What We’re Reading (Week Ending 27 October 2024) - 27 Oct 2024
Reading helps us learn about the world and it is a really important aspect of investing. The late Charlie Munger even went so far as to say that “I don’t think you can get to be a really good investor over a broad range without doing a massive amount of reading.” We (the co-founders of Compounder Fund) read widely across a range of topics, including investing, business, technology, and the world in general. We want to regularly share the best articles we’ve come across recently. Here they are (for the week ending 27 October 2024):
1. China’s Fiscal Policy Update – Leonid Mironov
Ministry of Finance top brass spoke at a press briefing, and outlined the extent of the Fiscal policy support they can offer *within the confines of the current budget*. However, while the steps laid out suggest a cautious and structured approach, notable gaps in specific figures leave room for market speculation…
…China is set to enhance its strategy for managing local government debt, which remains a critical issue. The central government will issue large-scale debt swaps, a move aimed at addressing the opaque “hidden debt” local authorities have accumulated off the books. Local governments still hold 2.3 trillion yuan in available funds, providing some breathing room to manage obligations in the final quarter of 2024. These steps aim to steady the debt situation, though the path forward will undoubtedly be closely watched…
…There is commentary out there to say that this is not new spending, I would counter with that yes, its not new per se, but its spending that would go in to this gap (authorised/unspent) but won’t anymore. So this is stimulative…
…With property markets showing persistent weakness, local governments now have the authority to deploy funds from special bonds to purchase unsold homes. These homes will be converted into subsidized housing—a dual-purpose measure to both alleviate property inventory and address housing affordability. It signals a nuanced, albeit gradual, approach to propping up the beleaguered real estate sector.
This is likely where most of that 2.3trn RMB mentioned in (1) will go. Again since the the property market is such a significant drag on the economy, this is reasonable…
…In line with recent People’s Bank of China (PBOC) directives, four major state-owned banks announced forthcoming cuts to existing mortgage rates. These rate reductions, effective from October 25, are part of broader efforts to ease financial pressures on households and further stimulate economic activity. Again given the sheer amount of total mortgages outstanding (38 trn RMB at the end of ‘23, see chart), this is significant. PBOC expects an effective cut of about 50pbs on average…
…Perhaps the most telling aspect of the press conference was what remained unsaid. There were no specifics on the magnitude of additional fiscal stimulus or further bond issuances. Additionally, there was no precise indication of how much the fiscal deficit might increase—a critical piece of information many market participants were hoping for…
…The Ministry of Finance’s approach at this juncture reflects a cautious yet deliberate strategy. While existing resources are being leveraged, and flexibility is maintained, major new initiatives have not yet been unveiled. All eyes now turn to the late October NPC meeting, where the prospect of more significant fiscal interventions could reshape the economic landscape for the year ahead.
2. The Bitter Lesson – Rich Sutton
The biggest lesson that can be read from 70 years of AI research is that general methods that leverage computation are ultimately the most effective, and by a large margin. The ultimate reason for this is Moore’s law, or rather its generalization of continued exponentially falling cost per unit of computation. Most AI research has been conducted as if the computation available to the agent were constant (in which case leveraging human knowledge would be one of the only ways to improve performance) but, over a slightly longer time than a typical research project, massively more computation inevitably becomes available. Seeking an improvement that makes a difference in the shorter term, researchers seek to leverage their human knowledge of the domain, but the only thing that matters in the long run is the leveraging of computation. These two need not run counter to each other, but in practice they tend to. Time spent on one is time not spent on the other. There are psychological commitments to investment in one approach or the other. And the human-knowledge approach tends to complicate methods in ways that make them less suited to taking advantage of general methods leveraging computation…
…We have to learn the bitter lesson that building in how we think we think does not work in the long run. The bitter lesson is based on the historical observations that 1) AI researchers have often tried to build knowledge into their agents, 2) this always helps in the short term, and is personally satisfying to the researcher, but 3) in the long run it plateaus and even inhibits further progress, and 4) breakthrough progress eventually arrives by an opposing approach based on scaling computation by search and learning. The eventual success is tinged with bitterness, and often incompletely digested, because it is success over a favored, human-centric approach.
One thing that should be learned from the bitter lesson is the great power of general purpose methods, of methods that continue to scale with increased computation even as the available computation becomes very great. The two methods that seem to scale arbitrarily in this way are search and learning.
The second general point to be learned from the bitter lesson is that the actual contents of minds are tremendously, irredeemably complex; we should stop trying to find simple ways to think about the contents of minds, such as simple ways to think about space, objects, multiple agents, or symmetries. All these are part of the arbitrary, intrinsically-complex, outside world. They are not what should be built in, as their complexity is endless; instead we should build in only the meta-methods that can find and capture this arbitrary complexity. Essential to these methods is that they can find good approximations, but the search for them should be by our methods, not by us. We want AI agents that can discover like we can, not which contain what we have discovered. Building in our discoveries only makes it harder to see how the discovering process can be done.
3. Austan Goolsbee Explains the Fed’s Big Rate Cut – Tracy Alloway, Joe Weisenthal, and Austan Goolsbee
Joe (12:30):
You mentioned lags. I want to ask you a question about that. When the Fed started jacking up rates aggressively, one of the theories for why it didn’t have a sharper impact on the economy is that so many households and corporations had in, say, 2020, first half of 2021, termed out their debt and so there was not a lot of sensitivity to debt.
The flip side of that now — and people have been writing about this — is that even though the Fed has now commenced a cutting cycle, that the weighted average cost of debt is probably going to rise in 2025 basically just mathematically, right? Because eventually that’ll have to be refi-ed at higher rates and so forth. How do you think about that dynamic now when you’re thinking about these lags? You’re starting a cutting cycle, but at the same time probably cost of debt is actually going to rise for a fair number of economic actors in this economy.
Austan (13:19):
You have remarked on this subject and thought it through. In my world that goes into the economic conditions and there are many things that have made this a hairy, strange time for central banks because the business cycle, both down and up, looked almost nothing like historical precedents. This is one aspect of that.
We’ve analyzed this specifically thinking about mortgages. Okay, so if I had told you, the premise of your question, I a hundred percent agree with, six years ago, if you said ‘The Fed is going to raise 500 basis points in a single year, what is going to happen?’ I think most all economists would say ‘Yikes, there’s going to be a major, major contraction and it’s going to be concentrated. Autos down the tubes. Consumer durables, bye-bye. Business fixed investment construction, all going to collapse because they’re very interest rate sensitive.’
We didn’t really see the economy go into the steepness of collapse that you would’ve expected. And so that brings us back to this question. It’s kind of a twofold. Is there something about this unusual business cycle that makes economic activity less sensitive to the interest rate? Or is there something strange about this moment that the lag effect is longer. And it can be both and they can run together, but in the case of mortgages, one of the things that has made monetary policy transmission less direct, is the fact that a vastly higher share of mortgages are 30-year fixed mortgages now, than they were in 2005, 2009, whenever you want to look at.
And so when they change the interest rate — in some countries virtually all mortgages are adjustable rate mortgages. So when their central bank raises rates, they bring out parents onto TV, ‘The central bank is killing us. You know, our mortgage payment went up.’
In the US, if everybody’s on a 30-year fixed, in a way that’s just a delay, but it’s a 30-year delay. So I do think that notion that there are companies that don’t have a lot of debt so they aren’t as especially sensitive to the interest rate, that the term structure of their debt may be such that the average rates they’re paying might even be higher as the Fed cuts. I think that’s not a problem, that’s just a fact and we just need to understand it and see what the magnitude is…
…Tracy (16:29):
Yeah, but my question is going to be ultra simplistic. Can you explain to us in excruciating detail what exactly you expect happens in the economy now, as you cut interest rates? How does that cut get transmitted?
Austan (16:45):
Oof. Okay, as a general matter, the Fed has only one tool really, which is a screwdriver that can tighten or can loosen and I always say if your problem is, you know, a loose fender, that’s great. If your problem is can you make breakfast? No, you kind of can’t do that with a screwdriver.
So the main channels of monetary policy impact on the economy, I think are on the real economy side and they are on interest rate sensitive parts of the economy — like consumer durables, business fixed, investment construction, things like that.
Now there are other channels of monetary transmission where there’s a lot of argument. How important are they and they are, well if you change the value of assets, like the value of housing, the value of stocks, etc., is there a wealth effect so that consumer spending might go up as the asset values go up. Or if you contract and asset values go down, would that limit spending?
There’s a dollar channel that if rates in the US are moving relative to how rates are moving in other places, can affect the currency and that could affect imports and exports.
Those are probably a lot of the main channels and it’s always in the counterfactual. What would be happening if we didn’t do this? So to the extent that there’s already a debt structure or to the extent that we went through a business cycle that for the first time ever was not driven by cyclical industries, but was driven by services because nobody could spend money on that, and services aren’t especially interest rates sensitive, that’s another reason why you might think the monetary transmission mechanism, which is actually a whole bunch of different transmission mechanisms, just looks different this time than before.
Now everything that looks different is not bad. Okay, in a way this is frustrating that monetary policy doesn’t have the same impact, but at the same time in 2023 we hit what I called the golden path. Inflation came down almost as much as it ever came down in a single year, and there was no recession. And that never happened before. And so the unusualness of this thing, sometimes it’s good!…
…Austan (21:27):
Yes, does not necessarily. I agree with [that]. So let me finish two thoughts. One, did the Fed have anything to do with it? That’s kind of the question. If it was all supply shocks, then the Fed didn’t really, yes, the Fed can’t be blamed for the inflation going up, but then the Fed shouldn’t take credit for it coming down.
There is some component that as supply shocks heal, you get immaculate disinflation. I do think that the fundamentally different thing that happened this time than the last time we were getting supply shocks, like at the end of the 70s, is that the market expectations of inflation basically never went up. In the 70s, as actual inflation went up, the expectations went up. And part of what made the Volcker experience so hard is you didn’t have to just slay the inflation dragon. You had to go convince everyone that we will hold this thing underwater for as long as it takes until it surrenders and that’s a brutal process.
I do think that expectations stayed — even as actual inflation was almost double digits — stayed exactly at PCE 2% as the inflation target said, was fundamentally the Fed making a promise it may look bad but we’re going to get it back, and that the market de facto believed it. And that is to the Fed, is about Fed credibility, and I do think it made a big difference…
…Tracy (38:43):
That was perfect. Can I ask one more serious question before we wind it down? But you talked about restrictiveness earlier in this conversation and I get where that comes from and people look at things like real yields and stuff.
But if you look at stock market prices, we’re recording this on October 9th, I think stock indices are at records again. If you look at credit spreads, those are at multi-year lows. Where’s the restrictiveness? Because I don’t see it in parts of the financial market, let me put it that way.
Austan (39:14):
I’d say two things. I told you, my focus is primarily on the real side of the economy. I think those are the biggest, most impactful parts of the monetary policy transmission mechanism, historically.
So I’m less of a fan of interpreting financial conditions indices as a measure of monetary restrictiveness or what monetary policy should do because, in my view, it’s got a major reflection problem that, let’s say the market, which is forward-looking, decides they think it’s going to work, that there will be a soft landing, that rates are going to come down because inflation has been tamed and is at 2%. Then equity markets go up, long rates would come down and that would then be interpreted as a loosening of financial conditions and it would be like ‘Oh, you better stop cutting, you better raise.’ But that’s just self-referential. So I think that’s a little problematic.
And the inverted yield curve, for two years, which everybody has been saying is an indication that there’s about to be a recession, that’s not normal. If we go back to a regularly-shaped yield curve like we’re in more normal conditions, that’s not the end of the world.
My view of restrictiveness is we set the Fed funds rate, we set it high and held it there for more than a year and as inflation came down, the real Fed Funds rate just kept going up, passive tightening. That’s the highest the real Fed Funds rate had been in decades. And so to me that’s where the restrictiveness is.
4. Investing lessons from a mini-Berkshire Hathaway – Chin Hui Leong
Gayner believes mistakes of omission are far more costly than mistakes of commission.
He shared a personal example of passing on investing in Berkshire Hathaway (A shares) in 1984 when he first discovered the company.
At the start of 1984, shares were trading at around US$1,300. By the time he got around to buying some shares, the stock price had risen to US$5,750. Hence, he missed out on a gain of over 340 per cent.
I’ll add a second lesson to his point.
Shares of Berkshire Hathaway (A shares) closed at nearly US$694,000 per share last Friday. In other words, even though Gayner did not invest earlier, his shares are worth about 120 times more than what he paid.
While his returns could have been over 530 times if he invested earlier, I don’t think anyone would lose a smile with a 120-fold return.
So, here’s my take: if you find a great company with a promising future, it may not be too late to invest, even if the stock has already appreciated…
…When selecting stocks to invest, Gayner looks for four key factors.
The first is about finding a profitable business with minimal to no debt and a good return on capital. The reason is clear; starting with this pool of stocks increases your chances of finding a winner.
Secondly, he wants to have a talented management team with integrity.
Gayner may have taken a leaf out of Buffett’s playbook here. As Buffett once said, without integrity, the other positive management qualities, will work against you.
Interestingly, Gayner also connected the use of debt with management’s character.
For him, debt is a character marker.
In a podcast recorded earlier this year, Gayner recalled the advice of Shelby Davis, another legendary investor and mentor. Davis pointed out that in the absence of knowledge about a new business, the use of debt can be telltale sign.
Simply said, if a business is entirely equity-financed, the management team will have no incentive to steal from their own funds.
To be sure, this does not mean that a debt-laden company is fraudulent.
However, Gayner argued that leverage creates conditions for a dishonest management team to exploit since the money does not belong to them.
5. A Message From the Past (Thoughts on Nostalgia) – Morgan Housel
I was recently asked at a conference how investors should feel about the stock market given that it’s basically gone straight up over the last 15 years.
My first thought was: you’re right. If you started investing 15 years ago and checked your account for the first time, you would gasp. You’ve made a fortune.
Then I thought, wait a minute. Straight up for the last 15 years? To echo my wife: What are you talking about?
Are we going to pretend like the 22% crash in the summer of 2011 never happened?
Are we supposed to forget that stocks plunged more than 20% in 2016, and again in 2018?
Are we – hello? – now pretending that the worst economic calamity since the Great Depression didn’t happen in 2020?
That Europe’s banking system nearly collapsed?
That wages were stagnant?
That America’s national debt was downgraded?
Are we now forgetting that at virtually every moment of the last 15 years, smart people argued that the market was overvalued, recession was near, hyperinflation was around the corner, the country was bankrupt, the numbers were manipulated, the dollar was worthless, on and on?
I think we forget these things because we now know how the story ends: the stock market went up a lot. If you held on tight, none of those past events mattered. So it’s easy to discount – even ignore – how they felt at the time. You think back and say, “That was so easy, money was free, the market went straight up.” Even if few people actually felt that way during the last 15 years.
So much of what matters in investing – this is true for a lot of things in life – is how you manage the psychology of uncertainty. The problem with looking back with hindsight is that nothing is uncertain. You think no one had anything to worry about, because most of what they were worrying about eventually came to pass.
“You should have been happy and calm, given where things ended up,” you say to your past self. But your past self had no idea where things would end up. Uncertainty dictates nearly everything in the current moment, but looking back we pretend it never existed.
Disclaimer: None of the information or analysis presented is intended to form the basis for any offer or recommendation. We currently have a vested interest in Markel (Tom Gayner is the CEO of Markel). Holdings are subject to change at any time.
