Inflation and Interest Rates: Separating Signal from Noise

I wanted to have an argument with myself.

I had developed a fairly convincing explanation for why central banks might be making a difficult situation worse. Convincing to me, anyway. Which seemed a good reason to challenge it.

My starting point was inflation. If conflict disrupts oil supplies, shipping becomes more expensive and harvests suffer, prices rise. Increasing interest rates cannot reopen a shipping route, repair a refinery or make it rain.

Meanwhile, households face expensive essentials and higher borrowing costs. Further increases might reduce spending, but how much of that spending can people comfortably give up?

I wanted to understand whether raising rates in those circumstances was sensible economic management or an attempt to solve a problem with the only tool readily to hand.

The evidence complicated my argument.

The first complication was that borrowing costs were already rising.

UK mortgage offers were becoming more expensive without an increase in the Bank of England’s base rate. Moneyfacts figures showed average two-year fixed rates rising from 5.59% to 5.63% in the first week of September 2026, with five-year rates increasing too. Mortgage-rate reporting

That initially felt like confirmation. The market was already applying pressure. Why would the central bank need to add more?

But mortgage pricing reflects expectations about future rates as well as lenders’ funding costs, risk and competition. Today’s base rate is only part of the picture. Fear of future inflation can make borrowing more expensive before a central bank takes any action.

The signal was useful: unchanged policy rates do not mean unchanged financial conditions. The conclusion still required judgement: was the pressure already sufficient, or would inflation persist without more?

Then I went looking for distress.

If households and businesses were struggling, surely rising defaults would demonstrate that demand was already being squeezed.

There was evidence supporting that concern. The Bank of England’s second-quarter credit survey reported increasing defaults on credit cards and other unsecured household borrowing. Yet UK Finance reported a slight fall in the number of homeowner mortgages with substantial arrears. These represented 0.89% of homeowner mortgages in the second quarter. Bank of England survey; UK Finance arrears data

Across the Atlantic, Fitch’s monitored sample of private corporate borrowers recorded a 9.2% default rate in 2025, up from 8.1% in 2024. That was concerning, but it was a particular sample of businesses, and the definition included distressed debt exchanges. It was not a measure of household defaults, or of every loan in private credit. Fitch findings reported by Reuters

I had to resist assembling different measures into one apparently definitive story.

A missed household payment, a mortgage in arrears and a company restructuring its debts all tell us something. They tell us different things, over different periods, about different people.

I also wondered whether private lenders were increasing interest charges to compensate for defaults. One index offered an awkward finding: estimated yields on performing private-credit loans rose between the first and second quarters of 2026, while the spread above the underlying benchmark narrowed. Benchmark movements and changes in the loans included helped explain the result. Higher yields alone did not prove lenders were charging more for credit risk. Houlihan Lokey’s private-credit index

This is where separating signal from noise becomes uncomfortable. Some of the noise is the story we would like the evidence to tell.

Food prices brought the argument closer to everyday life.

The official annual food-inflation figure for July was 1.3%. Read in isolation, that might suggest the pressure had largely passed. But it measured the increase over the previous year. It said little about whether people had recovered from the increases before that. ONS inflation release

The Food Foundation, using ONS data, reported that food prices were 30.1% higher in July 2026 than in April 2022. Slower inflation leaves that accumulated increase in place. Food Foundation tracker

More revealing was research from the Institute for Fiscal Studies. Between September 2021 and September 2023, grocery products initially among the cheapest tenth within their categories rose in price by 36.2%. The most expensive tenth rose by 15.8%.

Households in the bottom quarter of the spending distribution experienced grocery inflation 5.6 percentage points higher than those in the top quarter. The versions of products people bought mattered, as well as the categories they spent money on. IFS research

These are historical increases, not today’s annual inflation rate. But they demonstrate why the experience at the checkout can differ from an aggregate figure without either the shopper or the statistic being wrong.

Someone buying premium products may have room to switch. Someone already buying the cheapest acceptable food has fewer options. They can spend more, buy less or compromise on what they eat.

A household keeping its grocery bill steady might be managing successfully. It might also be eating worse.

Recent commercial evidence contained both reassurance and concern. Worldpanel reported grocery inflation slowing to 2.1% in the four weeks ending 9 August. Its survey found 39% of households feeling financially comfortable, alongside 20% struggling. Worldpanel report

An accurate average can still be an inadequate description of hardship.

That changed the question I was asking about interest rates.

I began by asking how a rate rise could destroy demand that was not there.

But weak demand is still demand. People still buy food, heat homes and service debts. Higher borrowing costs can reduce spending further, even when spending is already constrained.

And if the economy’s capacity to supply goods falls, spending does not need to be booming to exceed what can readily be supplied.

That is the difficult counterargument to my original position. Rates can influence inflation even when they cannot repair its initial cause. The question is how much benefit further restraint would deliver, how long it would take, and who would bear the cost.

The hardship evidence does not prove rates are ineffective. It makes the consequences of using them harder to dismiss.

I also had to reconsider my assumption that central banks could only restore stability by buying debt. They can lend temporarily against collateral, and they can distinguish support for market functioning from their broader interest-rate decisions. The Bank of England’s temporary gilt purchases in 2022 illustrated that separation. Bank of England announcement

None of these interventions makes genuine losses disappear. Each involves choices about risk and responsibility.

I ended with less certainty than I started with, but a better question: what evidence would justify imposing additional financial pressure, given the adjustment households and businesses have already made?

That seems more useful than declaring either that central banks are powerless or that another rate rise is obviously necessary.

Decision-making is hard. Data arrive late, describe different populations and sometimes point in different directions. Decisions have to be made before the picture is complete. Certainty can be an illusion created by the confidence of an explanation.

We should question institutions. Ask what their measures capture, whose circumstances disappear inside the average, which assumptions connect the evidence to the decision, and what would cause them to change their minds.

We can do that without turning every uncertainty into a crisis of confidence.

For all the institutional verbosity, the work is done by people, many of them capable, conscientious people trying to build the best understanding they can from incomplete evidence. Good intentions do not remove the need for scrutiny. Expertise does not remove the possibility of error.

Confidence worth having allows for both.

I wanted to have an argument with myself. I came away thinking that asking better questions, while accepting that the answers may remain uncertain, is part of how we help institutions make better decisions.