The art of predicting Bank of England’s main policy rate

Connecting business research with policy, practice and public debate Connecting business research with policy, practice and public debate Ahead of the Bank of England’s latest Monetary Policy ReportCostas Milasshows why UK interest rates peaking at 4.25 per cent in mid-2027 looks like a reasonable bet. But economic shocks, which now come more frequently and bigger in size than before, could derail even the best economic predictions. And some of those shocks might come courtesy of President Donald Trump. Greek philosopher and mathematician Pythagoras is often quoted as claiming that “number rules the universe”. For central bank policymakers, and academic economists with an interest in monetary policy, this number refers to the so-called “neutral rate”. This is the interest rate prevailing when output is at “equilibrium” and inflation is “stable”. We are keen on the concept of the neutral interest rate because it provides a straightforward monetary policy signal. If, for instance, the Bank of England’s (BoE’s) main policy interest rate (Bank Rate) is above the neutral rate, Monetary Policy Committee (MPC) members could infer that monetary policy is restrictive, therefore suppressing output but also inflation. If, on the other hand, the Bank Rate is below the neutral rate, then monetary policy is (too) loose or stimulative, therefore adding to inflation pressures and an overheating economy. (Some economists also use the terminologyR-starfor neutral rates which inspiredToby Nanglefrom FT Alphaville to produce these creative Star Warsgraphicsto illustrate the concept.) Nevertheless, “equilibrium” output cannot be measured accurately. At the same time, a 2 per cent inflation rate prevailing over (say) two years is mathematically as “stable” as (say) a 4 per cent inflation rate prevailing over the same period. These two unknowns, equilibrium output and stable inflation, that is, question why central bank staff and academics pay so much attention to the neutral rate. This is exactly why I sympathise with Federal Reserve Chair Kevin Warsh. When Mr Warsh was asked, in September 2026, on whether US interest rates were above or below the neutral rate, he replied that the neutral rate was an academic concept thatdid not enter into the Fed’s practical deliberations. Quite so! To get around the problem of calculating a single value for the neutral rate, one could estimate a range (or band) of neutral rate values. The BoE’s interest-rate setters believe thatthe neutral rateis between 2 per cent and 4 per cent. I estimate a monetary policy rule (or interest rate reaction function) in which the BoE’s main policy rate (Bank Rate) responds to (i) lagged Bank Rate (this captures inertia in interest rate decisions), (ii) deviations of expected inflation from the Bank’s 2 per cent inflation target, and (iii) a measure of the output gap. For expected inflation, I usetwo measures, the BoE’s five-quarter ahead and six-quarter ahead inflation expectations. This allows me to compare two monetary policy rules depending on whether I use five or six-quarter ahead inflation in my model. Bank Rate has to respond to inflation expectations. Indeed, monetary policy cannot wait too long to act because, as Milton Friedman noted,it affects the economy with long and variable lags. For the output gap, I use the Office for Budget Responsibility (OBR)measure, which monitors GDP deviations from “equilibrium” GDP. When inflation expectations are above the inflation target, MPC members raise interest rates. When output exceeds equilibrium (in which case the economy is overheating) interest rates are again on the rise. I estimate the monetary policy rule recursively, that is, by adding one data point at a time. The attraction of doing so is that I can derive, for every point in time, a band of neutral rates which changes according to shifting economic conditions. Figure 1 plots my estimates of the neutral band together with Bank Rate over the 2016-2026 period using quarterly data. Figure 1: Neutral interest rates and Bank Rate, 2016Q1-2026Q4 Between 2016 and 2021 the Bank Rate has been below the bound of neutral rates which suggests loose monetary policy over those years. Then Bank Rate enters the neutral band but only briefly, a period where monetary policy is neither loose nor restrictive. From mid-2022 onwards, and asinflationrose considerably (mainly triggered by the war in Ukraine and then the war in Iran), monetary policy in the UK tightens considerably. In late 2026, the Bank Rate and the upper bound of neutral rates are really close to each other. Therefore, Bank Rate, at 4 per cent, is close to being restrictive. According to thelatest forecastsmade by the OBR (in late October 2026), Bank Rate will reach 4.25 per cent by mid-2027 and close to 4.4 per cent by 2028Q1 (see Figure 2) under the assumption of CPI inflationpeakingat 3.8 per cent in early 2027. My monetary policy rules broadly agree with the assessment made by the OBR. In my worst-case scenario, Bank Rate couldinsteadrise to 4.6 per cent in mid-2027, but only if UK inflation remained as high as 4.2 per cent in early 2028. Such a development would, of course, indicateinflation stubbornnessdue to additional inflation shocks and/or strong “second-round” effects from wage pressures. Figure 2: Bank Rate central forecast. OBR, 2026Q4-2028Q1 There are other things to consider such as the impact of US monetary policy. Indeed, monetary policy decisions in America are watched closely by the BoE.Academic researchfinds that the foreign spillovers of higher US interest rates are notable. An interest rate tightening of 100 basis points (one percentage point) by the Fed reduces GDP in advanced economies (including the UK) by 0.5 per cent within a three-year period. The implication for UK monetary policy is that BoE’s MPC members might not necessarily rush towards following Fed’s lead in increasing UK interest rates since part of the intended tightening in the UK will be delivered through spillover effects from the US. All these predictions could turn out to be wrong. What we have learned since the mid-2000s is that economic shocks not onlyhappen more frequentlythan before but also come very big in size. See the global financial crisis, Brexit, COVID-19 pandemic, the Russian invasion of Ukraine and the war in Iran. Considering the unpredictability of President Donald Trump’s response to his party’s disastrous mid-term elections and a very slim immediate prospect of re-opening the Strait of Hormuz, even the most sophisticated (human or AI-generated) model would surely fail to accurately predict UK interest rates. This article gives the views of the author, not the position of LSE Business Review or the London School of Economics. You are agreeing with our comment policy when you leave a comment. Image credit:Jono Photographyprovided by Shutterstock. Costas Milas is Professor of Finance at the University of Liverpool. This site uses Akismet to reduce spam.Learn how your comment data is processed.