Britain’s gilty conscience

Base effects are an important part of inflation: a big monthly jump in prices can look like a sharp fall once, twelve months later, it drops out of the year-on-year rate figures. These effects can be hard to visualise, a situation this article won’t really rectify.

You can begin to crudely predict the next year-on-year inflation figure (eg for February) by taking the latest year-on-year figure (ie for January), deducting the month-on-month reading for the earliest month of the 12 preceding that January (ie the previous February), and then adjusting it by however much you think prices changed in your current February. Simples.

Apply that system to February 2024, UK inflation stats for which will drop next week, and you get:

(4pp [Jan 2024 y-o-y rate] – 1.1pp [February 2023 m-o-m rate]) + ???pp [February 2024 m-o-m] = ~~~2.9 per cent

Get the latest-month vibes roughly right, and you too could be a sellside economist.

Looking at year-on-year inflation figures as the sum of 12 batches of monthly components is one of those things that is somewhat visually interesting:

Another way of thinking about it is that every year-on-year price inflation figure starts its life as a single, embryonic month-on-month figure, and spends then spends 11 months developing into its final form.

By crudely adding together month-on-month figures for UK CPI inflation, we can kinda, sorta, ish, see in a forward-looking fashion how base effects work.

(Caveats: the following charts use figures that are all slightly out, which we think is a result of how m-o-m figures are rounded by the ONS but may also be our own incompetence. There’s almost certainly a better way that involves the year-on-year monthly figures, but we’d already committed too much time to doing it incorrectly.)

Using this probably flawed system, here are the figures as they developed (use the filter to select your favourite reading):

Viewed this way, you can hopefully see how big base effects kick in, creating big steps up and (particularly) down in inflation. This is especially acute in the UK, where the price cap system means that energy inflation undergoes sharp quarterly adjustments rather than developing more organically.

As an alternative, here’s how the readings look developing in parallel:

What, if anything, does this teach us? It might prompt sinful thoughts like: “Wow year-on-year inflation sure is meaningless outside of certain contexts such as annual spending decisions!”

Anyway, March’s inflation figures look like they may be roughly on target, while April could be the first of several undershoots.

A Bank of America note published yesterday says:

By a quirk of timing, the April energy price cap reduction is likely to help propel UK inflation to target, but things are not so pretty under the hood, as our Chart of the Day shows. The UK inflation persistence problem hasn’t gone away and will need materially higher real policy rates to deal with it, relative to peers

Said chart:

(If you’re finding that hard to read, it’s:
— Y-axis: % of inflation components
— X-axis: % inflation rate)

BofA analyst Mark Capleton’s argument is that, even as a target undershoot looms, “the UK’s underlying inflation picture is considerably less benign” than that of the eurozone (an argument for shorting gilts).

Part of his argument hinges on a long-standing error in how the Office for National Statistics measured clothing price inflation:

We’ve said before that the UK’s underlying inflation persistence problem long-predates Brexit, although it was aggravated by it. When we point out that the UK “had to wave a bigger stick” than its peers (needing much higher average policy rates, Exhibit 3) before the global financial crisis, in order to keep inflation in check, we are often countered with the fact that the UK did actually report lower average inflation than the US and Eurozone in that period.

Except it wasn’t really like that. The UK reported severe clothing price deflation that didn’t actually happen, as a result of a sampling discontinuity glitch (Exhibit4), and this persisted for a very long time. Despite the relatively small weighting of clothing and footwear, the BoE estimate that this error resulted in overall CPI inflation being understated by 37bp per annum, on average, until it was fixed in 2010.

Relevant charts:

Capleton, cont.:

This isn’t to say that the UK should price materially higher inflation than Eurozone (although it does). Our take would be that the UK has a much tougher fight on its hands, which will require meaningfully higher long term real policy rates, and this isn’t priced.

The important context to this, of course, is the UK’s massive levels of planned gilt issuance, which BofA says creates a key fragility.

Capleton deploys a chart from the Institute for Fiscal Studies, showing how poor the UK’s indebtedness position looks on a Public Sector Net Worth footing (which attempts to show indebtedness in the broader context of a state’s assets and liabilities):

It’s worth caveating that chart, from a chapter in the IFS’s Green Budget last year, a little. As its author Ben Zaranko wrote:

Performance against a public sector net worth target would tell us little or nothing about the government’s ability to access capital markets or service its debt… An increase in the estimated value of an asset the government cannot sell cannot be taken as a signal that the government can afford to borrow more.

Nuances aside, Capleton says:

[After] the Budget, we can say once more that the gilt market call on investors is simply huge – relative to GDP, relative to domestic savings, relative to the current stock of privately held Gilts, and relative to the stock of defined benefit pension liabilities (which are now half their 2020 peak market value).

This debt burden can be expressed in a variety of ways, none of which are very cheerful:

We’ve written plenty about quantitative tightening (and more is coming 🥳), and the impact of Bank of England active gilt sales piling on top of sizeable Debt Management Office issuance.

BofA’s estimate is that the DMO will make £125.4bn of net gilt sales this year, with a further £89.7bn in net supply from the BoE — “equivalent to almost 15% of the currently outstanding privately held stock of Gilts”.

The vicious cycle here is worth re-iterating: issuing debt from the DMO is more expensive because the BoE is selling gilts —

y̶͎̾i̵̝̓e̸͓͆ľ̵̳d̶̬̒s̴͓̍

̴̺̀m̴̗͊o̸̢͐ṿ̵͋e̶̤̓

̴̗̈́í̶͍n̸̲̈́ṿ̴̎ḙ̶̓r̸̟͛s̴̢̓ë̴̠́l̵͇̄ỹ̶̲

̴͚̽t̶̛͚o̸͚͘

̵̼͂p̷̥̓r̷͖̊i̴̙̍c̴̦̔ē̵̝s̷̟͂

— and more debt has to be issued because indemnifying the BoE against the losses on those gilt sales reduces fiscal space. It is a deeply unlovable situation, one that will potentially constrict the UK economy for years, and nobody really seems to know a way out.*

Capleton concludes:

This all leads, perhaps inevitably, to a reiteration of our bearish gilt bias, relative to other markets.

There will undoubtedly be some celebration when, in sort order, UK annual inflation falls below target. There might even be an election, according to some reports. We’d hold off on ordering the cake.

*simply yelling “growth” repeatedly doesn’t count.

Further reading
— The unbearable tightness of BoE’ing
— The Bank of England is misusing its fiscal powers

Source: Economy - ft.com