Warsh’s Fed and the return of inflation risk. Is the era of cheap money over for good?

Jackson Hole did more than raise the probability of a rate hike in September. Kevin Warsh, the current Fed Chair, has begun to explain what kind of Federal Reserve he wants to build and, if he really follows through on what he said, the change could be far deeper than 25 basis points. For fifteen years we grew used to an extraordinary combination of near-zero rates, gigantic balance sheets, bond purchases and central banks willing to stop the price of money from becoming too serious a problem for markets. We ended up treating as normal something that probably never was. Now inflation refuses to return to 2%, the US deficit forces the Treasury to place enormous amounts of debt, and the long end of the curve is starting to behave with an autonomy the Fed does not fully control. Warsh may raise rates in September or he may not. The real question is a different one. If the bond market once again demands a real price for lending money for ten, twenty or thirty years, can a Federal Reserve, however much it wants to, take us back to the world of cheap money?

For many years Jackson Hole has turned into a kind of linguistic interpretation contest. The Chair of the Federal Reserve gives a speech, changes a word, adds an adjective or drops an expression and, a few hours later, half the market is trying to work out whether it means a hike, a cut or simply that the man was tired when he wrote the paragraph. It has reached rather absurd extremes, but it works because the Fed taught us precisely that: to communicate months in advance what it intended to do and to lead the market almost by the hand.

Kevin Warsh had already announced other plans. At Jackson Hole he did not say he would raise rates in September and, consistent with his well-known aversion to forward guidance, he went to some lengths to avoid doing so. But he achieved something curious, because the market left far more convinced there will be a hike than before it heard him. The implied probability went from roughly 35% to around 55%-60%, while the two-year Treasury reacted with a rise of close to 13 basis points to around 4.36%. For a two-year bond, a move like that in a single session is hardly a minor detail.

We could now discuss whether Warsh will persuade most members of the Fed’s FOMC to raise rates by 25 basis points on 16 September. That will surely be debated over the next fortnight and will depend largely on the employment and inflation data still to come. To me, frankly, it is the least interesting part of what happened in Wyoming, because what really mattered was that Warsh began to explain what he understands by price stability and, above all, what kind of Fed he wants to chair.

He said the 2% target is firm and fixed. Some readers will think this deserves little attention because, if the Fed’s target is 2%, it stands to reason that it wants 2% inflation. Well, I am not so sure markets have been operating on that idea in recent years. We have grown used to accepting that 2.5% is close enough, that 2.8% can be explained by a temporary disturbance, and that any somewhat larger deviation always has some special cause that will disappear with time.

The problem is that time is passing. US CPI remains around 3.7% and the core component is around 3.3%. More uncomfortable still, roughly half of its components are still growing at annualised rates above 3%. After several years waiting for inflation to find its own way back to target, it is starting to be reasonable to ask whether it really will.

If the new Fed decides to intervene less at that end of the curve, the market will have more freedom to set the price of US debt. And we are back to the deficit.

Here Warsh introduced an idea that strikes me as far more important than the whole debate about September. Inflation does not necessarily revert spontaneously to its mean, and price stability does not deliver itself. Translated into language less typical of a central bank: if it does not come down fast enough, the Fed will have to make it come down. And that does mark an important change from a time when the priority almost always seemed to be preventing monetary tightening from doing too much damage to markets or to growth.

For much of the period after 2008 the central banks’ problem was exactly the opposite. They could not generate enough inflation. There was mediocre growth, excess savings, globalisation, relatively cheap energy and an enormous amount of capital looking for returns. Rates could stay near zero for years without serious inflation appearing. Today we are starting to discover that perhaps that world was far more exceptional than we thought at the time.

Tariffs are a first difference. I will not go back into the almost theological debate on whether a tariff generates permanent inflation or only raises the price level once. It will depend on what happens next. If the importing company absorbs the cost against its margin, the effect will be one thing. If it passes it on to the customer, the supplier does the same, trade retaliation appears and workers try to recover purchasing power, it will be something quite different. We can argue for hours about what to call it, but the consumer will not care in the slightest when paying the bill.

Then there is energy. Iran, Hormuz and everything happening on the shipping routes have reminded us once again that the economy still depends on things as old as ships, oil, gas and thirty-kilometre straits. The Fed cannot produce oil by raising rates, nor make gas, chips or transformers. What it can do is prevent an initial shock from feeding into wages, services and expectations. That difference explains why a central bank may be forced to tighten monetary policy even when the original cause of inflation is completely beyond its control.

And then there is artificial intelligence, which is probably one of the most interesting paradoxes in all of this. For months we have been hearing that AI will be disinflationary because it will raise productivity extraordinarily. It makes sense. If we produce more using the same resources, the economy’s potential capacity will increase and some price pressures will ease. I hope so too, not least because the sums of money being invested are starting to be large enough that we had better find productivity at the end of the road.

But first it has to be built. Models do not live in thin air. They need data centres, semiconductors, power grids, power generation, fibre, cooling, land, transformers and a vast amount of capital. All of that is in demand today, while much of the productivity gains will arrive tomorrow. So we may well find ourselves with an extraordinarily disinflationary technology in ten years’ time that, during its construction phase, adds pressure to certain parts of the economy. Anyone who says today they know precisely what the net result will be probably knows a lot more than I do, or is simply making it up and does not want to admit it.

But there is another issue that worries me far more than all of the above and which, curiously, does not depend on the Federal Reserve. The United States is running a fiscal deficit close to 6% of GDP with an economy that is not in recession and public debt held by the market that is already around 100% of output. If current projections hold, that debt will keep growing over the next decade.

At this point the question is inevitable. Is the era of cheap money over for good? I have no idea, and anyone who answers yes today with absolute certainty probably does not either.

Here we do have a problem that is fairly easy to understand. For years we have explained that the deficit should increase in recessions. The state collects less, certain expenses rise and demand is temporarily supported. Fine. The problem arises when you have a deficit close to 6% in an economy near full employment, because then it is inevitable to ask how big the deficit will be when the next recession arrives. And it will arrive. I have no idea when, but it will.

Some readers will tell me that the United States issues the world’s reserve currency and will always find buyers for its debt. I agree. The question has never been whether buyers will appear. The question is at what yield they will be willing to buy. At 3%, 4%, 5% or 6% they will appear, but someone will have to accept that the cost of financing the US government may be structurally higher than during the decade after the financial crisis.

And here two stories start to diverge within the same bond market. The two-year Treasury looks fundamentally to the Fed. If Warsh seems more worried about inflation and the chance of a hike rises, the two-year responds. That happened immediately after Jackson Hole. The thirty-year, on the other hand, has a longer memory and many more problems.

Anyone buying a thirty-year bond does not only have to worry about what Warsh does in September. They have to think about future inflation, growth, the amount of debt the Treasury will issue, international demand, the term premium and, increasingly, US fiscal policy. If the government needs to place enormous amounts of debt for years, someone will have to absorb it and, if the price offered is not attractive, the yield will have to rise until the buyer appears. There is not much mystery to it.

Bessent can buy back bonds to improve liquidity, change the issuance profile and ease certain points on the curve. It makes sense and may work. What he cannot do is make an investor accept, for thirty years, a yield they consider insufficient given inflation, debt and fiscal risk. And this brings us to a rather unpleasant possibility: the Fed could end up cutting rates and money could still remain expensive.

Imagine, some time from now, Fed Funds at 3% while the ten-year Treasury stays at 4.5% or 5%, the thirty-year remains above 5% and an American family is still financing a home at 6% or 7%. Technically the Federal Reserve will have cut rates. For anyone buying a house or financing a business, the joy will be rather limited.

For too many years we have confused low official rates with cheap money, and they are not exactly the same thing. The period after 2008 combined near-zero rates, a massive expansion of the Fed’s balance sheet, purchases of long-dated debt, forward guidance, contained volatility and a fairly widespread conviction that, when markets suffered too much, the central bank would eventually show up. The famous Fed put was not a written law, but it worked enough times for many investors to end up behaving as if it were.

Warsh seems uncomfortable with almost all of that. He has argued that short-term rates should once again be the main instrument of monetary policy and that extraordinary measures should be reserved for extraordinary situations. This matters far more than it seems, because the expansion of the balance sheet did not only increase liquidity. It also took duration out of the market and helped compress long-term yields.

If the new Fed decides to intervene less at that end of the curve, the market will have more freedom to set the price of US debt. And we are back to the deficit.

His rejection of forward guidance is no small matter either. For years the Fed told the market what it intended to do, the market adjusted its prices according to those signals, and then the Fed looked at those prices to interpret what the market expected. I admit the mechanism had something circular about it. I tell you what I will probably do, you change your mind because I told you, and then I use your change of mind as information about what I should do.

Apparently Warsh wants to scale back this game and return to a somewhat quieter Federal Reserve. That will probably increase volatility, but I am not sure that is a bad thing either. Perhaps it is not very healthy for trillions of dollars to be revalued because the head of a central bank has swapped one word for another at a press conference. Markets might even go back to doing something they used to do: analysing the economy, credit, earnings and risk for themselves.

If in six months the United States enters a severe recession, banks start failing or unemployment soars, Warsh will cut rates and use the balance sheet if necessary. Just as any reasonable Fed Chair would. Monetary doctrines are wonderful until someone smells smoke.

That is why I think the right question is a different one. It is not about whether we will one day have low rates again, because we surely will. It is about whether we will once again consider normal a world in which practically the entire price of capital is artificially compressed.

There I have many more doubts. The deficit is bigger, so is the debt, globalisation is retreating, energy has once again become a geopolitical problem, the need for physical investment is enormous, and the central bank that for years helped lower the cost of capital is saying it wants to intervene less.

This directly affects valuations. Growth companies suffer more when the discount rate rises, especially those whose profits lie far in the future. Private equity discovers that buying a company financed at 7% has absolutely nothing to do with buying it financed at 3%. Companies that survived for years by refinancing very cheap debt are starting to face far less pleasant maturities, and even cash is once again competing seriously with equities.

This last point seems especially important to me, because some of today’s investors have spent practically their entire careers in a world where the risk-free asset paid little or nothing. When a Treasury bill offers around 4%, buying a company at thirty times earnings requires a much better explanation than when that same bill offered zero.

Warsh may raise rates in September, or the next data may force him to wait. For the underlying question it matters relatively little to me. What is truly relevant about Jackson Hole is that the Fed is once again talking as if 2% really meant 2%, that the bond market is starting to demand a considerably larger premium for financing the US government for decades, and that the new Fed Chair seems to want to remove part of the monetary safety net that markets had grown used to.