A midterm reprieve; the Bank and the OBR can’t forecast Christmas
The Democrats’ midterm sweep pulled oil and bond yields down, then the AI trade into a rout. The Bank of England raised rates on an inflation forecast I think is wrong, and its forecasts and the OBR’s now point in opposite directions.
First published in Noise Cancelling
This week has been crammed with interesting economic and geopolitical events that have important implications for financial markets. Undoubtedly, the most significant was the result of the US midterm elections, which saw widespread gains for the Democratic Party in both the House and the Senate, both of which the party now controls.
I won’t repeat here the details of the election results or their domestic political implications, which have been extensively covered in the media already. However, the financial market reaction has been very interesting and worth commenting on, as were the election results’ implications for the conflict in the Middle East and for energy prices, which have fallen dramatically in the two days since they were announced.
Less impactful but nearly as thought-provoking was the Bank of England’s decision to raise interest rates, which was announced today. The thinking behind this decision has got me quite exercised and will occupy a few column inches in this update, as will the latest Bank forecasts that accompany the decision, which are especially interesting coming so quickly after the bewildering nonsense of the OBR’s comprehensive update last week.
Without getting too bogged down in the minutiae, I will compare and contrast the two to illustrate how these pillars of the UK’s financial establishment are quite simply not fit for purpose. I should say upfront that this is not a comment on the capabilities of the individuals employed in these two organisations, but it is a conclusion driven by what I see as structural failings in the way in which the Bank and the OBR analyse the economy and construct their forecasts. These failings produce the contradictory output I am going to highlight, but few other commentators will even be aware of it.
I will conclude by briefly covering the subject of this week’s Noise Cancelling Show, which looks at this month’s planned Anthropic IPO. If it goes ahead in under two weeks (there is a chance it will be postponed), it will be a major milestone in this profoundly important but very young company’s remarkable five-and-a-half-year journey.
The midterms: relief, then an AI rout
Tuesday’s midterm election results were consequential for domestic politics in the US and will usher in a period of divided government in Washington. A Democrat-controlled Congress will have all sorts of powers that it will use to constrain the President, his legislative agenda and his appointments.
Some suggest this could lead to the next political gridlock, but financial markets have tended to do well in these circumstances. Either way, the initial reaction was pretty positive in both equity and bond markets around the world, aside from those heavily influenced by the AI trade, including Japan and South Korea.
In broad terms, the immediate consensus conclusions were that AI capex spending growth would slow following Democrat governors taking control in five states with the largest data centre investment programmes, that Trump’s tax-cutting and government spending agenda would be reined in and, perhaps most significantly, that through budgetary control Congress would limit funding for further escalatory action in the conflict with Iran.
The immediate consequences of these broad conclusions were that energy prices fell significantly (Brent was down over $25 a barrel in two days and UK natural gas prices fell over 10%), and this catalysed a strong rally in US government bonds, with ten-year yields falling to 4.87% having peaked recently at 5.3%. Regular readers will know that this sort of move will have consequences for the UK gilt market too, where ten-year yields also fell to 5.1%, having peaked in early October at 5.5%.
Today, this reaction has become even more pronounced. Ten-year yields have fallen again in the US, but the AI trade is getting battered. The Nasdaq index is down nearly 5%, and some of the leading AI plays are down considerably more, including Micron, for example, which fell 15% by mid-morning.
As ever, these sorts of moves can become very ‘emotional’ quickly, and I suspect this one is already becoming somewhat hysterical as excessively leveraged exposures unwind rapidly. Clearly, the fundamentals driving this industrial revolution are not party political; they are foundational to the future prosperity and technology leadership position of the US economy and, as such, will continue even with a Democrat-controlled Congress.
Consequently, although this hysteria may last a little longer, it will, in my view, be over soon, and some normality will be restored. The growth rate of capex deployment may slow in the US, but my sense is that this would have happened anyway because of other externalities like access to grid connections, and might just as easily be seen as healthy for the economy and the sector.
Oil: a change of sentiment, not supply
As for energy prices, although the crude price has bounced a little today, it is still down over $20 from its high of over $117 less than a week ago. This is quite odd given that all that has changed is sentiment. There have been no breakthroughs in the negotiations and no announcements of progress in talks, which are ongoing. The volume of crude leaving the Gulf has also not spiked but remains high and close to pre-war norms.
Either way the falls are clearly welcome news and will be reflected promptly at the pump and, if sustained, in monthly inflation data in the not-too-distant future. (I will come to this in a bit more detail below.)
The Bank of England: a rise I think is a mistake
The second big issue of the week was the Bank of England’s decision to raise interest rates on Thursday by 0.25%, which I expected but which I think is a mistake. In the statement accompanying the decision, the Governor’s comments were also seen as indicating that this move could be followed in December by another quarter-point increase.
Today’s increase shouldn’t be seen as the first in an inevitable long sequence of hikes. But unless the conflict in the Middle East is resolved, neither does it seem to be the last increase.
(As far as I could tell, the UK gilt market completely ignored the interest rate increase and carried on following what was happening in the Treasury market.)
Interestingly, given what’s happened to the oil price in the last two days, this stance might already be seen as out of date, albeit that the conflict in the Gulf is clearly not over. But what is clear to me, at least in the short term, is that the petrol price has peaked and, having already fallen by a couple of pence per litre over the last few days, is likely to fall further in the weeks ahead.
In the short term, if maintained, this will mean that headline inflation in October and November, and possibly in December, could actually fall from September’s 3.5%. The Bank of England, of course, is much more bearish (and wrong, in my opinion) and is forecasting that CPI reaches 3.8% in November and December and then carries on increasing in 2027.
This is where the Bank starts to look really silly. In what to me looks like a serious judgement error (I will come back to its egregious 2026 errors later in this note), the Bank is now forecasting that inflation will peak at about 5% by mid-2027 and then fall only slowly from there to reach 4% by year end.
Whilst its energy and food price expectations are justifiable, albeit in my opinion too high, its ‘other’ category, which includes all the other elements in the basket that tend to be driven ultimately by wage growth, is where I think they have gone a bit mad. Here once again the Bank has invoked the ghostly second-round effects by stating that although absent so far (another thing the MPC got wrong) they will now come through because elevated energy prices have lasted so long. This view is reflected in its forecast for private sector wage growth, which it says will accelerate to 3.5% next year from its previous expectation of 3% – which is where it is at the moment.
Against a backdrop of a generally weakening labour market characterised by lower vacancies and fewer people in work, next year will see the growing impact of AI on job losses in the service sector. In an environment where people will be finding it harder to find a job, I cannot see workers demanding higher pay increases. It just doesn’t stack up with reality.
I also think the Bank’s view is a basic misread of how corporates will react to higher energy prices and higher costs thanks to the government’s tax policies. In an intensely competitive economy like the UK, companies don’t all enjoy the luxury of raising prices as a reflex in an environment of higher costs. Instead, companies limit the costs they can control, and they look to drive greater efficiency into their organisations by shedding labour where they can. A good recent example might be Ryanair’s announcement this week, in which the company said that it might be forced to raise airfares next year if jet fuel prices remain elevated but that it was also considering a wage freeze for all its staff, not a 3.5% pay increase.
In conclusion, it looks to me as if the Bank’s serious misjudgement on growth this year is going to be replaced by a serious misjudgement on inflation next year.
For what it’s worth, I see inflation peaking early in the new year at just above 4% (following the 22% or so increase in the energy price cap in January) and then declining from March onwards as the economy annualises the massive increase in energy prices that started when the war kicked off at the end of February. If we get a settlement in the war, then of course the declines will be greater and more rapid.
The Bank’s growth errors
Before covering the glaring inconsistencies between the Bank’s newly updated forecasts and last week’s OBR numbers, I wanted to write a few lines on the Bank’s growth forecast errors this year. At the start of the year the Bank had a reasonably sensible forecast of 1.6%. However, once the war started the Bank became very gloomy and it downgraded its annual growth forecast to 1.1% in July, despite data from the ONS showing strong growth in Q1 having continued throughout the summer.
Indeed, I said at the time it was wrong and was way too low. Part of the problem was that its forecast for business investment was laughably wrong. In fact, it was forecasting business investment to fall this year by 1%. Today it quietly upgraded that to growth of 4.3%.
As part of that annual growth downgrade, the Bank said that growth in Q3 would only be 0.1%. This week, in another very quiet change, with no proper explanation of its previous error, the Bank upgraded its Q3 forecast to 0.7%.
This is a massive increase which appears to have attracted no attention in the media, but which to me stands out like a sore thumb. Another notable line in the 92-page Monetary Policy Report is the Bank’s acknowledgement of much faster growth at the back end of this year. In fact, the Bank now says that the growth rate in Q4 2026 on Q4 last year will be as high as 2.2%, yet another pretty important change that has completely eluded the media.
These changes aren’t just tweaks in the normal course of forecasting error. They reveal giant-sized misjudgements and follow the Bank’s unfathomable misread of the country’s business community, reflected in its lunatic forecast for business investment.
For a central bank that claims to have its finger on the pulse of corporate Britain, this is completely inexplicable and leaves me wondering why anyone would believe that it can make good judgements about what’s happening in the economy and set monetary policy appropriately.
I am very confident that it is wrong again, and this time about inflation, and as a result its decision today to raise rates, combined with its heavy hint that it will repeat the move in a few weeks’ time, looks like a policy error to me.
Fortunately, it is not sufficiently significant to derail the strong growth momentum in the economy that I have been writing about all year, but it is far from helpful.
The Bank and the OBR, one week apart
Which brings me to what I can only describe as the comic discrepancies between the Bank of England’s latest wordy insights and forecasts and those handed down by the OBR only last week in its 200-page equivalent. This time I’m not going to dwell too long on both organisations’ catalogue of errors this year but focus on 2027. But, just as a reminder, in March the OBR forecast that the economy would grow by 1.1% and last week upgraded that to 1.2%. Today the Bank said growth would be 1.6% in 2026.
So, what we now know is that the OBR’s newly minted forecast that we were told about only last week, with the ink barely dry on the numbers, is already wrong – literally within a few days. In fact, I am very confident that both the OBR and the Bank are wrong and that the outcome will be closer to 1.8%, but with the year 85% complete, how can the OBR get it that wrong? After all, it had the benefit of three quarters of data to base its upgrade on, and it has still cocked it up. It really is staggeringly bad. But it gets even worse.
One week apart, this is what’s happened to the two organisations’ growth forecasts for 2027:
| 2027 GDP growth | Previous | Updated |
|---|---|---|
| OBR | 1.6% | 1.0% |
| Bank of England | 1.1% | 1.3% |
As an aside, I am confident that both are still wrong, and by quite a margin. I expect growth in 2027 of 2%.
One downgrades by 0.6% and days later the other upgrades by 0.2%, and the backdrop here is that the economy exits 2027 with a forecast quarter-on-quarter growth rate of 2.2%.
Quite what one is supposed to make of this Horlicks is anyone’s guess, but one has to wonder if these organisations are best placed to guide fiscal and monetary policy decisions when they can’t get these fundamentally important forecasts anywhere close to the outcomes and, within days of each other, their expectations move in different directions.
One might imagine that they were analysing different economies altogether. What’s more galling is the fact that these two pillars of the UK establishment are supposed to be making better judgements than the rest of us. They have the responsibility to guide policy, they have the resources they need, the data, the surveys, the network of contacts across the country and very big budgets.
Rightly, the expectation should be that they should be the most accurate forecasters. The reality is that neither of them could forecast Christmas.
Why they keep getting it wrong
This Weekly is already longer than I had planned and so I won’t write too much on why I think these organisations are getting their forecasts so wrong. However, in summary, I suspect that they are both suffering from the same availability heuristic bias.
Basically, their forecasting process is excessively influenced by what’s just happened rather than by a well-thought-through analysis of what might change in the future and how quickly. This bias works most obviously in the OBR model through its conditioning assumptions, which are based on market-derived futures prices, and at the Bank, my guess is that its reliance on consumer and business expectations surveys does exactly the same thing.
Futures prices and expectations surveys suffer from the same problem. They reflect current sentiment and are a terrible guide to what will happen beyond the immediate future. This fundamental characteristic is not a well-kept secret and has been reflected in countless academic studies.
So why do the OBR and the Bank of England rely so much on them? This is a much harder question to answer, but I suspect it’s because the decision-makers in these organisations are just not very good at making judgements or are frightened to do so. Either way, continuing with the current approach, which delivers such a poor track record, seems once again to be an excellent example of Einstein’s definition of madness.
Anthropic
As for the Anthropic IPO, following the tech rout on Thursday and the weakness of a number of high-profile stocks in the AI sector in recent weeks, there are rumours that the IPO will either be delayed or the valuation trimmed from $1.5trn to something meaningfully lower.
I suspect that more information will emerge on this after the weekend.
What to look out for next week
It’s a busy week next week, with lots of important economic data both in the US and in the UK. The highlights are US inflation on Tuesday (October’s data) and UK GDP data for September on Thursday. If September’s outcome is no growth, then Q3 as a whole will deliver growth of 0.7% – the Bank of England’s recently upgraded expectation.
There is also the usual diet of labour market and survey reports too. It’s also a moderately busy week in the UK for mainly interim results and scheduled trading statements, which I will be keeping a weather eye on, but I suspect most attention will be focused on the technology sector and the ‘AI trade’ in particular. My hunch is that the pretty extreme sell-off we have witnessed in these stocks in recent weeks is now largely past.
This week’s 3,000 words in three charts
A bad night for Trump, but not the worst
Clearly Tuesday was a bad day for President Trump but, interestingly, not as bad as the midterms were for some of his predecessors.
Tuesday was bad for Trump; in the House, 1994 and 2010 were worse
| Year | President (party) | House | Senate |
|---|---|---|---|
| 2026 | Trump (R) | ≈ −25 | ≈ −7 |
| 2022 | Biden (D) | −9 | +1 |
| 2018 | Trump (R) | −41 | +2 |
| 2014 | Obama (D) | −13 | −9 |
| 2010 | Obama (D) | −63 | −6 |
| 2006 | G.W. Bush (R) | −30 | −6 |
| 2002 | G.W. Bush (R) | +8 | +2 |
| 1998 | Clinton (D) | +4 | 0 |
| 1994 | Clinton (D) | −54 | −8 |
Mortgage approvals keep falling
The MPC’s decision this week will clearly not help here.
UK mortgage approvals are 19% lower than a year ago
Paying wind farms not to generate
This data contradicts the many government promises about the savings consumers would see from a predominantly renewable energy system. It appears that payments to wind farms not to produce are going up significantly.
Britain’s grid balancing bill has set two monthly records in a row
Important information
The views expressed in this article are those of the author at the date of publication and not necessarily those of Curated Capital Ltd. The contents of this article are not intended as investment advice and will not be updated after publication unless otherwise stated.