Buying Rembrandts
MUNGER: “To some extent, stocks sell like Rembrandts. They don’t sell based on the value people are going to get by looking at the picture. They sell based on the fact Rembrandts have gone up in value in the past. When you get that kind of valuation into stocks, some crazy things can happen. Bonds are way more rational because nobody can believe that a bond paying a fixed rate of modest interest can go to the sky, but with stocks, they behave partly like Rembrandts.
Suppose you filled every pension fund in America with nothing with Rembrandts. And of course Rembrandts would keep going up and up as people bought more and more Rembrandts or pieces of Rembrandts at higher and higher prices. Wouldn’t that create a hell of mess after twenty years of buying Rembrandts? And to the extent that stock prices generally become sort of irrational, isn’t it like filling half the pension funds with Rembrandts?
BUFFETT: Once it gets going though, people have an enormous interest in pushing Rembrandts. It creates its own constituency.
-- 2001 Berkshire Hathaway Annual Meeting [Source]
The annoying thing about investing is that past performance really doesn’t tell you anything about future returns.
Some assets, like Berkshire Hathaway stock, seem to be perpetually undervalued - they outperformed in the past, but they keep outperforming. Same for companies like Google and Netflix. This is the theory behind momentum investing.
Then you have assets like Treasury bonds, where high performance in the past means you are guaranteed to have lower returns in the future. The principal interest payments are fixed, so any price appreciation will come out of future return.1
No one bothers to label which is which, and a lot of assets are a mix of the two, or neither. It really is quite annoying.
This would be fine, except that our brains are wired to respond to feedback. If you get praised for doing good work, you will keep doing good work, and you will get paid more. If you touch a hot stove, it will hurt, and you will stop touching hot stoves, and you will get burned less often.
No doubt our ancestors who were highly responsive to feedback had a better chance of surviving - if you consume a poisonous mushroom and get sick, you are probably more likely to pass on your genes if at that point you stop eating poisonous mushrooms.
Then you come to the world of investing, where everything is backwards. Lots of stocks skyrocket one year, and then collapse the next. Others crash one year, and then take off the next. We can’t help but evaluate an asset in part based on how the price moved in the past.
Nothing about an asset price chart for the past year, three years, or even five years, tells you anything about whether the asset will perform well in the future, and you can prove this by revisiting some stock price charts from the 1990s dotcom bubble or 2000s housing boom.
You wouldn’t know this by the people taking victory laps today on Twitter. It’s hard to have the discipline to say, “We’ll see”, but one has to try to overcome one’s hardwired instincts.
As Charlie Munger said, bonds are more reasonable. Most people understand that if their long term bonds appreciated last year, that means that interest rates are now lower and they will now make less money if they keep buying bonds. No one bids bonds to the moon in the hope of selling it to a greater fool later on - everyone knows that in the best case scenario they will turn to cash on a set date in the relatively near future, and the logic works backwards from there.
Other assets are a little bit harder to analyze. It is often said that assets are intrinsically worth the sum of the future cash flow they produce, which is true by definition, but perhaps not always useful. Many assets are like Rembrandts, that produce no cash flow each year - the only cash flow you get is a function of supply and demand on the future date you sell it, and you make money by predicting that demand will be higher in the future, or supply will be lower.2
Here is a fun example from Bloomberg of an unusual way to make money from stocks [source]:
“What I used to say is, the flops are worth more dead than alive,” says Kerstein, a dealer in Virginia whose stocks and bonds can be found on Scripophily.com. “Like the Enrons. They were selling for a dollar or two apiece. Once the company died, they were selling for a couple hundred.”
In the past few years, there’s been fresh interest in the equities and bonds of Donald Trump’s failed casino company. A 1999 Trump Hotels & Casino Resorts Inc. stock certificate that’s in “practically pristine” condition, featuring a head-and-shoulders portrait of the Donald hovering over the Atlantic City boardwalk, is available on EBay at a buy-it-now price of $800. The stock itself was delisted after trading for about a buck or two following the company’s first trip to bankruptcy court in 2004.
If you bought a share or two of Enron or Trump Hotels, and had the foresight to order the certificate (no longer easily done), you might have turned a profit off of an investment in a bankrupt company. Sometimes people really do like looking at the picture.
The Federal Reserve estimates that US households have $124 trillion in net worth, mostly in the form of productive assets - that is, businesses and real estate. We describe income and wealth in terms of dollars, but that is merely an abstraction; we are really interested only in the goods and services that those dollars buy, such as food, shelter, and wifi. Generally, the more goods and services we have, the higher our quality of life.
For simplicity, let’s say we only consume two goods: coffee and socks. To produce anything, we need capital and labor.
In the case of coffee, we need people to pick the beans, operate the roaster, serve the coffee, and so on. We also need capital - the coffee plantation, the roaster, the espresso machine, the cafe and the land it sits on. To produce a lot of coffee, we need a lot of people, but also a lot of cafes and plantations and machinery. The same goes for socks - we need textile mills and packaging plants and warehouses and delivery trucks, and people to operate them.
If what we really care about is how much coffee and how many socks we can produce, then we also have to worry about making the productive assets we need to produce the coffee and socks. To do this, we convince everyone to consume less coffee and socks today so that we have the spare labor to build cafes and coffee roasting machines and textile mills, and we compensate the people who consume less today with pieces of paper which entitle them to more coffee and socks tomorrow.
In our economy, we attach these pieces of paper to the machines and cafes; if you sacrifice consumption today to build a cafe, you get so many cups of coffee a year in the future, based on how efficient the cafe is.
We have actually gone a step further, and we have engineered a lot of different kinds of pieces of paper with different levels of claims on productive assets, according to the different preferences of savers - we have bank deposits, and bonds, and stocks - but they are mostly ultimately backed by the productive assets that drive our economy, like houses, factories, laboratories, and software.
This $124 trillion of assets are a key input into the $21 trillion of annual output of the US economy, and their owners are entitled to a significant share of that output - historically, around 30% of total output, so about $6 trillion a year.
There are non-income producing assets, like art or gold, that may be nice to look at, and may bring joy to their owners, but in aggregate they are miniscule next to the giant pile of productive, income3 producing assets that drive the economy. Normally, non-income producing assets are barely even noticed. It is estimated that all the gold that has ever been mined in the world is worth only about $10 trillion at current market prices, a drop in the bucket next to the hundreds of trillions of dollars of productive income producing assets that exist globally.
Historically, it has been hard for non-income producing assets to keep up with income producing assets over a long period of time. An ounce of gold was worth $20.67 back when the US was on the gold standard, before the 1930s; that same ounce of gold is worth $1,772 today, 86 times your original investment, a respectable return. If you took that same $20.67 and put it in the S&P 500 in 1926 (the earliest date for that index), and reinvested dividends, you would have $225,905 today, or 10,929 times your original investment. This is through the Depression, the wars, and everything else.
In some version of the world with an efficient market, we should expect everything to return about the same amount over time. We can speculate on why productive assets historically return more - and I think one aspect of it is that ownership of productive assets in the US gives you the part of the 30% of GDP that gets allocated to owners of capital in this country - but I think it’s generally good for society that people can make so much more money in a game that is not zero sum. The more we invest today in research and development, housing, and factories today, the higher our living standards will be tomorrow. As Keynes put it:
The social object of skilled investment should be to defeat the dark forces of time and ignorance which envelop our future. The actual, private object of the most skilled investment to-day is “to beat the gun”, as the Americans so well express it, to outwit the crowd, and to pass the bad, or depreciating, half-crown to the other fellow.
In our example of a cafe, if you want to build a cafe today, you need to convince savers to fund it. You usually do so by issuing promises to pay savers a profit later, out of the future profits of your cafe. You can go to the bank and get a loan to help pay for it - you sign a contract that says you will pay the bank so many dollars in the future, and then the bank funds this by turning around and trying to attract the money from savers, writing their own contracts to borrow the dollars from them in the form of bank deposits.4 Or you can sell bonds directly to savers, cutting out the bank.
Usually, as a term of getting a loan or selling a bond, you will have to promise to contribute some money of your own to build your project, to limit lender risk and align incentives. In return, you get equity ownership, meaning you get all of the profits from your cafe, once you’ve paid your lenders. The profits from your equity ownership will be much more volatile and uncertain, and they will extend far into the future if you are successful, long after you have paid off all of your bonds and loans.
You can see how equity, particularly for younger companies, can take on a lot of Rembrandt-like characteristics. The future cash generated is very uncertain and very far in the future. People can trade equities like Rembrandts5 for a long time, without any economic consequences. For a lot of equities, people will focus on how much it went up in the past and how much cash you might get from flipping it tomorrow to another buyer, rather than the dividends from the business itself that may be twenty years off.
You can see that equity ownership also actually has a lot of bond-like characteristics, especially equity in a mature, stable and profitable asset. You lay out cash to buy a stock today, and you will get cash later. If you are good enough at predicting how much cash you will get later, you can make a lot of cash by buying it for less today.
On a long enough timeline, you don’t ever have to worry about selling a stock to another buyer; you can own it and capture enough cash distributions to pay off your investment, and then some. This is the business that Buffett and Munger are in; they buy an undervalued business, and then reinvest the cash flow into more undervalued businesses, and so on until they have $500 billion. But there is nothing magical about owning equities - the return on any asset is a function of what you paid for it, and the trick is to find equities that will actually throw off enough cash in the future to pay off your investment.
Returning to our quote from 2001, there is a key bit of historical context to understand here. At the time, Buffett estimated that based on the valuations prevailing at the time, stocks would only earn around 6%-7% per annum over the long term. (It is much easier to make these estimates for equities as a whole than it is for any specific stock - Pepsi’s loss may be Coke’s gain, but overall, corporate America will capture about the same share of GDP each year.) Over the twenty years that have elapsed since then, stocks have earned about 7% per year, so his estimate was about right.
The thing is, at the time, high grade corporate bonds yielded close to 8%. So, as a pension fund, it would seem that it would be foolish to put a large portion of your assets in stocks earning 7%, which are generally riskier and more volatile than bonds earning 8% backed by a senior position in the same corporate assets. Less return and more risk is not usually the combination you are going for as an investor.
Believe it or not, back in these days, city and state pension funds would actually sometimes borrow money on those terms to invest in stocks. This ended about as badly as you would imagine; New Jersey borrowed $2.8 billion in 1997 at 7.64%, and by 2002 they were already massively underwater.
(This is not the dumbest investing decision made during that era involving borrowed money. In 2000, the New York Mets owed aging slugger Bobby Bonilla $5.9 million, and had the brilliant idea that they would defer the money at 8% interest, paying it in installments between 2011 and 2035, largely because Mets ownership6 believed they were earning a higher rate on their money through their investments with Bernie Madoff. Today, Bernie Madoff is in jail, the Mets are under new ownership, and throughout baseball, every July 1 is known as Bobby Bonilla day, where he cashes another $1.2 million check toward his impossibly lucrative total payout of $30 million - a testament to the power of compound interest.)
At the time, predicting 6%-7% forward annual returns from stocks was a little controversial. Stocks had been returning 17% per year for the prior two decades. It is not unusual to open up an annual report from the time to see a corporate pension fund projecting 10% annual returns from the whole portfolio, not just the stock portion. Said Buffett in 1999 [Source]:
Today, staring fixedly back at the road they just traveled, most investors have rosy expectations. A Paine Webber and Gallup Organization survey released in July shows that the least experienced investors--those who have invested for less than five years--expect annual returns over the next ten years of 22.6%. Even those who have invested for more than 20 years are expecting 12.9%.
But as Buffett and Munger observed, a significant portion of these apparent returns were coming from a market where everyone was treating stocks like Rembrandts, buying them from each other at ever increasing prices without regard for cash flows. It turned out they were indeed stealing returns from the future (actual annual returns for the S&P from 2000-2009: -1.0%), and corporations and governments (and in some cases pensioners) were forced to fund the return shortfalls that followed.
The three year bear market that ensued from 2000-2002 would chop 49% from the S&P 500. It is incredible to think now that a $20 trillion dollar market could get so out of whack for so long - it makes the $1 trillion of bitcoin outstanding today or the small cap meme stock of the week seem almost irrelevant by comparison.
Defining everything in terms of dollars is a useful abstraction, as it allows us to easily compare any investment or consumption option at any moment in time. We can define a cup of coffee or a share of Tesla or a bitcoin in one uniform measure. That ease of comparison can do a bit of mischief, though, as we start to convince ourselves that a security that confers ownership in a productive asset and a painting are substantially the same thing.
There is value in digging that extra level below the abstraction, and really understanding an asset. Art and gold have value, but non-income producing assets are a small market and are at a fundamental disadvantage when compared to productive assets, which are crucial to the functioning of the economy and come with the right to trillions of dollars of income every year.
Even within productive assets, we saw last week that some housing is only partially a productive asset, and partially a medallion which is an anti-productive asset, conferring income on its owner by restricting economic investment and gathering government subsidies. It shows up on the balance sheet as an asset, but from a societal view it is a liability, and it is yet another angle that has to be underwritten. There are always additional layers we can dig into to better understand an investment.
Liberty on Twitter noted that Bill Gates stated on Clubhouse this week:
I haven't chosen to invest in [bitcoin]... I buy vaccines, I invest in malaria medicine, in companies that make things.. I don't buy something just thinking someone else will pay more for it later."
I couldn’t have said it better myself.
For example, if you have a 10 year, 2% Treasury bond, and it trades up to 120% of par, that means your return in the past was 20%, but your return in the future will now be 0%.
As Keynes famously put it in Chapter 12 of his General Theory, “Professional investment may be likened to those newspaper competitions in which the competitors have to pick out the six prettiest faces from a hundred photographs, the prize being awarded to the competitor whose choice most nearly corresponds to the average preferences of the competitors as a whole; so that each competitor has to pick, not those faces which he himself finds prettiest, but those which he thinks likeliest to catch the fancy of the other competitors, all of whom are looking at the problem from the same point of view.”
Many financial assets actually have negative income these days, particularly once you adjust for inflation. That is a topic for another essay.
This is not exactly the way it works in real life, but it is close enough for this story.
Cynics on Wall Street will often refer to a certain type of worthless speculative asset as a “trading sardine”, in reference to a classic story. I think a Rembrandt is a much better description for this particular phenomenon - a Rembrandt has value, but return is totally dependent on supply and demand rather than cash flow. The sardine story, as told by Seth Klarman (in a book that has itself ironically become a Rembrandt or trading sardine):
“There is the old story about the market craze in sardine trading when the sardines disappeared from their traditional waters in Monterey, California. The commodity traders bid them up and the price of a can of sardines soared. One day a buyer decided to treat himself to an expensive meal and actually opened a can and started eating. He immediately became ill and told the seller the sardines were no good. The seller said, ‘You don’t understand. These are not eating sardines, they are trading sardines.’”
The owner of the Mets at the time, Fred Wilpon, was in his youth a lefty pitcher at his high school in Brooklyn, during which time he recruited a friend to play first base on his high school team - a basketball star named Sandy Koufax. Koufax went to the University of Cincinnati on a basketball scholarship, but in his senior year he played a bit of baseball as well, this time as a pitcher instead of a first baseman. He then went home to the Brooklyn Dodgers and would go on to be perhaps the greatest lefty pitcher in baseball history, while Wilpon was building his fortune in commercial real estate. They would stay friends, and Wilpon got Koufax into business with Madoff as well, an act for which Koufax later forgave him.