Has Powell F****ed It?
Has Powell F****ed It?
The Fed's Joe Biden Mindset
On Thursday, the Federal Reserve announced that (as the markets more or less expected) it would not cut rates in July, but Chairman Powell heavily hinted at a rate cut in September - especially if labor market or output indicators worsened significantly. On Friday, the labor market report came out - and it was really bad. Among other signs of weakeness and causes for concern, the “Sahm Rule Indicator” (which is used to forecast recessions in real time) was triggered, signalling a seemingly inevitable downturn. So, has Jay Powell screwed it?
The state of play
Everyone is freaking the hell out. Why? Let’s go chronologically
As of today (Tuesday, Aug 6th) the markets have largely regained the ground they lost over Monday. Why did they lose ground on Monday? Well, because overnight (remember the 12 hour time difference with the West) the Japanese market, called Nikkei, dropped by roughly 12%, the biggest single-day drop since 1987 (which itself triggered a severe economic downturn). The country’s currency, the yen, also appreciated significantly, gaining from around 161 to the dollar to 144.4 (as of yesterday).
Why did this happen? Well, blog post readers might be somewhat familiar with the Japanese situation. The TL;DR here is that Japan has been in a pretty profound slump for the better part of 40 years due to excessively tight monetary policy, which has in turn depressed inflation expectations and overly boosted savings rates. Likewise, a series of weird corporate governance policies have kept private investment low, and savings mostly translated to government debt. In recent years, inflation increased because of global issues and the post-COVID recovery, and the Bank of Japan seemed to keep rates low - which was rumored to be on the cusp of starting an actual round of wage increases to keep up with inflation. This would be good, since higher inflation expectations would help break Japan out of this prolonged “too much saving, too little spending” trap, and reactivate its sleepy economy.
However, last week the Bank of Japan decided to raise rates from 0% to 0.25% to combat inflation. This, of course, undid all their work to boost their credibility on inflation, which had previously resulted in the yen depreciating to record lows and the stock market to soar around 35%. A part of this phenomenon was known as the carry trade: basically, international traders would borrow in Japan at 0% rates and invest in US securities at a safe, 5.5% Fed Funds Rate, Critically, this trade only worked if rates in Japan were very low, and rates in the US were high.
On July 31st, the Fed met (as per its schedule) and decided, asthe markets expected, to not raise rates. On August 2nd, the monthly job market report for July came out, and it was really bad: unemployment increased to 4.3% (versus expected decrease to 4%), and job creation slowed to 114,000, versus expected 185,000. The decrease in job creation is pretty bad, but the change in unemployment is even worse: economist Claudia Sahm at the Federal Reserve had, a few years back, introduced an indicator to predict a recession. This indicator tracks the difference between the average unemployment rate of the previous 3 months, and the lowest unemployment rate recorded over the past year; if the difference is greater than 0.5%, the economy is likely to be in a recession already, or to enter a recession soon. Well, in July, the Sahm indicator grew to 0.53… meaning the US is bound for a significant slowdown? This put significant pressure on the Fed to cut rates, and especially to cut them a lot: last week, there was a 90% chance of a 0.25 b.p. rate cut, and this week, there’s a 25% chance - the other 75% is for 0.5 basis points, twice as high as previously expected. This is what led ot the unwinding of the carry trade (as in, rates going down on the profit end and up on the cost end) and to the selloff in Japanese markets, which spread around the world while people were also starting to worry about a US recession. A bunch of emerging country currencies depreciated a lot, and their assets took a big hit.
Prime cut
Let’s get a bit abstract for a second: what is a recession? Well, it’s two quarters of negative GDP growth in a row. Right? Well, no. Recessions are, largely, what happens when everyone wants to spend too little money - they all cut back, which means a bunch of people lose their jobs, which means they all cut back more, etc. While many people think that the job of recession-fighting is done by fiscal stimulus, it’s actually handled by the Central Bank (aka the Federal Reserve) for the very simple reason that they are in charge of the amount of money in the economy. If people are spending too little, they simply print more, and people go out and spend it. Recession over. Similarly, if the Fed thinks people are spending too much money, they print less, and people spend less, too. That’s how the whole “raising interest rates” thing tends to go - they move rates around so people spend more or less, and thus, they get the economy to be better or worse.
Well, in this case, they obviously have made it worse. Too worse, in fact, since inflation had previously been on the way down without too much fuzz. The July 3-month average annualized core CPI (so inflation without all the noisy stuff like energy and food, at the same timeframe as the Sahm indicator) was running at 1%, significantly below what the Fed wants inflation to be (2% - so double). How much worse? Well, the Fed doesn’t use just GDP or unemployment to define a recession (we’ve done this exhausting debate already), but rather, uses a series of broader indicators on output, spending, income, and employment. If you look at those indicators besides unemployment, the economy seems to be doing fine beyond the labor market. This is important because unemployment usually increases after a recession has begun in all the other data, not before. Additionally, and this isn’t cope because it’s coming from Claudia Sahm herself, this doesn’t mean a recession is already happen or even that it will happen - just that we have to be on the lookout for one. The focus on the Sahm Indicator is particularly telling because Claudia Sahm came up with it as a guide to policymakers at the Federal Reserve and the Treasury - meaning, there’s obviously still enough time to do something. In fact, it’s rather obvious that if the Fed had met after the jobs report, they would have cut rates now instead of punting it to September.
Why don’t they just meet again and cut rates now, then? Well, because they’re scared that if they do anything it would signal to the markets that the Fed is scared, and then the markets would get even more scared, resulting in some kind of weird chain reaction. So they wait six weeks to give the impression of stability and reliability - above all, they want to be seen as an institution that operates according to stable, predictable rules, so that people will trust them over the long term. This is all fine and good in principle, but that’s not how credibility works. The Fed gets credibility by achieving its targets (low stable inflation and full employment) consistently and predictably. They don’t get credibility by doing Rich Dad, Poor Dad grindset daily routines. Adding in a bunch of weird kabuki about how everyone will get scared if they meet now is nonsense - people are scared because bad things are happening.
Could the stock market pseudo crash cause the recession itself? Well, maybe. Asset prices reflect expected future profits - for stocks, they represent the profitability of the company. This means that if they expect the economy broadly to do badly, stocks broadly do badly too. But the causality is obviously backwards - bad economic expectations affect the stock market, generally. But good economic expectations affect stock prices too - which is why the stock market soared during early COVID; everyone expected it to be a sort of one-and-done recession with big policy action (fiscal and monetary). The impact that the stock market pseudo-crash could have is on balance sheets (basically, that people’s assets become worth less so they cut back on spending), but they didn’t, so all we have left are general “bad vibes” stuff leading to a recession after the stock market goes down. I wouldn’t be worried on this front.
We’ll work it out on the remix
Fundamentally, it seem’s that the Fed’s comportment makes no sense - why did they let the economy do badly in the first place? Are they stupid?
One possibility floated online is that Jerome Powell is a deep-cover Republican operative who is secretly working to get Trump back in the White House. This is, of course, not true and dumb as hell- during Trump’s presidency, he came under enormous pressure to keep rates low during the brief but now forgotten hiking cycle of 2018-2019. The FOMC, which makes decisions collectively, is mostly made up of Biden appointees at this point - the two Trump “holdovers” are the lone hawk (Christopher Waller) and middle-of-the-road type Michelle Bowman1. And historically, Trump’s other (successful) apointees were all doves - Richard Clarida and Randal Quarles were both on the “low rates” team, which is why Trump appointed them in the first place (he was a low rates guy!). In fact, back in 2022, the trendy thing to do was to say that Biden was the one pushing the Fed around, and that they kept rates low for so long because of political congeniality.
The answer is quite obvious but also quite disappointing: they’re making a mistake. The biggest misconception at play here is that the Fed does not make mistakes, but in fact it does, quite often in fact. Those mistakes are usually an overcompensation of the previous round of mistakes. All the way back in 2022, my take was that a recession was not necessary to control the (then very high) inflation. A very important part is why I thought the Fed made the mistake of letting inflation run too high:
… the Federal Reserve had trouble meeting its 2% mandate for basically the entirety of the 2000s, which was due to them keeping total spending (NGDP) too low and therefore have too many people without jobs. In brief, this caused the economy to take basically a decade to recover from the Great Recession. The good faith explanation for why the Fed did what it did is that they thought that acting too early on inflation risked repeating this mistake. This also explains their nearly incomprehensible new monetary policy framework, announced in 2020: Flexible Average Inflation Targeting (FAIT). FAIT was designed very simply: the Fed promises 2% inflation. If it gets too much inflation one year, there will be less inflation later, so that the average inflation remains at 2%. Meanwhile, if it gets too little, it will allow more inflation in the future, so the average returns to 2%. (…) One of the most criticized aspects was that they were going to wait until inflation was actually too high to raise rates, which seems inexplicable now looking back but is perfectly rational: the Fed worried about having to spike the football every time inflation (and the economy) were below capacity.
The problem here is quite obvious: the Fed messed up its inflation response in 2021/22, because it was concerned with not tanking the labor market recovery. Why was the Fed so concerned with tanking the labor market recovery in 2021? Because they had done so before, during, and after the Great Recession.
To sum up an extremely complicated topic, before the Recession the Fed was extremely concerned with inflation, so it kept rates extremely high for multiple years, leading to a complete collapse of the housing market, which dragged several banks down to the mud because of their high exposure to collateralized assets built off the back of subprime mortgages (these subprime borrowers, being lower income and higher risk, were disporportionately likely to default). One of these banks was Lehman Brothers, which collapsed in late 2007 - and in a meeting held the exact following day, the Federal Reserve kept rates stable at 2% and said their main concern was inflation, and refused to bail Lehman out. This led to a broad-based market panic where other banks collapsed, such that the Fed had to cut rates and bail out several other banks. The Fed was, indeed, extremely proactive in their response to the Recession proper, but fell short afterwards - they started believing that they either couldn’t or shouldn’t do more for unemployment, and instead began to argue that the economy suffered from long term supply-side labor market issues that they couldn’t address (it did not). The Fed was also concerned with inflation, which remained persistently below its (implicit) 2% target for most of the 2010s.
But why was the Fed so concerned with inflation that it caused the worst recession since the Great Depression and then struggled to see its recovery through? Because during the Great Inflation, the Fed had refused to take decisive action to curb inflation, resulting in extremely high levels of price growth paired with sluggish real sector performance. This is an even more complicated topic, but the summary is that the Fed had extremely bad and sparse data, and also had extremely bad models to interpret the data, leading to its conclusions being persistently wrong. But the Fed also didn’t believe, at times, that it could effect inflation - and these beliefs, as a whole, came from a desire to avert a second Great Depression. The Great Depression is an even longer and duller and more complicated subject, but to quote Ben Bernanke:
… as an official representative of the Federal Reserve. I would like to say to Milton [Friedman] and Anna [Scwhartz]: Regarding the Great Depression. You're right, we did it. We're very sorry. But thanks to you, we won't do it again.
Sleepy Joe and Dark Brandon
In March 2024, the Joe Biden reelection campaign was doing okay. The President was neck-and-neck with Donald Trump. With a strong economy and an unpopular opponent, “Sleepy Joe” seemed likely to pull away. The main concern of voters about him was his extremely advanced age (81 years old). This was assumed to be fixable: the president’s public appearances went well, and he could cast aside doubts about his physical fitness by vigorously campaigning or, say, a presidential debate. They pushed for an unprecedented June debate to help Biden finally lay the questions about his fitness to rest.
Needless to say, this is not what happened. He came off quite badly in the debate - medical speculation aside, Biden just seemed extremely old. In the weeks that followed the debate, the incumbent lost 2-3% in basically all polls - not a large number on its own, but bigger than the margins in any of the swing states in the previous two elections. He had seemingly already lost the election. After weeks of negative press, Biden officially withdrew from consideration for a second term and endorsed the Vice President on July 21st. This made him the first president to decline a second term since 1968, and the fourth in the entire 20th century (after Harry Truman in 1952 and Teddy Roosevelt in 1908).
Was taking a bold course of action a good idea? In international politics, switching out an unpopular leader for a (at least relatively more) popular one tends to, at worst, not change the party’s fortune, and at best improve it. Historically, it’s also been a good choice in the US: in 1968, President Lyndon Johnson’s replacement was the Vice President, Hubert Humphrey. Humphrey ultimately lost the election to Richard Nixon, but extremely narrowly: he finished behind by just 0.7% of the national popular vote, and the margins in swing states worth 128 electoral votes were under 2% - meaning, Humphrey could have easily won. This is especially true taking into account the unusually acrimonious Democratic Party Primary, or the fact that a third party candidate (the segregationist George Wallace) siphoned off a lot of the party’s white, unionized voters in the Midwest, and cost it several winnable states in the South.
Fundamentally, replacing Biden with VP Harris was a good idea in principle - she did as well as Biden in polls and was able to campaign effectively, which Biden clearly was not. And the gambit paid off - the Democrats had a 27% chance of winning in late July, under Biden; now they have a chance greater than 50%. And Trump’s campaign, by their own admission, was completely unprepared for the switch. Biden is a lifelong party man (for better or worse), and he came through for what the party’s leaders, members, donors, and supporters all wanted.
But why did Biden make this decision so late? Well, he had already gotten pushed out of the presidential race by party elites - in 2016. Alongside family issues (his son Beau had recently died and his wife Jill opposed a run), informal pressure from Obama and other party leaders nudged him out of the 2016 Democratic Primary in favor of Hillary Clinton - who lost to Trump. It’s fairly likely that Joe Biden would have won back then. Similarly, in 2020, the Democratic chattering class and top party donors considered Biden a non-viable candidate, and instead supported various other electable middle-of-the-road liberals (resulting in out-of-control proliferation). Biden ignored them and not only did he win the 2020 primary, he won the election handily. And back in 1988, Biden was forced to withdraw from the Democratic Primary after a media firestorm over plagiarism - the Democrats, once again, lost rather badly. In fact, his 52-year-long political career started by running a longshot campaign as a 29 year old city councilor of a city with under 5,000 residents against a popular incumbent who’d been in statewide office for 26 years.won the race by under 2,000 votes.
In Joe Biden’s mind, dropping out in 2024 was repeating the same mistake he’d made in 2016, and in 1988, and which he hadn’t made in 2020 and 1972. He saw himself as a workhorse who had one more race in him, while other people saw him as the horse headed to the glue factory - and twice over, he was right and they were wrong. So the third time, he was fighting the last two wars - and was about to lose, until he changed track and decided to make a bold decision.
Powell’s Bidenist grindset
The Fed and Joe Biden both have a chronic case of fighting-the-last-war-itis. The decisionmaking process is fairly similar: looking back and overindexing from serious historic mistakes. A second parallel is quite obvious: the information to make such a decision is just not very abundant and not very good.
Let’s start with Biden: the polls seemed not very good. In 2016, they famously called the election for Hillary Clinton, and yet she lost. In 2020, Biden led comfortably in every swing state, but ultimately he won rather narrowly. In 2022, the Republicans seemed poised to make large gains in the House and Senate, and yet the polls whiffed it. The election models were also quite bad - particularly the 538 model. And the 2024 polls were also, themselves, not very good: they showed the largest electoral realignment in over half a century, with young voters and nonwhite voters all moving 20-40 points to the right - for example, young men of color were identifying as Republican as Evangelical Christians. These all went back to normal after Biden dropped out of the election (with some large shifts among Hispanic and black voters). The Democrats, meanwhile, are doing gangbusters with swing voters and the elderly. There are some explanations for this (that the realignment was real or that the electorate suddenly stopped being polarized), but they defy both common sense and like, the observable reality out there. It’s rather obvious that “agedep” and “racedep” were a polling artifact and the polls were doing a really bad job capturing the state of the race (mostly due to biased non-response rates from Democrats). Biden’s team didn’t have their own polls, but such internal polling is also plagued by the same issues, and fundamentally, they still didn’t believe the results. A problem with the discourse is that there’s not some group of “super-pollsters” out there who gets 100% accurate results - the President is looking at the same numbers we are.
Now onto Powell: the Chairman doesn’t have more data that we do. He doesn’t really have better data either. The Fed is the largest employer of economists on the planet, which means they can analyze and reanalyze, and rake it through a million layers of analysis. The Fed also feeds said data into models, and those models are not very good either - the Fed’s model said there would be no inflation in 2021, a recession in 2022, and a recession in 2023. Of course, neither was true. And the economic indicators, as listed above, were a wash until August 2nd, a day after the Fed met and decided not to cut rates unless future economic data was significantly worse than it was on August 1st. But the Fed doesn’t have a more accurate CPI (well, it does: the core PCE), nor it has a more accurate unemployment rate. In fact, it tends to focus on economic indicators that are quite bad, like vacancies to jobs. And its models are usually subject to a lot of theoretical critiques that focus on fundamental macroeconomic questions - they’re just as right or as wrong as all of us idiots.
Conclusion
The month of July leaves us with one big lesson: trying to consistently win the last war leads to consistently losing the current one. Doing something drastic is usually a sign of concern, unless the situation and the stakes are actually drastic. Obviously there’s a chance that the economy doesn’t actually get worse in the third quarter of 2024 and then we’re all a bunch of Chicken Littles. That’s not something I’d bet on.
The other lesson, besides “make good decisions on a timely manner” is that Market Monetarism is unambiguously correct once more.
The problem are mainly the regional Fed presidents, who aren’t appointed by anyone in particular and have mostly historically been, to put it nicely, idiots.