The Monetary Policy Committee held Bank Rate at 3.75% on 30 July, and almost nothing about the way it did so was straightforward. The vote was six to three, with Megan Greene, Catherine Mann and Huw Pill all pressing for an immediate quarter point rise to 4%.1 That is the largest bloc to vote for a rise at any meeting this year, and one dissenter more than the seven to two that economists polled by Reuters had expected. Yet the two year gilt had its biggest one day fall in over two months, and traders cut the tightening they expect this year from 38 basis points to 29.2

The reason for that apparent contradiction was Andrew Bailey, who used the press conference to tell the room in plain terms that nobody should leave it thinking the Bank was "edging towards a hike".3 A committee drifting towards tightening on paper, a governor pushing back against that reading in person, and a market that believed the governor. Underneath all three sits a forecast that quietly makes the strongest argument for patience of the lot.

The decision at a glance: Bank Rate 3.75% (unchanged) · Vote six to three, three for a rise to 4% · CPI inflation 2.6% in June · Central projection peaks at 3.2% in 2026 Q4, ends the forecast at 1.9% · Conditioning path carries Bank Rate to 4.2% · Two year gilt 4.34%, down more than 10 basis points · Next decision 17 September

A committee migrating in one direction

The story in the vote is the trajectory, not the split. In February, four of the nine wanted to cut to 3.5% and Bailey's majority was a bare five.4 The conflict broke out later that month, and by March the Committee was unanimous. In April one member dissented for a rise, in June two, now three. Over five meetings the balance has swung from a net four votes for easing to a net three for tightening, without Bank Rate moving a basis point.

Net balance of votes for tighter policy on the Monetary Policy Committee, February to July 2026. Bars below the line are votes for a cut, above it votes for a rise. March has no bar because the Committee was unanimous. Bank Rate has been unchanged at 3.75% throughout. Source: Bank of England Monetary Policy Summaries.

That drift is a response to a single variable. Brent crude settled near $90 a barrel on the day of the decision, roughly a quarter higher than a month earlier, after a fresh round of American strikes on Iran undid much of the relief that had followed the June memorandum of understanding.5 Energy is the shock, and the Committee has said repeatedly that monetary policy cannot influence energy prices. What it can influence is whether a one off rise in the price level turns into a persistently higher rate of inflation.

The transmission channel that has not fired

The dissenters are not arguing that the Bank should fight the oil price. They are arguing that the longer energy stays elevated, the greater the chance firms and workers build it into prices and wages. That is the second round effect, and the whole disagreement reduces to whether it is happening.

So far it is not. June CPI came in at 2.6%, below the 2.7% consensus. Core was flat at 2.6%, services eased from 3.7% to 3.6%, and goods slowed from 2.0% to 1.7%.6 Services is the measure that matters, being the part of the basket most closely tied to domestic labour costs, and it is moving the right way.

UK CPI inflation, annual rate, January to June 2026. The axis begins at the 2% target, so the whole plotted range sits above it. Headline inflation has fallen through the energy shock rather than risen with it. Source: Office for National Statistics.

The labour market is the reason. Unemployment stood at 4.9% in the three months to May, regular pay growth slowed to 3.4%, its joint weakest since October 2020, and vacancies fell to 712,000. Split the pay figure and it is starker: public sector earnings grew 5.5% against the private sector's 2.9%.7 That 2.9% flatters the position somewhat, since the Bank reckons compositional shift towards lower paid industries holds it down by around half a point. Even adjusted, though, this is not a labour market with the power to turn an oil shock into a wage price spiral, and momentum on a three month annualised basis remains subdued. Mann herself, one of the three, judged in June that private sector wage growth was near target consistent, which makes her dissent an argument about risk rather than evidence.8

Firms are behaving accordingly. Only 4% of respondents to the Bank's Decision Maker Panel expect the conflict to widen their margins, while roughly two thirds expect it to squeeze them. That is an economy absorbing a cost shock, not passing it on.9

The dissent is an argument about risk, not about evidence.

Growth offers no support either, and the published figures flatter the position. Three month growth to May looks respectable at 0.7%.10 The Bank's own underlying measure puts it at 0.1% in the second quarter against potential supply growth of 0.3% to 0.4%, heading to roughly zero in the third, with the output gap near 1% of potential GDP. Meanwhile financial conditions have tightened without the MPC lifting a finger: averaged over the next three years the forward curve sits around 77 basis points above its pre conflict level, and quoted two year mortgage rates have risen 79 basis points since February. The market has already delivered part of the tightening the minority wants, which weakens rather than strengthens the case for delivering the rest.

Households feel the squeeze from both directions. The Ofgem cap rose to £1,663 in the third quarter from £1,477 in the second, and the temporary removal of VAT from electricity bills in October, which runs for six months, holds the fourth quarter cap to roughly £1,680. On the borrowing side, a typical owner occupier rolling off a fixed rate over the next two years is projected to pay around £45 a month more than had been expected before the conflict.

The statements members gave to explain their votes point the same way. Bailey saw early signs that inherited inflation pressure was weaker than assumed and Breeden judged disinflation solidly on course, while Ramsden, for all the risks he listed, said he would consider restarting rate cuts if they faded. That is not the vocabulary of a committee preparing to tighten.11

What the July Report actually says

Here is the part of Thursday that markets read correctly and most of the commentary missed. The Bank's projections are conditioned on the market implied path for Bank Rate, a curve that rises to around 4.2% by the second half of 2027 and stays broadly flat thereafter. Run the economy forward on that assumption and the central projection has inflation peaking at 3.2% in 2026 Q4, easing to 2.6% by 2027 Q3, and then falling through the target to 1.8% in 2028 Q3 before settling at 1.9%.9

Central projection for CPI inflation, third quarter of each year, conditioned on a market curve that carries Bank Rate to around 4.2%. Inflation peaks at 3.2% in 2026 Q4, between the first two points shown. Source: Bank of England, July 2026 Monetary Policy Report, Table 3.B.

An undershoot is about as conventional a signal as a central bank sends. It says the conditioning path is tighter than the one required to hit the target. On this occasion the Report does not leave it to inference. Its own illustrative policy rules, which balance returning inflation to target against volatility in output, track the market curve for the coming year and then fall below it, pointing to a looser overall stance than the market implies. Under that looser stance, the Report notes, inflation would sit at target rather than slightly beneath it.

The second half of the argument is in the annex on financial conditions. A term structure model estimated by Bank staff finds that the expected path of Bank Rate over the coming year is broadly flat, against 50 basis points of cuts priced before the conflict. The July Market Participants Survey says much the same, with the median respondent expecting no change until June next year and declines after that. Which leaves a question: if nobody expects rises, why does the curve slope up? The Bank's answer is that the slope is driven primarily by risk premia, not by expected rate rises. Nor is that a one month judgement. Staff models pointed the same way in June, when respondents to that month's survey put most of the gap between their own expectations and the curve down to asymmetric risk and compensation for uncertainty.8

Bailey said the same thing in plainer language. Asked about the rising path priced by markets, he called it reasonable but cast it as compensation for the risk that the energy shock worsens, "rather than a central expectation", and added that a credible resolution of the conflict would let the Bank run policy looser than the curve implies.11

Why conditioning matters. The Bank does not forecast what it will do. It forecasts what would happen if Bank Rate followed the path markets have priced. A projection landing above 2% implies markets expect too little tightening. One landing below, as this one does, implies they expect too much. It is an indirect way of arguing with the curve.

The qualification is real: the Committee judged the risks around that projection tilted to the upside, and conditioned the exercise on the fifteen day average of energy prices to 20 July.1 The Report is built around scenarios rather than a fan chart, and their dispersion is wide enough to justify almost any degree of caution.

ScenarioPeak inflationCPI, 2027 Q3CPI, 2029 Q3Energy assumption
Milder3.0%, in 2026 Q42.4%1.7%Oil and gas roughly 3% and 6% below central
Central3.2%, in 2026 Q42.6%1.9%Oil declining from $76 to around $71
Adverse4.5%, in 2027 Q24.1%2.4%Oil 30% and gas 60% above central

The adverse case is the one that keeps the three dissenters awake. It peaks at 4.5% in 2027 Q2, is still at 2.4% three years out, and carries private sector wage growth to 4.4%. On that path, the Report accepts, policy would need to be materially tighter than the market curve. The scenarios do not simply argue against the minority. They set out the one world in which it is right.

And that world is closer than the conditioning implies. Brent averaged $78 across the fifteen day window and the central case has it easing towards $71. It was near $90 when the decision was announced. The window closed before the latest escalation, so the central projection was stale in the direction the minority fears on the day it was published.

Where rates go from here

Start with the disagreement, because it is unusually large. Even after Thursday's repricing, markets still fully price a rise by December.2 The curve those prices imply carries Bank Rate to 4.2%. The Bank's own illustrative policy rules sit below that curve. And the median respondent to its July survey of market participants expects no change at all until June next year, then cuts.9 Three different answers to the same question, from people reading the same data, and the two that come from inside the Bank both point lower than the price.

The market path is not a central expectation. It is a central expectation plus insurance against a war.

The base case is a hold on 17 September and probably through the rest of the year. The channel from energy to wages is not operating, the gilt market has already tightened conditions, and the Q4 peak is mechanically energy driven, so it drops out of the annual comparison during 2027 whether or not the MPC does anything.

Rob Wood of Pantheon Macroeconomics reads the majority as content to wait for hard evidence, provided the market keeps a tightening path priced in.11 That is the awkwardness of Bailey's position. The priced path does part of his work, so he needs the curve steep enough to restrain demand without it being read as a promise. Talk it down too well and he has to deliver the restriction himself.

September carries weight for a reason unrelated to Bank Rate. It is when the Committee holds its annual vote on the pace of quantitative tightening, and it has not yet set the target from October. The asset purchase facility will have fallen from a peak of £895 billion to £488 billion by then, and staff put the programme's contribution at 20 to 30 basis points of the roughly 200 basis point rise in gilt term premia since 2022. Investors have already settled on what comes next. The Bank's own survey, published the morning after the decision, has them expecting the pace to slow from £70 billion to £50 billion over the year to September 2027, unchanged from what they said in June.12 Slowing sales into a market absorbing heavy issuance eases financial conditions without touching the policy rate, and given how uncomfortable the long end has become, that is the more likely lever.

The case for an immediate rise succeeds only if several things go wrong together. Services inflation, which the Bank expects to tick up to 3.8% in October, would need to keep climbing rather than level off. Firms' price expectations, up to 3.9% from 3.5% before the conflict, would need to start overshooting their costs, which so far they have not. Settlements would need to break above the 3.5% the Bank's Agents report for this year. And energy would need to hold near current levels into the autumn.

Timing matters as much as the triggers. Firms typically settle pay by the second quarter, so the wage evidence that would vindicate the minority may not surface until 2027 Q2, long after the Committee has to decide. That is the strongest case for moving early, and it is an argument about lags rather than about the data. If the Committee does move, 5 November is the likely date, when the next forecast round lands.

Two complications sit outside the Bank's control. The first is fiscal. Andy Burnham took office on 20 July and installed John Healey at the Treasury. His early remarks about using any flexibility within the inherited fiscal rules revived concerns about a looser stance, and the repricing held: by the following day the ten year gilt stood at 5.06% and the thirty year at 5.77%, both two month highs, with sterling lower.13 A Budget that loosens would raise yields further, tightening financial conditions and raising medium term inflation risk at once, leaving the MPC to judge which dominates. The second is the Federal Reserve, which held on 29 July over three dissents of its own, with markets pricing roughly a 60% chance of a September increase.14 A Fed that tightens while the Bank sits still means a weaker pound and dearer imports, the channel the MPC least wants opened while energy is already doing damage.

The asymmetry is clear enough. The next move is still more likely up than down, but less likely than the curve implies, and the likeliest path is no move at all this year, followed by cuts through 2027 as the energy base effects unwind and the projected undershoot asserts itself. Ramsden has already set out the conditions under which he would vote to restart them. The market is pricing insurance against the adverse scenario. The Committee, on the evidence of its own forecast, does not think it will be needed.

Footnotes

  1. Bank of England, Bank Rate maintained at 3.75% - July 2026 Monetary Policy Summary and minutes (opens in a new tab), 30th July 2026. ↩ ↩2

  2. Reuters, Traders trim UK rate hike bets after BoE Iran comments (opens in a new tab), 30th July 2026. ↩ ↩2

  3. Bloomberg, Bank of England Leaves Rates Unchanged, Spurs Surge in Two-Year Gilts (opens in a new tab), 30th July 2026. ↩

  4. Bank of England, Bank Rate maintained at 3.75% - February 2026 Monetary Policy Summary and minutes (opens in a new tab), 5th February 2026. ↩

  5. Trading Economics, Brent crude oil (opens in a new tab), accessed 31st July 2026. ↩

  6. Office for National Statistics, Consumer price inflation, UK: June 2026 (opens in a new tab), 22nd July 2026. ↩

  7. Office for National Statistics, Labour market overview, UK: July 2026 (opens in a new tab), 21st July 2026. ↩

  8. Bank of England, Bank Rate maintained at 3.75% - June 2026 Monetary Policy Summary and minutes (opens in a new tab), 18th June 2026. ↩ ↩2

  9. Bank of England, Monetary Policy Report, July 2026 (opens in a new tab), 30th July 2026.Projections and scenarios are from Tables 3.A and 3.B and Section 3.3, energy bills and pass-through from Section 1.1, firm survey evidence from Box B, the market curve and household interest costs from Box F, quantitative tightening from Box G. ↩ ↩2 ↩3

  10. Office for National Statistics, GDP monthly estimate, UK: May 2026 (opens in a new tab), 16th July 2026. ↩

  11. Financial Times, Bank of England holds rates at 3.75% as it waits to see impact of Iran war (opens in a new tab), 30th July 2026. Source for the individual members’ voting statements, Bailey’s press conference remarks and the analyst comment quoted. ↩ ↩2 ↩3

  12. Reuters, UK markets expect £50 billion of QT in year to September 2027, BoE says (opens in a new tab), 31st July 2026. ↩

  13. Reuters, via Euronext, Sterling dips, gilt yields at 2-month high as investors assess Burnham’s spending plans (opens in a new tab), 21st July 2026. ↩

  14. Bloomberg, US 30-Year Yield Soars to Highest Since 2007 After Fed Stands Pat (opens in a new tab), 29th July 2026. ↩