cut in the Federal Funds rate in June leapt from 43 per cent to 83 per cent last Wednesday, before coming back a bit. While team transitory hunkered down, President Joe Biden made excuses.
Let us be clear. Regarding US inflation in the first quarter of this year, financial markets, economists and I got it wrong. The chart below shows how much professional economists were out. It compares their forecasts every three months with the out-turn for annualised inflation in the first quarter. As recently as February this year, the collective wisdom of economists was for annualised core CPI inflation to be 3.1 per cent in the first quarter and we now know the actual figure was 4.2 per cent. That is quite a miss.
In the chart, you can see the equivalent data for the personal consumption deflator series, with forecasts again compiled by the Philadelphia Fed. The likely out-turn is based on an expected March monthly PCE price index inflation of 0.3 per cent and the miss is proportionally greater, leaving more for the Fed to think about because PCE, which strips out volatile food and energy costs, is its favoured inflation measure.
There are three possible responses to forecast errors of this scale: pretend they did not happen; note that it is easy to be wise with hindsight and that everyone made the same error without it; or undertake an examination of what went wrong and learn some lessons.
Since I made the same error independently of other economists in January, it is time for a reckoning. Was inflation’s overshoot the result of data revisions that were not known at the start of the year? No. Was it the result of unforeseeable shocks such as Russia’s invasion of Ukraine? No. Could it still be a blip? Possibly.
Most likely, though, it represents an error of judgment in underestimating the strength of demand in the US economy and overestimating the supply response. This error was compounded by a second one, which was to get overexcited by the three months of good inflation data at the end of 2023. I was clearly seduced by those figures and was too hasty to extrapolate, committing what I termed “time crime” back in October when I outlined the mistakes people make analysing inflation.
There is now a need to wait for more data before taking a definitive view and Fed officials queued up to make that point late last week.
There’s absolutely, in my mind, no urgency to adjust the policy rate,” Mary Daly, San Francisco Fed president.
“I would prefer to be patient and wait for clear and convincing evidence that inflation is on track to hit our 2 per cent target before adjusting the stance of policy,” Jeffrey Schmid, Kansas City Fed president.
The current economic environment “calls for a lot of patience”, Susan Collins, Boston Fed president.
We are “not yet where we want to be”, Thomas Barkin, Richmond Fed president.
A false knot across the Atlantic
I am going to stick my neck out and say financial markets are being far too quick in extrapolating poor US inflation data across the Atlantic to Europe.
The chart below examines financial market interest rate expectations since the start of the year, showing that they expect a slower path of rate cuts. This is true for the Federal Reserve, European Central Bank and the Bank of England and there is not a lot of difference between how much rate expectations have changed on both sides of the Atlantic. There has been little movement in expected Japanese interest rates even after the Bank of Japan raised rates out of negative territory in March.
If I take two points — December 2024 and December 2025 — and examine how much higher financial markets expect rates to be at these points than at the end of 2023, you can see below that rate expectations have moved together across leading central banks.
Some of that reflects common trends. But more recently the movements have been crackers. Look at the jump in rate expectations last Wednesday in the chart below when the US CPI inflation data was published. The expectation of UK interest rates in December 2024 rose 0.26 percentage points in one day because US inflation was 0.1 percentage point higher than expected. Madness.
ECB president Christine Lagarde reversed a similar movement in the eurozone by confirming the likelihood of a June rate cut. And by pointing out the obvious. “We are data-dependent, we are not Fed-dependent,” she said.
What I’ve been reading and watching
-
Andy Haldane, former chief economist of the Bank of England, gives his own verdict on the Bernanke review with some lovely detail on how forecasting really works and is “largely performative”.
-
It is cheering that ahead of the IMF and World Bank spring meetings this week, the Brookings-FT indicator of global economic growth is looking up a little, spurred on by the US and India.
-
The US data drew out team persistent in the US Democratic Party. Lawrence Summers was on multiple outlets saying a US rate cut in June would be a dangerous and egregious error. Jason Furman thinks underlying US inflation is between 2.5 and 3 per cent and getting that down to 2 per cent will be difficult.
-
You need to read the quotes from anonymous ECB governing council members demonstrating their independence from the Fed that my colleague Martin Arnold collected. “We are not Switzerland,” was my favourite and will annoy everyone.
A chart that matters
Before you get carried away by a hot US economy, this chart in IMF managing director Kristalina Georgieva’s curtain-raising speech for the fund’s spring meetings puts the damper on things. Yes, the IMF will nudge up its forecasts, but five years ahead it thinks the sustainable world growth rate has been coming down fast.