can build on that. I think we really had to learn from the experience of 2022 where I think it took the economy a while to understand what the pressures were on food prices going into that summer and then winter, coming from energy prices, but then fertilizer and other inputs. Then we can talk about second-round effects and all of that. What's happened this year, as Andrew says, is interesting up to now and has supported inflation perhaps not rising as much, certainly in food, as we might have expected. The latest reading for July was about 1.7%. We're expecting food to go up to 3.5% by the end of the year, and the trade bodies have come down a bit. They were originally saying, when the conflict started, maybe 6% or 7%, and they've come down a bit.
But you do then have to look at your fertilizer that you've got to buy for next year. You've got to take account of the difficulty you've had with yields this year, as you say, already using up winter silage for feed. And then, as Andrew was saying—and we heard this again when we were in Basel at the weekend; the great thing about the Basel meetings is you get central bank governors from right around the world—so the central bank governors from some of the countries more exposed to El Niño now, because with us it will happen with a lag, were talking about the impact on food prices that they're already experiencing. They're some of the biggest producers in the world, so those will feed through to us.
Thank you. Given the fact that climate change is a reality now, are we now at a point where this volatility, this pressure on food prices, and the pressures on supply mean that we're dealing with a chronic inflationary problem?
Well, I'm not an expert in climate science, so El Niños do obviously happen. They're not necessarily to do with climate change. This is a big one. I think everybody's saying this is a big one, but again, I'm not sure the link necessarily follows. We've obviously had very hot summers in this country before. I remember 1976, which was a very hot summer—sadly, I'm that old. Certainly when you look at global food prices, we are seeing more volatility in recent years. If we go back to 2022, that was not to do with climate; that was to do with Ukraine predominantly.
I think it's the case that whether we touched on geopolitics, talked about energy, now we're talking about food, we seem to be in a world where there are more shocks and more supply shocks. In the short term, the drought and El Niño will have a supply impact because they will affect production. But then back to what we were discussing right at the outset, what actually happens to consumer prices, which is after all what we target, will depend on demand. Obviously food is a necessity, so that really affects the cost of living. But also the incidence of these shocks, it will depend on is it global or can other areas compensate for it, and then how can demand adjust? If meat prices are very high, for example—you were talking about the lamb auction you were at—does that mean consumers substitute away to the extent that they can? But I would say that this environment of more and more significant supply shocks is more likely to have a ratcheting up in prices than a ratcheting down. We haven't seen many positive supply shocks of late.
Can I just make a simple but important point? I think to see ever-rising inflation, you need to see these shocks get worse and worse and worse every year, all else equal. What I think we're looking at is more frequent supply shocks, but that does mean there's just going to be a lot of volatility rather than just ever-rising inflation.
And if I may, just one more. Just talking about a very traditional sector there, but you've looked at the AI boom recently and the pressure that that's exerted on the supply of chips. Where does this factor in just generally in terms of that inflationary pressure?
Well, there's a box in the MPR on the subject. Because we are seeing some upward pressure on chip prices, I think the staff have done a bit of modeling on it. They've come down about 0.1 on the CPI potential effect. But again, we have to watch this because the whole issue here is obviously the scale of demand for chips and how that evolves. So we're going to have to watch this quite carefully. But yes, it is having an impact, and it has an impact obviously because it's not just AI, because there's chips in lots of things. So to the extent it affects broader chip prices, that feeds through. But our staff have done work on that, yeah.
Can I just add a quick point on that? One thing that nobody knows how to model is how these different supply issues interact with one another. I think that's a big degree of uncertainty for all of us. This AI piece might interact with the Iran war piece, which might interact with other supply shocks to create an effect that you can't just model directly.
Thank you very much. Chris Coghlan MP.
Chair, I turn now to quantitative tightening. Dave, the estimates of the impact of QT on interest rates have consistently edged up over time. Is that because the actual impact of QT has edged up or because the estimates have changed?
Well, it's fair to say—and look, I think you give the right framing when you say that they've edged up. A year ago, we said that we thought the overall impact was between 15 and 25 basis points. We've gone up to saying it's an impact of 20 to 30 basis points. Remember that's in the context of an overall increase in yields of about 350 basis points and an overall increase in term premia, which is the bit that we think QT impacts, because remember our principles are all about being gradual and predictable and QT operating in the background and bank rate being our primary tool. So 20 to 30 in the context of a 200 basis point increase in term premia, I would argue, is still very much in the background.
Whether it is because another year we've done a bit more QT or our estimations have been refined, I think you're talking about such small amounts that I wouldn't want to say it's one or the other. The points I've made are that it's a very small percentage, somewhere between a sixth and a tenth of the overall impact on term premium. And remember, over that period, QT has amounted to 400 billion pounds—we've gone down from 895 to under 500 billion, so this has been a very significant adjustment.
I've seen some analysis by Bloomberg that apparently the fiscal losses being incurred on QT today currently relate to the QE that happened during the pandemic. You, of course, if I'm right in saying, voted to curtail the final 150 billion of QE during the pandemic. In hindsight, do you think quantitative easing was the correct policy response to the pandemic given the inflation we're seeing now, or was hindsight a wonderful thing?
Well, look, this committee—I'm not sure you were on it then, but this committee carried out a very extensive inquiry into QE and the start of QT, I think in 2023, and I was asked a very similar question to that one. I think it's fair to say that Andrew and I did that inquiry with Ben Broadbent, who was then a deputy governor. We were all asked what lessons we would learn. The one lesson I would learn is a second-order one. I remain very supportive of having QE as a tool when you're at the zero lower bound and when you're dealing with an economic crisis like we were with COVID. So I obviously can't say if I went back in time would I vote again for the first and second and third rounds, but I'm pretty confident I would because of what they did to support the economy through a horrendous economic period.
However, that third round of QE—thank you, always have to correct the record—that third round of QE was, if I recall, 150 billion over 12 months. It was an unusually long period going into 2021. At the outset of that, we didn't even have a vaccine.
But with the benefit of hindsight, I would have had a review point for such a long period. We didn't have a review point at the time, so I started to vote against completing it. I was in a small minority, but the reason I voted against completing it was because I was beginning to worry about inflation.
And so it was the first tool available, and then I started to vote, I think it was with Michael Saunders, both to stop QE and to start putting up bank rate, but it was only a month or so before the rest of the committee.
Can I just make a rather dull and arithmetic point, sorry about this, but I think it would be true to say, if I remember the analysis, that in cash flow terms there will always be more of the bonds from the last round of QE before you raise interest rates in the portfolio. At this point about where's the relative cost, I think that's probably—there will always be more, yeah.
I thought you might be interested in coming in. I'm really interested to hear your thoughts on something from 1844, the Bank Charter Act.
Specifically in terms of the accounting treatment, the Treasury receives the full fiscal benefit or losses from QE and QT, I think as a result of that act, whereas the Fed and the ECB do not because of different accounting treatments. Therefore, if the government was to change accounting treatments similar to the Fed and the ECB, would that be one way forward on this issue in terms of not having a taxpayer accumulate 20 billion of losses halfway through?
To make a point which you may say goes against my interest, but I think it's important: in economic terms, this makes no difference at all. First of all, this is all about accounting, but it is also about transparency. That's the point. Because in terms of the taxpayer, let's think about the consequence of the 1844 Act. Let's be clear: the 1844 Act—because I hope very few of the public carry around in their head how the 1844 Act works—splits the Bank of England balance sheet into two parts. There's what we call the Issue Department, which has on it—the only liability is banknotes and they don't bear interest—and the rest of it is what's called the Banking Department.
Now, the reason this was done in 1844 was all to do with the gold standard, so that's history. It was a way of actually making the gold standard work because the gold was on the Issue Department to back the notes, because the Bank originally was a commercial bank, actually. But it's persisted. These days, the so-called seigniorage, which is the profit of the note issue—because every note we've got is effectively an interest-free loan to the Bank of England—because of the way the 1844 Act works, all the profit of the Issue Department is paid directly over to the government, to the Treasury. The only bit that we withhold is the cost of actually printing and issuing banknotes, which is very small compared. So it's about 4 billion pounds a year.
Now, in other central banks, it's all in one balance sheet. The seigniorage accrues to the profit of the central bank, and then they usually have some process in normal times for deciding what gets paid over. The consequence of that is that other central banks therefore are able to do two things actually, but they're able to essentially offset the seigniorage against the cost of the QE/QT. They also do, by the way, a second thing which is a consequence of that, which is that they also book a negative asset. In other words, they book the negative cost of the QE/QT because obviously the one-year seigniorage doesn't cover it. They're therefore projecting forwards to say, for as long as we need to offset this, we'll take seigniorage into the future. Actually, we'd have to change our accounting standards to do that, but they do that.
We can't do that because we have got very little income outside seigniorage. The Banking Department has a relatively small income. We pay a dividend to the Treasury, but it's about 100 million a year, I think, under that. We've only just started resuming that. So the consequence of that is that we are, in a sense, paying money to the government with one hand and it's having to come back with another hand. Other central banks don't do that. In economic terms, for the taxpayer, it's the same because obviously—
It is, but the fact is there's a fiscal hit to the taxpayer today from the current-year losses.
