In early March, we interviewed Citi strategist Matt King on the role of central bank balance sheet activity in driving markets. His contention then was that, contrary to popular belief, balance sheets were expanding, and that we were already seeing monetary easing, causing markets to be “in thrall” to their activity. Since the recording, of course, we saw the blowup of Silicon Valley Bank, a significant expansion of the Fed's balance sheet, and (perhaps surprisingly) a rally in the stock market. In addition to the normal conversation, Matt brought along with him a bunch of charts, that you can see here, or by watching a video of the episode here on YouTube. The transcript has been lightly edited for clarity.
Key points from the pod:
Where are we now in the macro cycle? — 2:50
How are central bank balance sheets still expanding? — 6:50
How do balance sheet actions actually impact markets? — 9:23
Can market moves be explained simply by rate policy? — 12:41
What is the role of Chinese credit expansion? — 15:46
How tight is the US economy right now? — 21:57
Are long and variable lags still a thing? — 26:46
Is monetarism back? — 28:49
What should investors do? — 30:48
Tracy Alloway: (01:01)
Hello and welcome to another episode of the Odd Lots podcast. I'm Tracy Alloway.
Joe: (01:06)
And I'm Joe Weisenthal.
Tracy: (01:07)
Joe, it feels like it's been a pretty whiplashy start to the year. Yes. Yes. It felt like there was that moment in February where maybe there were signs that inflation was cooling. People were talking about a soft landing, and then just a few weeks later we're talking about inflation being entrenched. Maybe the Fed has to go even harder on the terminal rate. It just feels like everything changed so quickly.
Joe: (01:33)
Absolutely. I mean, I think even in January, it was still recession watch. It went from recession watch to soft landing, to no landing, and then to overheating fears again. Quite a lot of ambiguity for this short time into the year.
Tracy: (01:50)
When we have ambiguous macro environments, there's one man we like to turn to, an Odd Lots favorite. We need to talk to Matt King over at Citigroup.
Joe: (02:01)
Absolutely. Has anyone gotten the last few years right completely? No. But the last time we talked to Matt was in late 2021, and he said inflation isn't transitory. It's going to be hard. This isn't coming down anytime soon. And I think in early 2023, March 2023, people would say, yeah, that's pretty vindicated.
Tracy: (02:20)
I remember in that conversation he also talked about the possibility that the Fed might need to induce a recession to bring inflation down, which again, in late 2021 was not conventional wisdom.
So we need to check in with Matt, and I am very happy to say that we have him here with us right now. Matt, thank you so much for coming back on Odd Lots.
Matt King: (02:40)
Thank you very much for inviting me.
Tracy: (02:42)
So how would you characterize the current environment? Where are we in this sort of macro cycle?
Matt: (02:50)
I would say that markets are still in thrall to central bank liquidity to a much greater degree than is widely appreciated. And this is contributing to the uncertainty about the underlying economic outlook.
So, as I see it, the central puzzle is: How is it that with inflation proving stickier than many people imagined, and with central banks basically being hawkish on the back of that, and with yields and real yields rising, how is it that risk assets are doing so well? Most people would say, oh, it's because the economic data have surprised positively, and the economy is more resilient, and maybe we can have this soft landing or no landing or whatever.
Unfortunately, my work puts this in a rather different light. For me, the big factor which has contributed to the strength of risk assets beneath the surface is the way in which, even as the central banks have told us that they're going to be doing quantitative tightening (QT), actually, when you look at the details, they've ended up doing quantitative easing (QE). They've injected, over the last three months, a trillion dollars of liquidity. In my framework, that equates very directly into 10% directly on equities. The moment that you think of it in those terms, it just puts the whole outlook in a very different light.
Joe: (04:04)
Can you explain the mechanism by which central banks are still adding liquidity? We're in a historic hiking cycle, not just in the US but elsewhere. What's actually going on with this liquidity expansion?
Matt: (04:19)
The main thing I look at is reserves, or changes in reserves, on central bank balance sheets. The main mechanism at work here is the portfolio balance effect, or how much money we've given to the private sector in the form of reserves or deposits. It's reserves that correlate best relative to how many securities are available to absorb that. It's become really apparent that changes in reserves at the Fed in particular correlate much better with changes in risk than looking at securities.
Money growth has always been important for markets, but over the last decade, changes in money growth have overwhelmingly come from technical changes on central bank balance sheets. When the Fed or other central banks add or withdraw 4 or 500 billion of liquidity, sometimes even in a single week, there's nothing the private sector is doing on anything like that scale. So it has this outsized impact on markets.
Tracy: (05:18)
On that note, talk to us about where the liquidity has been coming from. Most people, when they think about central banks at the moment, are thinking about balance-sheet reduction. The Fed has said that it started QT, although there's a lot of disagreement about whether or not it practically has, and in other parts of the world, central banks have been raising rates and withdrawing liquidity. So, where is that excess coming from?
Matt: (05:46)
So, this gets quite geeky quite quickly, but I think it's odd. Most funding markets, roughly depending on when you measure it, have around a trillion dollars. About 250 billion comes from the Bank of Japan, about 450 billion, again depending on when we measure it, comes from the PBOC, and about 300 billion from the ECB.
In addition, the Fed was draining liquidity and reducing reserves last year. This year, even as QT has continued, and securities have been coming down, reserves have not been falling. So the Fed's contribution is technically zero. But as you say, that's surprising, when notionally they're doing QT. Each of these has its own story, and I'm probably more confident in the framework than I am in the outlook. But when you talk about a trillion dollars over three months, it just has this massive impact on markets.
Joe: (06:36)
How has the Fed been doing QT, and yet we haven't seen a decline in reserves in the US?
Matt: (06:47)
The way I tend to analyze all of this is empirically, what correlates best with markets. Specifically, what's been going on is the change in reserves. The Fed is influenced by not only the change in securities, what most people would think of as QT, but also the change in the Treasury General Account, where the US Treasury deposits money at the Fed, and the change in RRP, where money-market funds deposit money at the Fed. It's even reasonably intuitive as to why both of these have an impact. If, for example, the Treasury is issuing a lot more bills, and you take money from your bank account to buy those bills, but they don't spend the money in the real economy, paying employees or whatever, and they just lock the money away on the Fed balance sheet.
QT means the private sector has less money, and there are more securities needing to be absorbed in markets. Empirically, in periods where that's happening, like January and February last year and April and May last year, risk trades off as reserves fall, regardless of whether securities are going down.
Over the last six months, even as securities have continued to roll off, that impact has been offset by declines in the Treasury General Account and, to a lesser extent, by moves in RRP. That means that even though the Fed has notionally been tightening and reducing the balance sheet, it hasn't in terms of what matters for markets. Over the last few months, what's become relevant is to look at equivalent processes going on elsewhere.
Tracy: (08:56)
You mentioned the portfolio substitution effects, and I think when we talk about the impact of liquidity on markets, it's sort of an abstract thing. Could you explain to us in detail what is the process by which a liquidity injection, and I suppose it will depend on the form, gets transformed into a greater bid for risk assets?
Matt: (09:23)
I do all of this empirically, by looking at what correlates effectively with market moves. Everything I've observed over the last decade or so, where QE has dominated, suggests more or less the opposite mechanism from what you hear from the central banks. Bill Dudley had a lovely article on Bloomberg just the other day saying it's the level of reserves that matter, and all this stuff about money changes is irrelevant.
But everything I see from markets suggests the opposite: It's the changes that matter. The central banks tend to assume once they've announced it, it's in the price, but instead, it's only as liquidity hits the market or is withdrawn from markets that we see the market pricing it in. For me, it's all about this balance between how much money the private sector has relative to how many assets are available to absorb that money. When the Treasury or another private borrower borrows in markets, that creates some bonds or bills that somebody needs to buy.
Similarly, the central banks tend to assume once they've announced it, it's in the price. Instead, it's only as liquidity hits the market or is withdrawn from markets that we see the market pricing it in. The central banks always look for an impact on government bond yields and think in terms of reducing duration from bond markets. That's not what I think is going on at all.
For me, it's really about this balance between how much money the private sector has relative to how many assets are available to absorb that money. When the Treasury or another private borrower borrows in markets, that creates some bonds or bills that somebody needs to buy.
If they spend that money in the real economy, that kind of nets out, it's closer to being self-funding than people imagined. And in fact, that process of money creation as the system gains assets and liabilities is associated with risk-on QE. QE is kind of doubly powerful because it simultaneously gives the private sector more money in the form of reserves or bank deposits and deprives them of safe assets like T-bills or bonds to invest in, crowding investors into riskier assets as a result.
It's sort of unsatisfying in a way, because you can't see all these moving parts. But this helps to explain what is otherwise a significant puzzle, which is how come all of the action is mostly in Treasuries and government bonds around the world, but all these lovely charts point to strong relationships between QE and risk assets like equities and credit spreads.
Joe: (12:00)
I like this daisy chain. It's like, someone buys the bill buyer buys bonds. The bond buyer buys Treasuries, the Treasury buyer buys corporate, the corporate buyer buys junk. The junk buyer buys stock, and the stock buyer buys Dogecoin. That is like the little domino meme, right? It's like that slight change that some degen way out at the end of the chain is buying crypto.
Matt: (12:22)
Just since you've mentioned that, ironically the best correlations I find all are exactly with the most popular assets, like cryptocurrency or like Tesla stock, for example. It's just amazing how it shows up there in the hottest assets, even though the correlation applies more broadly, too.
Joe: (12:41)
Let me ask why there isn't a simpler answer to all of this.
I can look at the S&P, or the relationship between QQQ and the S&P, something more high-beta. In Q4 of 2022, we got a string of encouraging inflation prints that said it's finally happening, that the inflationary process has peaked. But starting in the middle of January, people started saying maybe it hasn't. We're seeing upward surprises in used-car prices and ongoing firmness in rents. Why is that real activity not a useful way to explain the trajectory of risk assets, and why inflation isn't coming down as we thought?
Matt: (13:43)
I agree that's probably part of the explanation, but if fundamentals were as strong as people traditionally think, my charts with central bank balance sheets shouldn't work at all, but instead, they work better. When we look at the moment, the economic surprises on the Citigroup Economic Surprise Indices are very positive. But the economic data changes are not. We've raised our growth forecast globally by 30 basis points, but it's still at one of the lowest levels, 2.2%, over the last 40 years. Is that really enough to make you super-excited?
Joe: (14:26)
To push back, it might be, if I thought two months ago that we were staring down the barrel of a hard landing.
Matt: (14:33)
Yes, true. But the explanation people normally come up with is that the central banks are being really dovish, and then you have Powell and the Fed saying, no, we're not being dovish, we're going to stay the course because the most important thing is inflation. You get a disconnect if you try to explain it in those terms. People are trying to retrofit fundamental explanations to price action that was actually driven by these technicals.
Tracy: (14:57)
So your contention is that the recent rally doesn't really have anything to do with what's been happening in the real economy, as evidenced by the collapse in private money. But if you look at what's going on with public money, i.e., central bank balance sheets and things like that, the correlation is much stronger.
Matt: (15:18)
That puts it slightly too strongly, but yes. Basically.
Tracy: (15:22)
Okay. So one thing I was wondering is if you look at China, China is currently a place that wants to stimulate, but seems to be having a hard time convincing private companies to actually go out and borrow. Can you talk a little bit more about what you're seeing there?
Matt: (15:46)
I think that's a very good description of it. The Chinese liquidity injection in December was so large that they've been a little bit late publishing the January number. But as you say, what we've seen is that total social financing, in particular, has been really quite disappointing. Even M2 growth has been a bit stronger, but has not surprised to the upside. And this is part of a broader story that has two legs. The first is that economies have been getting saturated with debt even as it's been cheap to borrow. Japan drove the world’s borrowing until 1990. Since then, they haven't done much. The US and Europe drove the world's borrowing until 2008. Since then, the private sector hasn't wanted to do much, and they're opting for share buybacks. That's where China has stepped in.
But even in China, it feels as though you're getting this saturation, where it's the state-run banks lending to state-owned enterprises rather than the private sector voluntarily wanting to borrow. The authorities are intervening, and providing support, and injecting liquidity to banks, but it's not that they want a new investment and real estate-driven boom in the same way as they've targeted in the past. Instead, they're trying to achieve the rebalancing towards the consumer that they always wanted.
As a result, you see relative disappointment in the broad credit metrics and in the credit impulse. The implications for things like commodities in the rest of the world are much less positive than they have been in previous investment-led booms. We still see the positive in things like China equities, but if what I care about is those credit numbers, it all feels a bit more half-hearted than it used to be. Even as there have been some narrow liquidity injections on the central bank balance sheet recently, it feels to us as though those have been a bit extraordinary, and it would be a mistake to extrapolate them through the rest of this year.
Joe: (17:59)
So this era of large central bank balance sheets really started in the wake of the Great Financial Crisis. All the central banks cut rates to zero and could not cut further, so they had to use balance sheet activities to compensate for their inability to cut rates any further. By and large, they didn't want to go negative, but the reversal of that in 2020, we saw some reversal of that in the mid-2010s with the rate hikes and the quantitative tightening.
What we're seeing now is the reversal of that. You've been talking about what's the trajectory as a balance sheet. What is the role of the tightening because, yes, it's true that in some cases central bank balance sheets still are growing, but we do know that they're all tightening, and that part has not really changed. What is the rate effect, and how does that affect markets or what we see in the real economy?
Matt: (18:49)
So, this is unclear, and I have a very different view from the central banks and traditional economists, and this comes back to what I was saying about flow versus level. And Mervyn King has been doing some nice talks for Citigroup clients, where he echoes the points that I'm making about the flows of money being important. So, in the central bank models, not only is inflation potentially self-reinforcing, but also the level of rates almost mechanically, without looking at flows of money growth, is thought to translate through into inflation. And never mind that over the last decade, until recently, that didn't seem to be happening again.
They don't have money growth or the financial markets more broadly. And therefore, again, it's the rate levels which for them are super important. And the way I think about it is instead, no, that low level of rates counts only if it drives someone to borrow. And even when it comes to things like unemployment, again, it's the changes that are actually more associated with recessions rather than the levels in themselves.
And while I'm open to the possibility that actually, there is now more momentum in the economy because of green investment, or very well some of the other things that people are speculating about as drivers of an increase in R* in general, I don't see that. What money growth I did see is often defensive stuff, like credit card borrowing, and seems now to be reducing, being killed off by the rises in rates. The bank lending surveys are all showing tightening.
And my general impression is that actually, while the M2 and the M3 numbers may exaggerate the tendency and the negativity, because to some extent those are influenced by QT in general, I'd say central banks and economists assume that there is momentum there, which would've been the case in the past when it was the private sector driving the money growth, because rates were too low and they had a great investment idea or whatever.
And this time around, while we had the biggest surge in money growth since the Second World War, it never came from the private sector. It came from the fiscal stimulus. It came from the QE. It had already been turned off, even before the rate hikes started. And all this, to my mind, points to the risk of overtightening. I'm not convinced that there is this super-strong momentum, which the central banks tend to assume, and yet they're in a really difficult place because the lags are so long and it becomes almost impossible to tell the difference between a two-year lag on inflation and then conversely up relative to money growth. And then conversely, ooh, a genuine decoupling, especially when you claim that the inflation was transitory to begin with, and then were disappointed that it took longer to go away than you thought.
Tracy: (21:37)
This was going to be my next question, on the long and variable lags, but talk to us, like, what evidence are you seeing right now of higher interest rates impacting not markets, but the real economy, and what are you looking for in terms of signs or evidence that they are, in fact, having an effect?
Matt: (21:57)
In many respects, this is hard, because those lags are long, and also because there's been so much terming out of debt in recent years that makes it kind of difficult to tell. So let me answer that a different way, because some of that I leave to our economists, and you're seeing weakness in the housing market, and then in the US, the housing market has strengthened. You're still getting weakness in other places. But let me maybe answer this a different way.
So, one of the puzzles of the last few cycles, in 2018 in particular, is that in general, it's taken lower and lower levels of real yields to kind of end each cycle until now. And each time, what caused the pivot was effectively dysfunction in financial markets threatening to feed through into the economy. And each time, there's been more debt in the system.
And maybe that's part of the explanation as to why it was a lower rate each time. In 2018 to 2019, in particular, it's not the case that anyone was running around saying, "Oh, I can't roll my corporate debt," or "Oh, I can't pay my mortgage." Instead, what you had was weakness in equities and the weakness in the housing market threatening to feed through into something broader. It was that, coupled with broader fears about deflation, which allowed the Fed to pivot relatively rapidly.
Now, this time around, we haven't had that to anything like the same extent. Again, we've seen this year in particular, last year we had an orderly sell-off in markets, this year we are rebounding. But you get this debate as to whether the recession is postponed or avoided entirely. While there are some signs of more ongoing momentum, in the US consumer in particular, perhaps than I had imagined previously, and my economists have had a better call on that, I'm still deeply suspicious that a lot of the exuberance in markets has come because of this stealth QE from the global central banks.
And then in addition, the lags are long before you see equity markets and house prices correct downwards, and the stock of accumulated savings begins to diminish. The main thing that would convince me that I'm wrong on all of this is if we saw a significant upturn in loan growth, and money growth numbers that looked resilient. However, that's not what I'm seeing.
When I speak to corporates, everyone is having difficulty recruiting workers in hotels and restaurants, in particular, and there is pent-up demand for things that weren't possible during lockdowns. But the question I keep coming back to is whether the corporates are saying, we need to build more hotels and restaurants again, and whether there is this longer-term demand. I'm not nearly as convinced as many people are that everything has turned around as much as people like. Maybe some of the backstory here, if I may, is the difference between how I think about it and how the central banks think about it.
Tracy: (24:59)
It's a hair chart, it's a Medusa chart, I love those.
Matt: (25:01)
Exactly, the hedgehogs or whatever, porcupines. Not only have the central banks been surprised by the inflation being higher than expected over the last couple of years, but for the preceding decade, they kept expecting more inflation and then there was less. So they're really having to scratch their heads and say, what has turned 180 degrees and caused our models to go wrong in the opposite direction?
Joe: (25:25)
This is one of the themes that I like going back to, this idea of the 2020s being the inverse 2010s. That hair chart, hedgehog, Medusa chart, what have you, is a good example. First, there was a decade of perennially over-optimistic, or too-high, inflation expectations. And now, we may have a decade of continuing to expect that inflation will come down sooner. It seems like part of this debate is the variability of the long and variable lag impact on inflation, and it feels like the Fed is under the belief that these lags are much shorter than they used to be. The instantaneous financial-market effects reflect the speech.
If Powell gives a speech, even if he doesn't raise rates, then it all reprices, and the actual rate rises are a mere formality after that. From your point of view, it sounds like you still have these long lags because of things like how companies termed out their debt, and eventually, they will have to roll them over, even if they haven't yet. When that rollover happens, there will be a kick-up in their interest costs, and that will create a burden on investment. So it sort of sounds like that's where the tension is, in your view. Simply put, the impact of all the rate hikes we saw in 2022 is still coming.
Matt: (26:46)
Basically, yes. Maybe I have a particular view here. I'm not an economist, I'm a strategist. For me, asset-price inflation and CPI inflation have always been two sides of the same coin. The central banks gave up on money growth in the '80s and '90s when they said, there's lots of money growth, but there's hardly any inflation. Our job is to control CPI inflation, so this money growth thing is useless, and let's stop using it. Or, in the Fed's case, even stop measuring some of the metrics they ran previously.
For me, and I think virtually anyone in financial markets, it's kind of obvious what went on. We had asset-price inflation instead. Even if you build quite crude models, where you put together asset-price inflation and CPI inflation, those correlations that break down with CPI inflation, they basically carry on.
And so for me, what we're seeing is not this drastic turnaround where something like globalization is shifting to de-globalization, and long and persistent inflation instead. For me, it's just these long time lags, and there's a very clear pattern. The surge in money growth showed up first in asset-price inflation, then in goods-price inflation, and now in services inflation. And yes, in things like wage growth, but the lags are long enough that it's really difficult to tell whether this is genuinely persistent.
Joe: (28:08)
Let me ask you the devil's-advocate question. Could the correlation go the other direction, where the surge in asset prices creates the surge in monetary aggregates?
There are models of the economy where managers and bank lenders look at the price and are more likely to make a mortgage loan if they feel like this is an era where house prices are going up, or approve a business loan if this is an era where stock prices are going up, and the company is likely to be able to tap the equity market. Could it be that some of these charts, which do seem to show a compelling relationship, go from assets first to money supply next?
Matt: (28:49)
You are certainly right that the relationship often works both ways. It's not only that credit growth stimulates the housing market, it's also that a buoyant housing market encourages more credit growth. But if it only worked that way, then this chart shouldn't work. I shouldn't be able to find a relationship going back to the early 1900s where the money growth in the US links through to real estate, but with about a one-and-a-half-year lag. So yes, you are right. That's part of it.
But for me, money growth is still the best driver. Regarding lags, it's reasonably short when I look at things like the equity market, especially now that central banks are driving it. But the link to real estate is about one and a half years.
The link to commodity prices is about one and a half years. The link to CPI is harder to tell because the relationship is weaker, but as far as I can see, it's something like a two-year lag. That puts the Fed in a terribly difficult spot, because you're not going to see the impact of even the first rate hikes until late April 2024, never mind the hikes that you're doing at the moment.
Tracy: (30:07)
Matt, you emphasized that you are a strategist and not an economist. On this podcast, you are famous for the "flows before pros" idea - this idea that, for many years post-financial crisis, it made sense to just follow the money, and never mind whether valuations were reasonable or not. If we assume that liquidity does have a big impact on markets, which you argue it does, and if it does seem like all these one-off sort of stealth liquidity injections are now going away, what should investors do here? Should they flee en masse, or what would be your recommendation?
Matt: (30:48)
Actually, the outlook for the various liquidity factors is complicated for all of them. The surge feels as though it's been extraordinary. There is a lot of debate as to just how negative, or at least less-positive, they all become. On balance, though, what I think we have seen is an extraordinary three months that has left equity, and especially riskier credit, valuations at levels where I don't like chasing them at this point, especially for the more expensive equities, which are still the tech sector and the growth sector, and still the US relative to the likes of Europe.
If we want risk-on positions, we would tend to do them through currencies — euro versus dollar — or through regional preferences. European versus US equities, or maybe the same thing about China. Yes, you are right. The biggest problem that we see generally is that, for every individual asset class, that considered in isolation might seem attractive, what really matters is the valuation relative to money-market funds, especially in dollars. That argues for significantly-increased allocations to cash and cash equivalents. Exactly, those things which were basically uninvestable over the last decade.
Tracy: (32:19)
Matt, we're going to have to leave it there, but it was fantastic having you on the show once again. Really appreciate it.
Matt: (32:25)
Thanks.
Tracy: (32:39)
Joe, you know what I just realized?
Joe: (32:57)
Tell me.
Tracy: (33:00)
The last time we spoke to Matt, I think we ended the discussion by saying that we wished we had a video product because during the conversation, Matt was bringing up all these different charts and showing them. Now we're finally able to do it and show off some of these.
Joe: (33:21)
So, if you just listened to this episode on Apple or Spotify or something like that, go to YouTube, where we have the charts that he was bringing up during our conversation. We're going to have a video, and we're also going to write a post with some of the charts, or as many of the charts as possible. You can read a story about this with all the charts because I love the way talking to Matt, how he brings up all his charts in real time. It's very fun.
Tracy: (33:50)
Yeah, and the charts are excellent, especially the hair charts, which I can never get enough of. But I do think he hits on something, you know, this overall feeling in the market at the moment, which is, it does feel a little uncomfortable that we still have this overarching question of, is inflation coming down? Are central banks going to have to go harder? People are talking about terminal rates at, like, 6.5% now, which seems extreme, and you would think that would have more of an impact on asset prices.
Joe: (34:43)
It is really striking. The fundamental story still seems like it explains something, especially if we set aside what markets have done in 2023. As you talked about in the intro, a lot of people have suddenly gotten anxious about the overheating. On the other hand, one of the charts he showed during the conversation specifically had the title of "monetarism has gone out of fashion, and it's coming back," and it's 100% true. The out-of-fashion part has been especially in the 2010s. No one was talking like M2 and all that stuff. Does this become, once again, a sort of like the '80s, like the vulgar era, where these monetary aggregates become a central focus for how investors view the economy?
Tracy: (35:05)
Bringing back M2? Let's do it. It is true, though, that you can have both things. You can have technicals that are a tailwind for risk assets, and also have fundamentals that, so far because of the long and variable lag that Matt was also laying out, look pretty strong.
Joe: (35:22)
Those charts look pretty good to me.
Tracy: (35:24)
All right. Shall we leave it there?
Joe: (35:26)
Let's leave it there.