# The synchronised mistake

_The world's central banks turned hawkish in unison this spring, just as the shock that frightened them was already reversing. My mid-year report card, honestly marked, and where I think the second half goes._

Neil Woodford · 1 July 2026 · 16 min read

![Double SpaceX Falcon Heavy Landing](https://cdn.sanity.io/images/v3acfbvo/production/5d318528c986e80a3a6c5059e8f7cb45f938f628-3000x2000.jpg?w=1600&fit=max&auto=format)

---

In December, I wrote about the year ahead, and the picture I painted was an unfashionably calm one: inflation easing, rates falling, the doom-mongers wrong again, and the UK, of all places, the most attractive major market in the developed world. The consensus probably thought I had taken leave of my senses.

Then, on the last morning of February, the United States and Israel put nine hundred warheads into Iran in the space of twelve hours, killed its Supreme Leader, and the calmest forecast on the Street was suddenly staring at the largest oil-supply disruption in the history of the seaborne market.

A week into that war, on The Show, Jon asked me the one question worth asking in a crisis. I gave him the framework I have used for thirty years.

> Is this a four-week problem, or a four-year problem?
>
> — The Show, Episode 27 · 6 March 2026

My answer, six days into the shooting, was the former. I was a little optimistic about the timing. The war ran nearer sixteen weeks than four, and I will take the few months on the chin. But sixteen weeks is a great deal closer to four than to the two hundred that four years would have been, and on the only axis that matters for your savings, the duration of the damage, the judgement held. The conflict that was going to reset the world economy has just been closed by a memorandum signed at Versailles, with oil cheaper than when it started.

So let me begin where an honest letter should at the year’s halfway mark, with my own homework marked in red.

_[Watch: Neil gives his economic outlook in this episode of the Noise Cancelling show — Three Central Banks Made the Same Mistake in One Week](https://www.noisecancelling.co/the-show)_

## The report card

_Six months, one war, and a card I am mostly happy to stand behind._

**$144.42** — Dated Brent on 7 April, the highest oil price ever recorded

The headline calls have held up better than the year deserved. I said US rates would settle at 3.75 per cent after three cuts in 2025, and they did; the consensus that screamed "higher for longer" was, in the end, clean bowled by the data. I said American hyperscalers would spend more than $200bn building out artificial intelligence; they are now guiding to something nearer $700bn, which is less a forecast beaten than a forecast made to look timid. I said the UK would embarrass the perma-pessimists, and in the first quarter, it grew 0.6 per cent, the fastest in the G7, while the Bank of England's own staff were still talking the country into a recession that has not arrived.

And I said the great machine of consensus forecasting, the IMF foremost among it, would be wrong in the same direction again. It was. Its January forecast put global growth at 3.1 per cent, wrapped in the usual doom-laden commentary; six months on, the same institutions have marked the world down towards 2.5 per cent and called it a calamity. I would gently point out that an economy growing at more than 2.5 per cent amid a regional war and the highest oil price on record is not a calamity. It is a remarkably resilient organism, being narrated by people who would find the dark lining in a silver cloud.

Now the calls that have not worked.

I told you long-dated yields would fall by a hundred basis points this year. They have done the opposite. The US ten-year is higher, the gilt spiked towards 5 per cent at the height of the war, and the thirty-year reached its worst level since 1998. I will make the case below that this is delayed rather than wrong, that the energy spike which lifted yields is the very thing now reversing. I am not going to dress up six months of being offside as anything other than being offside. _(Neil in the margin: A basis point is one-hundredth of a percentage point, so a hundred of them is a full percentage point. The convention exists because bond markets move in such small increments that talking in whole points would be hopelessly imprecise.)_

I was also too sanguine on the European Central Bank, which I thought had reached the floor. It had not. It found a way to make the floor lower by going up, of which more shortly. And the call for domestic UK equities to lead the market higher is, at the halfway mark, unresolved. The FTSE 100 crossed 10,000 for the first time in its history in the first week of January and has gone broadly sideways since, which in a year with a war in it I will take, though it is not yet the vindication I am looking for.

That is the card. Two firm wins, several tracking, two offside. For six months that contained an actual war, I will take it.

## The shock that wasn't

_A genuine oil shock arrived, did its worst, and left almost no mark on the thing that actually matters._

**$72 → $144 → $72** — The biggest oil shock in a generation, round-tripped in ten weeks.

The sequence matters. On 28 February, the war began. By 4 March, Iran had declared the Strait of Hormuz closed, and roughly a fifth of the world's seaborne crude and a quarter of its liquefied natural gas had, on paper, nowhere to go. The insurance market moved faster than the missiles: war-risk premiums went from a quarter of one per cent of a vessel's value to as much as ten, and there was a morning in early April when the war-risk premium on a single cargo of crude exceeded the freight revenue for the voyage that carried it. On 7 April, Dated Brent, the physical price rather than the paper one, printed $144.42, the highest figure since Platts began publishing in 1987. _(Neil in the margin: The Strait of Hormuz is the narrow neck between Iran and Oman through which Gulf oil must sail. Its perennial place in war scenarios is precisely because so much supply funnels through one chokepoint a few miles wide.)_

This was not a drill. This was the energy shock the bears had warned about for a decade, arriving in full.

And then look at what it did to the thing that is supposed to matter. G7 core inflation in March averaged 2.3 per cent. American headline inflation did climb to 4.2 per cent by May, a number I will come back to, but more than sixty per cent of that rise was energy, and core goods prices actually fell. By the time the Islamabad Memorandum was signed at Versailles on 17 June, oil was back at $77, below where a good deal of the market had it before the first shot. The largest oil shock in a generation round-tripped in ten weeks and left the underlying price level of the Western world essentially where it found it. _(Neil in the margin: Core strips out food and energy. With oil at a record, it barely moved. That is the entire point: the shock landed in the volatile line, not the underlying one.)_

Why? Because oil prices are not set by the people who burn the stuff. It is set by an estimated twenty to thirty times that volume in financial transactions, and financial markets price the resolution, not the headline. The seaborne oil market is fungible in a way that Russian pipeline gas in 2022 was not. The American shale patch is a shock absorber that the 1970s did not have. And the G7 economy now uses roughly half as much oil per unit of output as it did when the textbooks our central banks still lean on were written. _(Neil in the margin: This is the paper-to-physical ratio: futures, options and swaps changing hands many times over for every actual barrel delivered. It is why the screen price reacts to expectations and positioning rather than to who is queueing at the pump.)_

![An oil price spike, round-tripped in ten weeks](https://r4at4qm6kmohrtvq.public.blob.vercel-storage.com/charts/AWlTQI01dswpK5un9WI1zK-5fc65926a360-light.png)

So the lesson of the first half is not that the world is dangerous. The world is always dangerous. The lesson is what happened next. A textbook supply shock hit the most feared chokepoint on earth, oil doubled, and within a single quarter the price had round-tripped and core inflation had barely moved. The shock was real; its lasting imprint on prices was not. The people whose entire job is to understand that mechanism appear to have missed it completely.

## The synchronised mistake

_Faced with an energy spike that was already reversing, Frankfurt, Washington and Tokyo all reached for the brakes at once._

**2.25%** — The ECB's deposit rate after June's hike, its first since 2023

There is a particular kind of error only a committee can make, and we have just watched three of them make it together.

Start with Frankfurt, the most egregious of the three. On 11 June, the European Central Bank raised its deposit rate by a quarter point to 2.25 per cent, its first increase since 2023, citing the energy shock. It did this the day before the peace process became public, on a spike that had already rolled over and sat some $38 a barrel below an April peak from which it would ultimately halve, and by unanimous vote of a twenty-six-person committee that then spent the following week explaining, at length and on the record, why it had been right. I called it tokenism at the time, and I see no reason to soften the verdict. The ECB spent two years cutting because growth was on the floor; the euro-area economy grew one tenth of one per cent last quarter; and the institution chose this moment, on this provocation, to tighten. It will be reversed, and the reversal will be dressed up as foresight.

Tokyo at least has a case. The Bank of Japan took its policy rate to 1 per cent on 16 June, the highest since 1995, on the back of a genuine wage round; the shunto delivered 5.46 per cent, a third straight year above five. That is real, domestically generated inflation of the sort Japan spent thirty years praying for, and yet the yen still sits near 161 to the dollar, weaker than where it started, which tells you the market thinks even this central bank is behind. Japan is the one place tightening into strength rather than into a mirage. It has my sympathy. _(Neil in the margin: The shuntō is Japan's annual 'spring wage offensive', the coordinated round of pay negotiations between big employers and unions. After decades of near-zero settlements, a number above five is genuinely a regime change.)_

![Everyone hit the brakes at once](https://r4at4qm6kmohrtvq.public.blob.vercel-storage.com/charts/RPLhY3joW8ryZXw3laIpSZ-f598750f53c3-light.png)

Which brings us to Washington, and the most consequential change of personnel in the whole story.

## The Warsh Fed picks a fight with the wrong number

_Kevin Warsh wanted a quieter, more disciplined Fed. He has one. He has also aimed it at an inflation rate that is about to fall._

**4.2%** — US CPI in May, of which more than sixty per cent was energy

Jerome Powell has gone, kept on the Board in a gesture towards continuity, and Kevin Warsh has the chair, confirmed in May by the narrowest margin in the modern history of the institution. On the question of how a central bank should communicate, Warsh and I are in violent agreement. His first statement ran to a hundred and thirty words against three hundred and forty-one for the last of the Powell era. He has stripped out the easing bias, the forward guidance, the constant tending of the market's feelings. A central bank should be a little obscure and a great deal quiet. Greenspan, who sadly died aged 100 this month, understood this. The twenty-six-person ECB, issuing press releases to justify a hike nobody needed, is the living counter-example. _(Neil in the margin: Alan Greenspan was an American economist who served as chairman of the U.S. Federal Reserve from 1987 to 2006, making him one of the most influential central bankers of his era. Born in New York City in 1926, he became known for steering monetary policy through decades of growth, low inflation, and major market shocks)_

So far, so good. The trouble is the number he has chosen to fight.

American headline inflation reached 4.2 per cent in May, the highest in three years, and the new projections now carry a hike: the median official sees the funds rate ending the year at 3.8 per cent, up from the cuts that were pencilled in only in March. _(Neil in the margin: The federal funds rate is the overnight rate US banks charge each other, and the lever the Fed actually sets. Everything else — mortgages, corporate borrowing, the dollar — keys off it, which is why one committee's median dot moves markets worldwide.)_

But take the number apart, as Warsh's own staff surely have. Core goods prices fell in May. The tariff pass-through the whole profession spent a year frightened of has been a damp squib: the Supreme Court struck the original tariffs down in February, the replacement is a ten per cent surcharge that expires, conveniently, around the 24th of July, and in any case the inflation is not in goods. It is in energy, which is to say it is in the oil price which we all watched round-trip over the last three months. The labour market that is meant to be generating a wage-price spiral added 172,000 jobs in May against an 80,000 consensus, with wages growing a perfectly civilised 3.4 per cent. This is not 1979. It is a temporary, energy-led bump in a headline number, and the man in charge of the world's most important interest rate is preparing to raise into the teeth of an energy reversal.

![The pop was energy, not the core](https://r4at4qm6kmohrtvq.public.blob.vercel-storage.com/charts/c9638216-8b67-48a8-84c7-da3990e7789d-171aeacc92a8-light.png)

My base case is that he never gets the chance. As the energy base-effects roll off through the third quarter, US headline inflation falls back towards the threes, the September hike the market prices at two-in-five odds quietly leaves the table, and the conversation by the turn of the year is about cuts again. The growth backdrop can wait: the Atlanta Fed has the second quarter running near 3 per cent. But the direction of the next genuine surprise is not in doubt. It is lower inflation and lower yields. _(Neil in the margin: A base effect is an arithmetic quirk: annual inflation compares today's price with a year ago, so when last year's spike drops out of the twelve-month window, the rate falls even if prices simply hold still. No actual disinflation required.)_

![Yields rose instead of falling.](https://r4at4qm6kmohrtvq.public.blob.vercel-storage.com/charts/e2d76e2e-1c14-4064-9acf-8e0a9a9cb37e-0374fb41974d-light.png)

**−100 bps** — Wrong today. Let's see what the next six months holds...

That is the heart of it. My January call for a hundred basis points off long yields was early; the energy shock pushed them the wrong way. But the shock has reversed, and the policy panic it produced is, I believe, the last domino. When the data confirm what the oil price is already showing, the rates market will reprice, and yields will fall again, back on the trajectory they were on before the war started.

## Cheap, leaderless, and quietly growing

_Britain has the fastest growth in the G7, inflation below the Bank's own forecast, the cheapest major equity market in the world, and, as of last week, no Prime Minister. Only one of those is a real problem._

**+0.6%** — UK GDP in the first quarter, the fastest in the G7

I have spent a career being told that to be optimistic about the British economy is to misunderstand it. Permit me a moment of vindication and then a genuine warning.

What is happening: the economy grew 0.6 per cent in the first quarter, the fastest in the G7, led by services. Inflation came in at 2.8 per cent in April and again in May, a full half-point below the Bank of England's own projection, the same Bank whose chief economist spent the spring warning about the second-round effects of an energy shock that, as we have established, did not have any. Private-sector wage settlements have fallen to 2.9 per cent, the lowest in five years; vacancies are falling, as is the number of people in work. Despite these pretty obvious signals that so-called second-round effects were not kicking in, the MPC predictably held rates at 3.75 per cent, and the market is still clinging to the idea that rates will increase over the second half of the year by 0.25 per cent.

**10,000** — The FTSE 100's first-ever close, 5 January 2026

In my view the consensus is wrong and the next move in rates is down, and possibly before the year end. I also expect inflation and rates to continue to fall in 2027. The former will fall possibly below the 2 per cent target by about the middle of the year, and base rates may well dip below 3 per cent. This will, I think, be the biggest economic surprise of 2027, and should in turn lead to better growth outcomes too.

![Britain, marked at half-time.](https://r4at4qm6kmohrtvq.public.blob.vercel-storage.com/charts/0f45577e-b044-4caf-b294-a0225926287c-5f7605582b8a-light.png)

Now the warning, and the politics, which collided last week. On 22 June, Keir Starmer resigned as Prime Minister, brought down by his own party, with a leadership contest, expected to install Andy Burnham, that will not be resolved until September. Rachel Reeves is unlikely to stay in No 11, but the fiscal rules, iron-clad apparently, will remain in place, which will clearly act as a significant constraint on the new Prime Minister’s political ambitions. Encouragingly, he has recruited some sensible, some might argue highly regarded, people to advise him on economic policy, including Andy Haldane, the wise ex-chief economist at the Bank of England. Let’s see what they come up with, but some of the rumours sound quite interesting: getting rid of stamp duty, which I would welcome, and abandoning the triple lock.

Underneath the politics, the slow puncture continues. The UK equity market is the cheapest in the developed world. The All-Share trades on around 12 times earnings, a 15.6 per cent return on equity, and a 3.6 per cent yield, implying an implied return north of 8 per cent for anyone willing to look. The world has noticed in the worst possible way. Foreign buyers are taking whole companies private at the fastest rate in a decade, Schroders Schroders plc (SDR), Beazley Beazley plc (BEZ), Intertek Intertek Group plc (ITRK) and Tate & Lyle Tate & Lyle Public Limited Company (TATE) among them this year, to be joined potentially by EasyJet and SEGRO, both of whom are being courted by US-based investors. The listed universe has roughly halved in twenty-five years. This is a market being quietly dismantled and sold for parts, and the policy response has been to wonder, occasionally and aloud, why nobody wants to list in London. _(Neil in the margin: Return on equity measures profit as a percentage of shareholders' capital — how hard the firm works what it owns. A figure near 16 paired with 12 times earnings is the crux of his argument: decent profitability going cheap.)_

**10,000** — The cheapest major market in the world, and they are selling it for parts.

The investment conclusion has not changed, only sharpened. The UK equity market is quite clearly valued at a significant discount to its peers, and most obviously in its domestically exposed sectors. Whilst UK-based and international portfolio investors continue to ignore or walk past this obvious opportunity, foreign corporates and private-equity investors will continue to exploit it with what appears to be increasing frequency.

_Companies are named here as worked examples of a market-structure argument, the UK's valuation discount and the takeovers it invites, not as recommendations. Nothing here is advice to buy or sell any investment._

## Old problems, new money, and a melt-up in the East

_Germany has finally found its chequebook, China is running two economies at once, and the most violent bull market on the planet just had its first heart attack._

**€500bn** — Germany's new infrastructure fund, and the end of the debt brake

Europe's misfortune is that its best news this year is fiscal, not monetary. Germany, under Friedrich Merz, has done the thing it swore for a generation it never would: reformed the debt brake, established a €500bn infrastructure fund and the largest defence budget in its history. This is genuinely significant, and in time will be seen as the most powerful growth impulse on the continent. For now it is swamped by a stagnant euro-area economy growing by a tenth of a per cent a quarter. France is paralysed under yet another minority government, and Europe’s central bank, as we have seen, has tightened into the gloom. Euro-area equities have had a good half in spite of all that. The Stoxx 600 is up a little over 7 per cent, with some of its best-performing constituents not surprisingly being AI-related stocks like ASML, STMicro and Infineon for example. _(Neil in the margin: The Schuldenbremse is Germany's constitutional cap on structural borrowing, in force since 2009. Loosening it required a two-thirds parliamentary majority, which is why Merz doing it counts as breaking a generational taboo rather than routine budgeting.)_

China is now a two-speed economy in the plainest sense. In the first quarter the economy apparently grew by 5 per cent; industrial production was firm; but retail sales actually fell year-on-year in May, as did fixed-asset investment. Property investment was also down sixteen per cent. Since these data were released, Beijing has quietly cut its growth target this year to between 4.5 and 5 per cent, the lowest since the early 1990s. There were also important geopolitical developments in the first half of the year. The first face-to-face Xi-Trump summit happened in May, in Beijing, and produced deals on soybeans, Boeings and some warm words, but there was no extension of the tariff truce that expires on 10 November. The second summit between the two leaders is set for the 24th of September, and I suspect it is this meeting, not the data, that matters most for the world's second-largest economy.

![Figure](https://cdn.sanity.io/images/v3acfbvo/production/6c88b240bc0f47cb75b13f13dcc3fa5b36e3af55-1800x1800.jpg?w=1600&fit=max&auto=format)

Then there is the East Asian melt-up, which earns a paragraph of respect and a paragraph of caution. The respect: driven by the artificial-intelligence hardware chain, the Nikkei is up 40 per cent and made an all-time high last week, Taiwan is up more than 60 per cent, and Korea, amazingly, has doubled, up around 90 per cent on the year even after the setback on Monday this week. The caution is the events of Monday. The Kospi fell almost ten per cent in a single session, a Black Tuesday in all but name. It’s not at all clear whether this is the pause that refreshes or the start of a more meaningful correction. My sense is that the underlying fundamentals that have driven enthusiasm for stocks like SK Hynix and Samsung have not gone away, but equally, the market’s relentless rise was looking overdue for a moment of pause.

![The half in one chart: dispersion.](https://r4at4qm6kmohrtvq.public.blob.vercel-storage.com/charts/9e344e5d-df88-4c8f-ac21-f0f6d434416b-f3c6d95b04aa-light.png)

## The frenzy rang its own bell

_Gold round-tripped, the small-caps woke up, and the largest IPO in history priced a loss-making rocket company into the record books._

**$1.77trn** — Bells do not usually ring at the top. This one may have done.

Three things in the markets this half mattered, arguably, as much as the index levels.

First, gold made an all-time high of $5,595 an ounce in late January, into the teeth of the pre-war scare, pretty much exactly what it was supposed to as a safe-haven asset, but for obvious reasons, it has since fallen by more than a quarter. Possibly not the sort of insurance that buyers of the precious metal might have been expecting.

Second, the rotation I have gone on about for two years finally arrived. The Russell 2000 made its first-ever close above 3,000 and is up around a fifth on the year, comfortably ahead of a mega-cap index that, for long stretches of the spring, went nowhere. The market is broadening out beyond the handful of names that carried 2024 and 2025, which is healthy, and which is what you would expect as the AI trade matures from a story about a dozen or so stocks into a story about the picks and shovels: the power, the grid, the memory, the silicon.

Third, SpaceX Space Exploration Technologies Corp. Class A (SPCX) came to market on the 11th of June at a $1.77 trillion valuation, the largest flotation in history, for a business that does not make money, and briefly traded, by way of a short-squeeze in the crypto futures written on it, at an implied three trillion dollars. On the one hand this extraordinary short-term trading history might sound alarm bells, for good reason, but the truth is that none of this changes the underlying thesis.

The AI build-out is real, it is contracted years out, and the money has to travel through bricks and mortar, power and silicon long before it reaches a chatbot. That is a good reason to be more careful about where one stands in this queue, and it explains, in part, why I have always favoured the picks-and-shovels approach to participation in this remarkable industrial revolution.

## Peering into the third quarter

_The shock has passed. The policy mistake has not. The second half is the story of the unwind._

The first half of 2026 saw a war, a record oil price, and a partly synchronised misread of the crisis and the scale of the energy price shock that it created, by the people who set the price of money across the developed economies of the world. The second half, I think, will be the story of that mistake being corrected, albeit slowly. Energy-price-driven inflation will dissipate, and I expect headline inflation in the United States and in the United Kingdom to peak soon and then decline. The interest-rate increases anticipated on both sides of the Atlantic will, in my view, not arrive, and the Bank of England may even cut before the year end. Bond yields will follow suit, falling across the curve. The UK equity market stays cheap, but should keep delivering, and will likely see more bids and M&A. Excitement for all things AI-related will remain, albeit slightly tempered.

The risks to this relatively benign outlook are for the most part geopolitical. The Islamabad ceasefire is a sixty-day framework, not a treaty, and its clock runs out in the middle of August, though clearly that timetable can be extended. The American tariff surcharge expires on the 24th of July. The second Xi-Trump summit falls on the 24th of September, with the tariff truce behind it due to expire on 10 November. Britain will have a new Prime Minister, but my guess is that the fiscal constraints that boxed in his predecessor will be precisely the same for the new boss and his team. Having said that, it appears that Mr Burnham has recruited some sensible people to advise him, and so there may be some interesting developments that just might help the economy. I live in hope.

The consensus spent most of the first half of 2026 being very frightened by the war and the energy price shock that the consensus miscalculated. My job in the second half is the same as it was in the first: to keep watching what is actually happening rather than what we are told to worry about. To look at the data and where they can be established, the facts, and to challenge everything. I expect to be very busy.

I will report back in the autumn.

Neil

_This letter is commentary and perspective, not investment advice or a recommendation to buy or sell any investment. Holdings and companies are discussed only as worked examples of the reasoning. Past performance is not a guide to the future._