I am going to talk about two different ideas. Floats. And Moats.
Over the last few months, I have been obsessed by these two ideas.
I have spent a lot of time researching them and thinking about them and also about they might be related.
It’s been a fascinating journey so far.
This talk is the story of that journey.
The story starts in 1964. Warren Buffett is a young, dynamic investment manager. He makes his first big bet.
Between 1964 and 1966, he buys 5% of American Express. That’s 40% of his assets under management.
And Amex is embroiled in a scandal involving of things, salad oil.
This guy can’t get credit from a bank because he is a convicted fraudster. So he comes up with a neat plan.
Amex is a prosperous company with a stellar reputation. It also has a subsidiary which owns and rents warehouses.
Don’t ask why in the world is Amex in the warehousing business. Anyway, Antonio De Angelis goes to Amex warehouse and deposits tank loads of sea water. Except that he tells the warehouse that those tanks contain salad oil. No one checks out this guy or the tanks. I guess there were issues with KYC even back in 1964.
The warehouse takes the sea water and issues a warehouse receipt certifying that it has in storage a large amount of vegetable oil. The receipt carries the name of American Express, which is a name that stands for trust. An elated De Angelis takes the warehouse receipt to a bank and offers it as collateral and gets a loan, whereupon he goes gambling in futures and options. As you’d expect, he loses, and promptly goes bankrupt. Now the bank has a warehouse receipt as collateral on what it thinks is valuable salad oil, except that it’s sea water.
Shit hits the ceiling and Amex discovers that “it has a problem subsidiary.” The extent of the problem? About $150 mil. That’s a very large sum of money in 1964. Amex’s Warehouse sub files for bankruptcy but for Amex trust is everything. Its CEO says that Amex has a moral obligation to pay the bank even though its not legally obliged. So much for ring fencing using limited liability companies as subsidiaries. Recall this is what happened in the case of Tata Corus.
The market gets spooked. The stock drops from 60 to 35.
Ok, now let’s look at the magnitude of the problem.
Cash is $263 Mil. Securities are $515 Mil. Some are quite liquid. The loss on account of the scandal is $150 mil. So, what’s the problem?
The half a billion dollar problem is that of the outstanding TCs. Those TCs are are cashable on demand. What if there’s a run on Amex? The consequences could be devastating.
Death spiral is the word that comes to mind. Amex has not missed a dividend in 94 years and suddenly market feels that risk of insolvency is high.
Warren Buffett disagrees with the market. We know that because he goes out and puts $13 million into the stock for a 5% stake in Amex. That’s 40% of his partnership’s money.
What was he thinking?
Let’s speculate on that.
First, imagine that instead of T/Cs, Amex had bank debt outstanding. Would Buffett have invested?
I think not. Why? Too risky.
But there’s is no Bank debt! T/C outstanding: $526 Mil.
Risk of run on Amex. Consequences of a run are severe but what’s the probability? Buffett finds out by going shopping. He finds out that customers don’t care. The tarnish on Wall Street has not spread to Main Street.
He buys the stock.
By 1968, he has sold his sake which cost him $13 mil for $33 mil.
When Buffett correctly figured that trust behind that half a billion dollar promise is not evaporating, he gets convinced about buying into Amex.
He also discovers something else. That AMEX T/Cs represent an unusually attractive form of financing. To see how, let’s compare this with traditional debt.
Plain vanilla debt is onerous for three reasons. Buffett has discovered that Amex’s half a billion dollar liability represented by T/Cs have none of these onerous terms.
When Amex took this money, it simply issued a piece of paper without giving any collateral. So there’s no collateral.
There is no interest either. The paper is redeemable at demand, but there is a lag between issue and encashment and sometimes people don’t en-cash them. More importantly, even if some people encase them, there are others who buy new ones, so the balance in the liability account has become a “revolving fund.”
General point. When we look at a consolidated liability account or for that matter any account which consists of balances from a large number of accountholders, then what matters is not what the individual account holders are doing, but what the account balance is doing. Even if such a liability is classified as a “current liability” it could indeed be a perpetual one, if you are sure that the account balance won’t shrink.
I know this sounds like a ponzi scheme and it IS a ponzi scheme but without the derogatory connotation that goes with the world Ponzi.
See the full PDF below.