Unlock the Editor’s Digest for free
Roula Khalaf, Editor of the FT, selects her favourite stories in this weekly newsletter.
The UK is just a few days from a new fiscal event, and Rachel Reeves is hemmed in by her past decisions.
Having pledged to get Britain down to one budget per year, the Chancellor maintained the Office for Budget Responsibility’s requirement to produce fiscal forecasts twice a year, meaning she has to now respond to updated forecasts without the luxury of tax tweaks.
So this time around, sating the fiscal rules gods means relieving a lot of people of their benefits. Everyone is entitled to their own views on whether that is good or bad thing, but hopefully we can agree on this: it’s not a great thing to be in effect forced to do so by made-up rules.
But this is all big-picture political stuff, which is why here at FT Alphaville we’ve decided to deliberately miss the wood and stare intensely at the branches of one particular tree.
When we last wrote about the OBR’s assumptions for the Bank of England’s quantitative tightening, we were pretty bamboozled. To remember why, let’s go back to basics.
Back to basics
As part of their twice-yearly forecast suite, the budget watchdog’s analysts estimate future losses from the BoE’s Asset Purchase Facility. At present, the BoE is whittling the APF pile down in two ways: through the passive roll-off of maturing bonds (“passive QT”), and the active selling of others (“active QT”).
The OBR’s predictions matter a lot, because the UK’s fixation on maintaining “fiscal headroom” — which is currently extremely tight — means every billion counts when Budget day comes.
In the first two years of QT (2022—23 and 2023—24), the Monetary Policy Committee aimed for a total reduction of £100bn per year, consisting of the passive roll-off plus whatever active sales level was needed to hit that 💯.
This worked out as £42bn of active sales in the first year, and £54bn for the second. At its March forecast last year, the OBR decided the best way to extrapolate these numbers forward was to look at the active sales levels and average them. And so it did, duly forecasting that future active sales would be £48bn ((42+54)/2) per year throughout the five-year forecast window.
Notably, this resulted in a striking figure for the current year (2024—25), during which about £87bn of passive is expected. 85+48=£135bn, which was a lot of QT to predict, especially following two fixed £100bn years.
Within a few short months, the OBR’s prediction had been proven to be completely wrong: at its August meeting, the MPC backed another £100bn envelope, implicitly consisting of £87bn of passive and £13bn of active.
As we noted last autumn, this presented a learning opportunity for the OBR. Heading into October’s budget, analysts speculated on four possible ways the OBR might predict future active sales:
— An average of the first three years of planned active sales ((42+54+13)/3=£36.3bn per year through the forecasting window).
— Extrapolating the latest-year active sales figure to project £13bn per year through the forecasting window.
— Extrapolating the BoE’s previous decision to predict a £100bn envelope, in which active sales top up passive to hit that level.
— Following the path expected by markets, as they do with things like Bank Rate.
The difference between these was highly material in the fiscal context. There was a projected spread of £15.5bn between the most gentle of these (£13bn active forward) and the harshest (£100bn envelope) in terms of how they would affect Reeves’ headroom.
Absurdly, the OBR didn’t do any of these, choosing instead to ignore the third-year envelope and simply maintain their £48bn-per-year prediction, based on the first two years of active sales. As we wrote then:
The effects this will have had on Reeves’ headroom aren’t as extreme as if the OBR had presumed a continuous envelope of £100bn p/a — in effect, nothing has changed except the current year — but it’s definitely at the tougher end of things compared with some of the possible alternatives.
Back to the near future
Of the five apparent ways the OBR could have made up this number, they chose the second-most-harsh in terms of how it cuts into Reeves’ fiscal headroom.
Now, first, let’s get four things straight:
1) The OBR is not to blame for the sorry situation in which their QT predictions could contribute, in some way, to things like welfare cuts.
2) The OBR should not be pressured into choosing the model that is most favourable to the Chancellor of the day.
3) Other factors (like the expected path of Bank Rate) may well matter as much as the overall scale of active sales.
4) A lot of this may prove to be academic if the BoE finds it has hit a minimum viable level of reserves and stops QT early for stability reasons. Based on its current trajectory, it will reach the upper band of its Preferred Minimum Range of Reserves by the end of the year.
Writing for FTAV ahead of that fiscal event, T Rowe Price’s Tomasz Wieladek advocated for the OBR tracking the market-predicted path, as indicated by the BoE’s Markets Participants Survey (MaPS).
It’s a decent moment to revisit MaPS. According to the latest survey, released last month, the median prediction is for there to be almost zero active QT by 2028—29:
Even the 75-percentile estimate is less than half the OBR’s current assumption. Meanwhile the 25th percentile assumption is that active QT is over by the ends of the next fiscal year. Which to trust, the aggregate view of about 70 firms who have a vested interest in getting this right, or an arbitrary average?
Let us say without ambiguity: how the OBR predicted active sales in October was an indefensible fiscal Numberwang, and it ought to change. The current framework doesn’t help anyone, and the OBR loses credibility by using it.
We asked the OBR if it will be revisiting its QT assumptions as part of the upcoming economics and fiscal outlook. A spokesperson told us:
I’m afraid we can’t provide any guidance on what we might consider for the forthcoming forecast but will of course explain any necessary changes fully in the EFO on 26 March.
Roll on Wednesday!
