What the gilt move does to the Budget: a pre-registered nowcast of the OBR's headroom number for 28 October 2026

The Chancellor's headroom is the OBR's forecast of the current budget in 2029-30. It was a £23.6bn surplus in March. Since the OBR's market window closed in January, gilt yields have risen by about two thirds of a percentage point across the curve and markets have gone from pricing Bank Rate cuts to pricing hikes. Five sell-side houses put the pre-measures number at £11.6bn to £15bn, all of it attributed to debt interest. Our estimate is £14.6bn before Budget measures, with an 80 per cent interval of about -£4bn to +£34bn. The point is inside the sell-side range. The route is different, and the route is the finding: Debt interest is the predictable part. Across the 33 OBR forecasts since 2010, each percentage point on gilt yields between one market window and the next has moved the third-year debt interest forecast by about £11bn, with an R squared of 0.81 and a walk-forward error of £3.6bn. On today's move that is £7.4bn; the OBR's own ready reckoner, adding the Bank Rate path, gives £9.2bn. We take £8.5bn, not the £12bn in circulation. Receipts give some back. Tax revenue this year is running £2.8bn ahead of the OBR's profile, and the record says the OBR carries in-year receipts news into its third-year forecast at roughly one for one. That is worth about £3bn. Spending drift takes it away again. In 27 of 33 forecasts the OBR has found more non-interest spending in year three than it had before, welfare above all, with a median revision of £3.4bn. This year's welfare outturn is above profile. We take £3.5bn. The interval is wide because the OBR's judgement is. The part of the OBR's third-year revision that no market variable predicts, its view on productivity, migration and the tax gap, has a standard deviation of about £16bn on the record. Anyone quoting a headroom estimate without that width is quoting a point on a distribution they have not looked at. The gilt move is mostly imported. Regressing daily gilt changes on Bunds and Treasuries over the two years to January, then applying the fit to the months since, attributes 74 per cent of the rise in the 10-year to global rates. The UK-specific part was concentrated in March and has been flat since June. Against a floor of £10bn, which the Treasury is reported to defend, £14.6bn leaves a little over £4bn for anything the Budget wants to do. The sell-side median of £13bn leaves £3bn; Bloomberg Economics' £11.6bn leaves under £2bn. The gap between the estimates is the gap between a Budget with something in it and one with nothing.