Research and Writing

A ridgeline is also my favorite way to describe central bank management of inflation. To the left you have deflation, and deflation begets deflation. Why buy something today when it will be cheaper tomorrow? And with that small leap of logic, spending and the economy grinds to a halt. You roll down the hill and it is hard to stop. Similarly, inflation begets inflation. It's easy for 3% to quietly become 4% and then 5% as small decisions and behaviors compound. At 2% you laminate your restaurant menus, at 4% your menu is a QR code that links to a webpage you can edit. Workers who feel the squeeze of their grocery bill demand higher wages, which businesses pass along to customers to the maximum degree they can get away with. Each incremental step higher makes the next incremental step higher just a little bit easier and more likely as inflation expectations get baked in. In that context, a 2% inflation target makes good sense for a credible central bank. Walk in the middle of the path - far enough from deflation without spilling into a world where inflation fears become self-fulfilling.

One thing I like about this metaphor is that it shines light on the weirdness of the "average inflation targeting" policy that the Fed has been touting since 2020 and that has become a pillar of the rationale for cutting interest rates today. If you were out for a ridgeline hike with a friend and he said "I walked too close to one edge for a while, now I'm going to walk too close to the other edge to average it out" you'd think he was having some kind of psychiatric break. When Powell says it, pundits nod solemnly and pretend it makes sense.

Core PCE - the Fed's preferred inflation gauge - was 2.6% in April, 2.8% a month later, and 2.9% now. The 1yr forward consumer expectation, according to the Fed's survey, is 3.4%. In other words, consumers expect things to keep slowly getting worse. Does that mean they are changing their behaviors accordingly? Will it get worse? I lean that way, but I don't know, neither do you, and neither does Powell. But I do know that if it does, the 60/40 portfolio is unlikely to preserve your purchasing power and you'll be grateful for any and all inflation protection you own.

So does this mean we are going to run out of copper? No, definitely not, this chart is ludicrous. It's ludicrous because commodity markets evolve and adapt, and this chart assumes the entire market will march off a cliff like a bunch of lemmings. Price always closes the gap - first by changing behavior, then by incentivizing new permanent capacity. When the market gets tight, the first responder is scrap. At one price, the old transformer in the yard is junk. At a higher price, it is feedstock. Motors get unwound. Cable is stripped. Tesla charging stations get vandalized. Scrap metal moves because the bid is finally worth the work. At the same time, buyers change behavior. Projects slip a quarter to the right. Engineers reopen drawings and trim copper usage. Aluminum takes more of the load where physics allows. Each decision looks small in isolation, but in aggregate it brings the market back into balance.

We think we have entered one of those periods where price has some heavy lifting to do. The catalyst is the disaster last month at the Grasberg mine in Indonesia. Seven people were killed in a mudslide that knocked a few percent off of global copper supply, supply that will probably not return to full production until 2027. We think this has created a moment where - barring the Chinese and/or global economy falling out of bed - the copper price will have to rise further to get things back into balance. We are positioned accordingly.

A reasonable question one might ask is, "why will it take so long for the production to come back?" The short answer is - it's a nightmare to even get there, never mind to get equipment there. The better answer is this fascinating (if a bit dated) documentary about why the construction of the mine took decades.

The huge table below reinforces the point by showing the returns of US stocks and bonds during and immediately after every government shutdown in the last 50 years. The most interesting thing about the table is how many rows it has - some version of a shutdown or shutdown bluff has happened on average roughly every 2 years. The average returns during those periods basically look like the average returns of non-shutdown periods, i.e. it doesn't matter. If you want the spreadsheet that generated this table for some ungodly reason, let me know and I'll give it to you. Otherwise I hope this email prevents you from spending another second on this topic and saves you some time and brain damage.

This question popped into my head again this week as I was stepping back on the big geopolitical forces that are evolving and reshaping the world. Gold has now passed the euro in terms of reserves at global central banks, and if current trends hold it may pass the US dollar before the end of the decade. So it seems that at least from a "store of value" perspective, the globe is voting with its wealth that gold is money. In reality, the USD itself has never passed the "store of value" criteria. The old-timey stories about going to the grocery store with a quarter and coming home with milk, bread, and a dime are proof enough of that.

USD financial assets have acted as the "store of value" - based on the belief that the US would steward the dollar responsibly enough that you could lend the government USD for a decade and end up on the other side with more purchasing power than when you started. The cracks in that belief are obvious and warranted.

Gold is not perfect either. It fails the "medium of exchange" and "unit of account criteria," because the real world stuff we buy like pizzas and baseball tickets are way too volatile when priced in gold to make any sense. Crypto has the same failings, on steroids.

What is the importance of these musings? Maybe nothing, maybe everything. Maybe it's just another way of getting to the same point I've made many times, which is that you should spread your wealth across different types of "money" because they all have warts. I do know that as investors it is all too easy to assume tomorrow will look like yesterday, and the evidence is building and building that this is currently unlikely. Heck, OpenAI has a disclosure on their website that says "it may be difficult to know what role money will play in a post-AGI world." Stepping back and thinking about this stuff is probably healthy, in doses.

I'll sign off with 3 of my favorite pages of writing ever. Milton Friedman succinctly describes the dynamics of "money" on a small Micronesian island, and then artfully shows how conceptually similar it is to many of our "more advanced" conventions. Enjoy.

During those days I would spend a lot of time at the Bellagio in Las Vegas. Their poker room was right next to the nightclub. When the club got out at 3 am, the drunk patrons would often sit down to play cards. They were well-dressed, while the poker pros were wearing sweatpants and basketball jerseys. The key was to "count the neckties" as the expression went - the more dapper people at the table you played at, the easier it was to win. But there was a tension - the smaller stakes games would get more clubgoers, while the highest stakes games would get fewer. So you could win smaller amounts of money more consistently, or try to win more but have to compete with better players and greatly increase the chances you would get smoked by someone better than you.

Little did I know how valuable that framework would be for my future career in investment management. Knowing who you have to beat in your markets - and their strengths and weaknesses - is a critical part of being a trader. Also, too many managers shun the "small stakes" games and instead let their AUM bloat to a point where they can only trade the biggest, hardest markets to beat. Some can still make money. Most cannot. We like the "small stakes" games quite a bit at Avos.

But there was a much bigger and more important lesson to be learned from those days in casinos - the casino always wins. They always win because the casino doesn't gamble, or in our analogy, actively trade. They set up a diversified portfolio of assets - roulette tables, slot machines, craps tables - each of which might lose money tonight but will make money in the long run. Casinos are the ultimate laboratory for understanding the prospect of disciplined buy and hold investors - in the long-run you win. The beauty of financial markets is you get to decide how much you want to gamble and how much you just want to be the casino, and it is ok to more or less do 100% of the latter.

At some point I realized that spending my days sitting in a casino getting fat off of casino food and lying to strangers wasn't the most fulfilling life. And, my first child was born, so it seemed like a good time to get a grown-up job. Bridgewater had been wooing me for a while so I succumbed to their persistence - with a big head start thanks to the Bellagio.

"Value" is one such imprecise word. Probably the most common usage is to mean "I think this thing will go up," which is actually a different concept. So people contort themselves to discern the "value" of things that don't have cash flows. Crypto and commodities are classic examples - bullish investors love to throw some squiggly lines on a chart showing that the gold/oil ratio or the gold/silver ratio or the gold/S&P500 ratio is out of whack and therefore something is good "value." Those charts are pure rubbish.

"Value" in investing is a much narrower concept. It means the current price is below what you get if you take the future cash flows and discount them to the present using a normal rate of return. There is nothing in the definition that makes a specific prediction about subsequent returns, though empirically you will find if you study the topic that "value" has virtually no predictive power in the short run but a pretty great track record in the long run.

Classically, the way value investing worked was a manager would find some under-the-radar company that was poorly understood by the market. They would do the analysis and decide the stock was undervalued. They would buy the stock and then yell from the mountaintops about how amazing the stock was, and other stockpickers would do their own analysis and if they agreed they would buy it too and the price would get pushed to fair value.

This isn't how things work anymore. The stockpicker cavalry doesn't exist to nearly the same degree, as it's been replaced by cheap index ETFs. So if you find a stock that is good "value," you often have to be willing to just own it and collect the outsized cash flows, instead of getting some kind of windfall when other investors notice what you noticed. Maybe you get lucky and it gets added to a big index, giving you some of that sweet passive investor honey, but you can't count on it. You need patience, more than most investors possess.

While I've given a precise definition of "value," I've glazed over the reality that calculating value is actually rife with nonsense. Here, the issue is that analysts decide what they want the output to be, and then retrofit the earnings growth rate to get them the answer they want. I would venture to guess that over 90% of Wall St. published price targets fit this bill, which is why you should aggressively ignore them. I leave you with my favorite case study of this: Tesla. If you try to value Tesla, you get nonsensical results. We ran it through our process and it spit out that by 2035, Tesla sales would have to be about ~25% of US GDP. In other words, all money in New York, California, and Texas spent by every entity on literally anything would be used to buy Tesla products. And yet...that doesn't stop Wall St from contorting themselves to make sense of the "value" of the company, instead of the common sense approach of saying "value has nothing to do with Tesla returns, the stock is just a call option on Elon Musk becoming a real-life Tony Stark."

The chart below is complicated but is also one of my all time favorites, so I'll explain it as carefully as I can. If you still are having trouble let me know and I will help you through it. The white line is the price of Tesla stock, on the left axis. It's a bumpy ride, but it goes up about 30% over this period. The colored lines are what Wall St was putting into their models for earnings per share for each future year, as of the date on the chart. So for example, the light blue line is their 2029 earnings per share estimate. In 2022, it was about $15/share. Now, it is $6.92/share. Each year follows a similar glide path down, culminating in my favorite line, 2025. In 2022, analyst models had 2025 EPS at about $7.50/share to make the price of Tesla stock make sense from a value perspective. Now we know the actual number is looking like $1.73. They missed by ~4.5x. This is what I mean when I say "value" and "I think this thing will go up" are different concepts!

I'll finish with some quick procedural points. These notes are public, you should forward them to anyone you want. If you think someone would appreciate being added, just let me know. And if you have a topic/article/question you'd like me to write about, also let me know. I hope these notes are helpful to people, and if you have a question others probably do too. - Josh