Snap, crackle, and pop: understanding inflation and how it is measured
Economists and commentators love to use movement-related words when describing the economy. Is the economy speeding up or slowing down? Coming to a stand-still? Is inflation accelerating? Are interest rates pausing or pivoting? Such words are often helpful in conveying a broad sense of dynamics, but can be baffling to the lay reader.
Why is that? Well, these words are essentially borrowed from physics and maths and have very precise meanings in those contexts. A further complication is that both “GDP growth” and “inflation” are variables that are not measuring a level of something, but the change in the level of something. GDP growth is an increase in the level of output. Inflation is an increase in the level of prices. And both are often (but not always) measured on an annualised basis, i.e. output or price today relative to output or price one year ago. That means that, by definition, the figure we get — GDP growth or inflation — depends a lot on what was happening a year ago, which is not that easy to remember.
A further implication of all this is that when we talk about an increase in inflation, we are already onto the second derivative of prices relative to time: a change (upwards) in the rate of change (upwards) of underlying price levels. “Accelerating” inflation would, technically, mean a change in the rate of change in the rate of change in underlying price levels. Since acceleration is the 2nd derivative of position in respect of time, and inflation is the 1st derivative of prices in respect of time, an acceleration in inflation would be a “jerk” in prices, i.e. the third derivative of prices in respect of time. [For anyone interested — and I know there’s only so far we can take this analogy— the fourth, fifth, and sixth derivatives of position with respect to time are called snap, crackle and pop.]
Why does any of this matter? Well, I think the derivative nature of inflation is making it really hard to explain what might happen on the way down. Almost invariably, when I mention to people that inflation might drop quite sharply next year, they find it hard to understand why (and I find it hard to explain, hence this post!) Surely, we are still going to see high energy and commodity prices? Doesn’t that mean high prices? Well, yes it does, but it doesn’t necessarily mean increasing prices.
Energy futures markets are predicting that, despite continued tightness, prices of gas and oil are about to decrease (and prices have, indeed, already reduced significantly from the peaks reached earlier this year). Oh, so inflation will come down? Well, yes, but it would have come down even if prices had stayed the same. A stable, high price that doesn’t change delivers zero inflation. A reduction in prices would mean that inflation will go negative.
The chart above is a simple illustration of this, showing price levels (white lines) and inflation (blue lines) side-by-side. The dashed lines are based on currently traded forward prices; the solid lines are out-turn prices already registered. To keep things simple, I’m using energy prices (or, more precisely, natural gas prices at the UK’s National Balancing Point Virtual Trading Point) as the variable here. Clearly all kinds of other prices are also going up (and down) and are showing up in the consumer price index, but it gets a little complicated to illustrate the point with the full basket of items. (It is also the case that household prices do not change on a monthly basis in line with the wholesale gas price, but let’s ignore that further complication for now.) In any case, according to the ONS, natural gas contributed 1.3%-points of the 10.1% inflation in September 2022. So it’s an important driver, and one that has been very volatile.
So let’s look at what has been going on with natural gas prices. Take point 1 in the graph, March 2022. Gas prices went up to nearly 300 p/therm and (annualised) inflation really shot up, to 540%. In August 2022 (point 2 in the chart), prices climbed even higher, to 460 p/therm, but inflation dropped significantly, to 260%. How is that possible? Well, inflation is the price level today relative to prices a year ago. In March 2022, 300 p/therm was being compared to a relatively low number, of around 50 p/therm. In August, the 460 p/therm was being compared to an already high number, 130 p/therm. So, prices went up, inflation came down. Because of the mechanics of the calculation, this happens all the time.
What about July 2023 (point 3 in the chart)? Looking at the white dashed line, futures traders for gas are anticipating a continued very high price, around 300 p/therm. But the blue dashed line shows that with those prices, inflation would be negative. Why? Well, while 300 p/therm is a very high price compared to, say, the last 10 years, it is still considerably lower — in fact 14% lower — than the price in July 2022, at 350 p/therm. This is one of the key reasons the Bank of England expects inflation to drop sharply in 2023. For most people, this doesn’t make any sense, because they are anchored on the idea that prices will still be high. But, as is clear by now, that has nothing to with whether price increases will be high. [Well, it has something to do with it but that’s another blog for the future.]
Finally, let’s look at the blue dashed line for 2024 and 2025. Despite the forward price (white dashed line) remaining elevated relative to long-term historical averages, inflation (blue dashed line) is showing up as negative. This is because, even if high in absolute terms, gas prices are successively lower as time goes by. Mathematically, that means negative inflation. Or, in a broader context, given that natural gas is only one part of the consumer price index basket, that the contribution from gas prices will be negative.
I hope this helps explain some of the counterintuitive statements on where inflation might go in the future. These are not projections, and in fact the forward price curve has not been a great predictor of future prices historically. But it is a meaningful benchmark. Nevertheless, I hope the illustration is useful.