Fed Chair Kevin Warsh made an unusual announcement this week — one Fed Open Market Committee (FOMC) meeting in 2027, and two meetings in 2026 will not be “live”: they will sit around and yack, but will not change rates. And as a reminder to those who are unfamiliar with the FOMC, they have 8 scheduled meetings per year (roughly every six weeks), although they can make emergency rate moves in between scheduled meetings (usually panicked cuts).
Losing one live meeting is not that big a deal, but cutting it to six meetings is starting to change the complexion of things. The problem is straightforward: central bankers generally like to have smooth rate change regimes. Until the last cycle, the preference evolved towards 25 basis point hikes per meeting (“gradualism”). The last cycle featured more aggressive rate hikes (which surprised me), but they felt that they got behind the curve. That cycle was unusual, and if we have more “normal” inflation behaviour, the Fed is going to be in the position of pre-emptively tightening before inflation gets out of hand (which is what the Bank of Canada is doing now).
The problem with six meetings per year is that if gradualism is followed, that only allows for 150 basis points in hikes a year. That is not going to be enough if the Fed follows its previous track record and doing hefty rate cuts when it panics as a recession hits, reversion speeds need to be faster than that. (The panicked nature of rate cuts in practice means that the meeting schedule is not an issue for rate cut campaigns.) This means that the Fed would likely have to intersperse fatter rate hikes — 50 or 75 basis points. This in turn will cause markets to panic, as they will likely feel that the Fed is panicking about inflation (like in the early 2020s). The end result is the bond market having to price in a greater risk (term) premium, which is an outcome that all financial commentary wrings its hands about. (“Oh no, the fiscal risk premium will go up, raising costs for borrowers!” is standard fodder for the financial press, particularly in the United Kingdom, where the financial press are fiscal risk nutters.)
This does not matter if policymakers think the rate settings are roughly correct, but it is somewhat stupid to premise counter-cyclical policy settings on the basis that we will never get another recession.
The only justification I see for this is that the Fed Chair is tired of being hassled by the President about cutting rates, and this move gets him peace and quiet for a couple months during the year. This is an unusual situation, and will not persist.
Does the Fed Need Time To Discuss Strategy?
The argument for declaring meetings not to be “live” is to allow more strategy discussions. Believe it or not, there exists modern telecommunications tools that allow groups of people to yack from locations all over the world, they just need to connect to the “internet.” Since such meetings have very limited coordination time costs, one could have dozens of such meetings in a year without greatly affecting other duties.
If you are doing macro strategy in the private financial sector, there is an argument for fairly regular strategy meetings. The reason for this is straightforward: you are buying and selling securities (and derivatives) that reprice continuously during the business day (and after hours), and timing buys and sells matters.
This is not true for central bankers. They are adjusting policy variables in response to macroeconomic conditions (and very occasionally, responding to financial crises). Useful economic data for the real economy is generally collected at a monthly frequency, and the data are noisy. Outside of a crisis, central banks are concerned with “supply constraint” measures (like the output gap) that move slowly. If you are a believer in some variant of an output gap, if the output gap indicates that there is a lot of spare capacity right now, in the absence of a crisis, there will spare capacity three months in the future. Unless you subscribe to extremely wacky theories about monetary policy (e.g., Market Monetarism), deviations in the policy rate by 25-50 basis points will have almost no measurable effect on the economy. Outside of a crisis, monetary policy is about adjusting policy variables slowly based on forecasts one- to two-years out. Unless your models stink, those forecasts should not move by much month-to-month, so we should expect policy rate changes to follow sweeping up/flat/down trajectories.
The reason why central bankers need periodic meetings is to discuss whether new data require revising their forecasts, and to make those slow changes that their long-term beliefs require.
Cutting the number of live meetings forces policy away from gradualism, it implies a more hyperactive monetary policy. Policymakers need to jump immediately to the “correct” value. In his “gradualism” speech, Bernanke called “bang-bang control”, which is niche control systems jargon. If you ever spent afternoons as a teaching assistant in a control systems laboratory, you will discover that the description of “bang-bang control” is a literal description of how the systems operate. “Bang-bang control” was a child of 1960s optimal control theory, which was based on the typically bad assumption that you had an extremely good model of the system and know its state exactly. This worked for path-planning problems (getting to the moon was the apex of optimal control), but it breaks down in the presence of uncertainty. Central banking is shrouded in uncertainty, which is why more gradual policy changes are prudent.
No comments:
Post a Comment
Note: Posts are manually moderated, with a varying delay. Some disappear.
The comment section here is largely dead. My Substack or Twitter are better places to have a conversation.
Given that this is largely a backup way to reach me, I am going to reject posts that annoy me. Please post lengthy essays elsewhere.