Archive
A Price-Linked USD
Let me introduce you, for those who aren’t already acquainted, to the Unidad de Fomento.
The Unidad de Fomento (UF) is an almost-unique currency in the world.[1] It was established in 1967 by Chile as a non-circulating currency – I will get to that in a minute – and has survived a period of hyperinflation and a rebasing of the circulating Chilean currency from Escudos to Chilean Pesos in 1975. That is an amazing testimony.
What is unique about the UF is that it is directly indexed to the price level. The value of the UF increases (or in theory decreases) every day with the inflation index. This means that unlike the actual circulating currency, the UF maintains its purchasing power over time. If you could buy a physical UF, it would be like buying a 1967 Chilean peso. As long as you hold it, you will be able to buy exactly as much actual stuff as you could have in 1967, or for that matter last year. If you put enough UF in your pocket to buy an empanada when the price of an empanada was 1,000 pesos, then when you take it out of your pocket you should still be able to buy an empanada no matter what the price of an empanada is today.[2]
Every day, what is changing is the exchange rate between 1967 pesos and ‘today’ pesos, and the only part of that which is changing is the price level. That is, after all, exactly what a price level means. It means snapshotting the value of this basket of goods and services in, say, 1983 (as for the CPI) and then telling you what roughly the same basket of goods and services (allowing for changes in the consumption basket over time) would cost today. So the NSA CPI today at 314.54 means that if the consumption basket cost $100 in 1983, today the same basket would cost $314.54. Except that if we had a UF for the dollar, we would simply say the basket cost 100 UF$ in 1983 and 100 UF$ in 2024. That’s powerful.
I noted that the Chilean UF is a non-circulating currency. Then what good is it if you can’t actually buy goods and services with it?
The purpose of the UF was to facilitate contracts and wage agreements in a period of inflation uncertainty. When the future price level is unknown, negotiating longer-term agreements becomes more difficult. Consider a labor union negotiating a multi-year labor contract. The union, who is contracting to provide labor at certain future prices, will want larger increases to protect itself from the possibility that those future wages are eroded by higher-than-expected inflation. Management, on the other hand, wants lower increases because agreeing to larger increases if inflation is lower than expected will make its cost structure less competitive. For a short-term contract, the risk on both sides is low. But the longer the term of the contract, the more the risks grow for both sides and the harder it will be to reach an agreement. Where this manifests is that in low and stable inflation regimes, contracts tend to have longer tenors; in high and unstable inflation regimes, contracts tend to have very short tenors or to be completely untenable.
This is the problem that UF solves. The parties to a negotiation no longer have to protect their nominal wages and prices. They have offsetting risks to inflation, so they agree to real wage increases or price escalations by negotiating not in Chilean Pesos but in UF. Then, as time passes, those agreed-upon UF amounts are translated at the then-current exchange rate between UF and Chilean Pesos – reflecting the change in the price level that has actually occurred.
There are many salutatory effects to this. Suddenly, the need to get ahead of wage and price increases and negotiate larger increases to protect against inflation vanishes – and, with it, the feedback loop where higher wages induce higher prices, which in turn induce higher wages. (As I’ve discussed previously here, that doesn’t necessarily accelerate inflation, but it makes inflation much stickier going down than it is going up.)
Why am I mentioning this?
We do not have anything like this in the United States at this point. We have CPI, and companies do negotiate contracts on the basis of CPI. Longer-term construction contracts, such as for power plants or airplanes, often have escalators tied to particular price indices. Frankly, the ability to base a contract on a non-circulating currency is not itself something that is necessary in the USA today[3] although in 1967 in Chile it was. However it has escaped no one’s notice that over time, our currency is a medium of exchange but becomes less and less a good store of value when inflation moves away from the zero bound. Obviously, there are investment products which help address this issue but to the extent that there are any cash balances, higher inflation implies higher monetary velocity partly because the money itself becomes a worse and worse store of value.
One solution to this would be inflation-linked savings accounts, which don’t exist (although I’ve tried to convince people to do that!) Another solution would be to have a currency – a circulating currency – which you could hold which would keep up with inflation. Such a currency would be superior to USD, because it would be USD preserving the purchasing power of the base year.
Why don’t we have such a thing? Well, the US has a monopoly on issuance of its currency, and every time they issue more it is a pure gain to the government, called seigniorage. But that’s not really the reason, because the government would also earn seigniorage on “USDUF.” The real reason is that the USD is a successful fiat currency because, and only because, everyone believes that everyone else will accept it as worth $1. Any new currency, if unbacked, would have to generate trust anew that everyone else would accept the USDUF-USD exchange rate to be equal to the price level. Unless the US government guaranteed that exchange rate by freely exchanging dollars for USDUF on demand, there is no guarantee that USDUF would trade at the appropriate level.
I don’t see the government doing this any time soon. But it really should.
[1] In 2001, Bolivia established a similar currency, the Unidad de Fomento Vivienda, which is based on the same mechanism but is not as widely used as the Chilean UF due to the latter’s long head start.
[2] I am taking some liberties for the sake of illustration. Obviously some goods keep up with the price index, some run ahead and some fall behind…it’s really only true that the value of the UF is constant in terms of the overall consumption basket, not each specific item.
[3] …although having historical financial statements and financial projections in real terms instead of nominal terms, using real discount rates instead of nominal rates, would make a ton of sense and you can show that corporate finance in real space is more efficient and more sensible if there’s any volatility in inflation.
Inflation Guy’s CPI Summary (July 2024)
It was only a few months ago (with the March CPI report in April) that I was talking about a ‘Potential Pony Situation’ in my podcast when, after an unsettling Core CPI, I pointed out that the Median CPI was much less disturbing. Trying to tell the story of the economy is about figuring out where the underlying trends are, and trying to figure out what you can ignore as ‘noise.’ Back then, it was clear that inflation was heading lower, but not as fast as people were saying, so the bad core CPI was off-putting. It messed up that story. But because we were focused on Median CPI, that month was not so unsettling and we focused (successfully I think) on the fact that inflation was decelerating…but not collapsing back to target imminently. Fast forward, and the story we are looking at coming into today’s CPI is that inflation is still declining, but people are probably getting a bit out over their skis in anticipating (again) a rapid collapse in inflation after a couple of weak CPI prints. Once again, that’s not the story the data is really telling, but deviations from that belief are likely to be painful.
For what it’s worth – I saw a lot of commentary this morning about how “PPI is encouraging,” or “PPI means this or that.” No one in the inflation trading community cares much about PPI. There are some elements of the PPI report that can help with some of the parts of other inflation reports, but the overall number has very little correlation (and no lead) with the CPI. You and I are exposed to CPI. The Fed looks at consumer prices. My best advice about PPI is to ignore it.
When CPI actually came out, it was a touch better than expected on the surface. Economists had been looking for +0.19% m/m on core, and got +0.155% on the actual number. What was fascinating to me was the market reaction. Equity futures appear to be completely flummoxed by an as-expected number, vacillating around unchanged 20 minutes later as I write this. I think this tells you something, actually – folks coming into today weren’t trading the actual number but rather planning to trade what other people thought about the number. Everyone thought everyone else knew what a higher-than-expected or lower-than-expected number would do. An as-expected print means you have to dig into the details, and equity guys don’t like details. They like big pictures. Thick lines. Crayons.
So let’s look at some pictures. Here are the last 12 core prints and the 8 major subcomponent pieces.
The first thing that jumped out at me was that core goods again plumbed new 20-year lows. Yes, that’s 20-year lows, as the following chart shows. -1.9% y/y.
Folks, I am still waiting for the turn and I say every month “surely, it can’t go lower than that.” So far, so wrong. The dollar is no longer strengthening in a straight line, and hasn’t been for a while. If anything, it’s weakening. Apparel this month was -0.45% m/m, and only 1.1% y/y. Apparel is almost entirely imported, and at some point a steady-to-lower dollar will mean that core goods heads back to flattish. (Also, keep in mind that both Presidential candidates have expressed pro-tariff positions, but that’s a 2025 story at the earliest).
Within Core Goods, we also saw Used Cars decline yet again. This month it was -2.3%. CPI had diverged a bit from the private surveys, but with this month has basically converged back to the number implied by Black Book. That doesn’t mean Used Car prices won’t decline further, but there’s no longer a reason to expect “bonus depreciation” going forward.
Now, in the first chart above note that Core Services dipped to 4.9%, the lowest it has been in a while also. Within core services, we saw Airfares decline again (-1.6% m/m after -5% last month), but the interesting thing is Hospital Services. The other parts of Medical Care, that is Physicians’ Services and Medicinal Drugs, were both in line with recent trends and on top of last month’s figures. Hospital Services plunged -1.1% m/m. The y/y is still pretty high at 6.1%, but if this number is prologue (I sort of doubt it) then this upward pressure will abate.
The fact that services dropped so hard helped to bring “SuperCore” down a little bit. It is still elevated, and frankly the trend doesn’t look wonderful. You want 50bps in September? You need more than this, pal.
Do you know what I haven’t mentioned yet? Shelter. Shelter is the biggest and stickiest piece, and the foreordained deceleration of shelter is part of the religion of everyone who thinks we will decline to 2% core inflation and remain there (which is basically where breakevens are these days). Bad news – this month, Primary Rents rose 0.49% m/m and OER rose 0.36%, compared to 0.26% and 0.28% last month. This is where it’s useful though to look at the y/y numbers. That big surprise in Primary Rents produced an unchanged y/y number and OER still decelerated to 5.30% from 5.45%. The wonder of base effects!
So let’s harken back to the beginning of this piece. In ‘A Potential Pony Situation,’ the Median CPI warned us to not get too worried about the surge in core because Median was pretty well-behaved. In the current circumstance, Median tells us to not get too excited by all of those people who will be talking about how low the 3-month average is (I guarantee that old chestnut will make a reappearance this month), because Median will be something like 0.268% (my early estimate). This will be the highest since April, if I am right.
The bottom line remains the same, and that is that inflation continues to decelerate but median is going to end up in the “high 3s, low 4s.” I keep thinking that we will dip below that for a little while when the base effects of shelter pass through, before reaccelerating to what I think is the new ‘normal’ level, but shelter is being persnickety and resistant to that deceleration. Either way, there is nothing here that would encourage the Fed to aggressively ease 50bps. Or, for that matter, to ease at all. If the Fed eases in September (which I expect, even though if I were a member of the Board I wouldn’t vote for one), it will be because its members fear recession and not because there is evidence that inflation is licked. That evidence is still elusive.
