A commenter writes,

AWS (the Amazon “cloud”), is wildly profitable (20%+ margins), and while only 10% of their revenue currently, is growing much (over 2x) faster than the rest of the business.

He offers these figures for the second quarter:

Amazon Web Services $4.1 billion revenue, $0.9 billion profit, revenue growth rate of 50 percent per year.

Amazon excluding AWS $33.5 billion revenue, $1.5 billion loss, revenue growth rate of 22 percent per year.

So let me play with these numbers a bit, to try to get to $12.5 billion in profit, which I think is needed to justify the current stock valuation. We have four numbers to play with: AWS revenue, AWS margin, ex-AWS revenue, and ex-AWS margin.

1. Start with a combination of AWS margin of 20 percent, ex-AWS margin of 0. Then we need AWS revenue to be at $62.5 billion, compared to $4.1 currently. That might not happen for . . .ever.

2. Start with a combination of AWS margin of 20 percent and ex-AWS margin of 4 percent. At current revenues, that gives them $.9 billion and $1.34 billion, respectively, for $2.24 billion. That means that they have to grow revenue in each sector by a factor of 5 to get to $12.5 billion in profit. Even if AWS maintains a 50 percent growth rate, it would take 4 years to go up by a factor of 5. And it would take ex-AWS a lot longer. All while supposedly goosing their profit margins in retail to 4 percent.

Another commenter notes that Dean Baker thinks the way I do about Amazon.

Other commenters have said that profits are understated, because Amazon has plenty of “free cash flow” that it is currently spending to try to enhance its capabilities. OK, but Amazon’s revenues are not understated. They had $38 billion in quarterly revenue. How much of that could possibly be profit, under the most generous accounting?

I’ve got $100 that says the market cap of Amazon is lower on July 31, 2020 than it is today.

Mark J. Perry writes,

Amazon and Facebook were in the news this week for “joining an exclusive club open to only the richest companies in the world: both crossed the half-a-trillion mark.” Those two companies join Apple ($785 billion as of today), Alphabet/Google ($652 billion), and Microsoft ($564 billion) in the exclusive club of companies whose market capitalization tops $500 billion. To put the size of that market cap into perspective, consider that the entire Mexican stock exchange has a current market valuation of $438 billion, Thailand’s publicly traded stocks are worth $468 billion and the Russian stock exchange has a market cap of $554 billion! Together, the five US companies in the “$500 billion” club have a combined value of nearly $3 trillion… To put that into perspective, just those five US companies, as a separate country or stock exchange, would be the eight largest stock market in the world (see table above), and 40% larger than No. 9 Canada’s entire stock market and almost 50% (and $1 trillion) larger than the German stock exchange!

Some arithmetic:

Amazon just reported quarterly earnings of $200 million. At that rate, it would take 2500 quarters to earn $500 billion. That is 625. Years.

Call me impatient, but I do not care to wait 625 years to earn back my investment. I would rather see the money paid back a bit sooner. Like maybe 10 years.

If you want a $500 billion investment paid back in 10 years, then you need quarterly earnings of $12.5 billion. What does Amazon have to do to achieve that?

Their earnings are their sales times their profit margin. Investopedia writes,

Each year, the S&P 500 releases industry specific returns on equity and net margins, and each year the retail industry is among the least profitable. This is especially true for web-only retailers, which often see net margins as low as 2.50 to 3.50%.

So if we assume that Amazon is capable of achieving a 4 percent profit margin, we are being generous. At that profit margin, to get $12.5 billion in quarterly earnings, you would need $312.5 billion in sales. Actual sales last quarter were $38 billion.

Unless I have made some errors (check to see that I haven’t messed up by a factor of 10 somewhere), this arithmetic says that Amazon needs to be 8 times as big as it is, with a generous profit margin, in order to make me want to own the stock.

In July of 1999, I wrote Arithmetic in a Bubble. My first blog was an extended exploration of the topic. Maybe it is time for that blog to make a comeback.

John Cochrane writes,

For $100 of assets, and $100 of bank equity, let, say, $10 of that equity be traded — enough to establish a liquid market. Then, let $90 of that equity is held by a downstream entity or entities— a fund, special purpose vehicle, holding company or other money bucket. I’ll call it a holding company, and return to legal structures below. The holding company, in turn, issues $10 of holding company equity and $80 of debt.

The good news is that the bank can never fail, and if the holding company becomes insolvent, Cochrane suggests that it could quickly force debt holders to accept a new mix of (less) debt and some equity.

The bad news is that the debt of the bank holding company, which is the closest thing to bank money in his world, will not trade at par. It will trade at a discount, which will be small if the bank seems to be doing well and large if the bank seems to be doing poorly. For transaction purposes, that sort of money is noisy, so it will not really work. If it would work, then nobody would be bothered by a money market fund that can “break the buck.” In order to avoid breaking the buck, money market funds have to stick to non-risky assets.

I think that in order to work as money, financial liabilities have to trade at par. If a firm has risky assets and most of its liabilities trade at par, then I think it has to be subject to runs (assuming no deposit insurance). I don’t see any way around that.

A reader writes,

I would describe the posture of big banks (and other well-heeled financial institutions) as “pleading for mercy,” more than “buying friends and influence.” Banks are giant piles of treasure and if in the olden times they would send hordes, spells and dragons, now they send the DOJ, regulators (state, fed and quasi-public, like FINRA) and the plaintiffs’ bar.

Big banks are subject to shakedowns by grandstanding government officials. But I think that in exchange for paying this “tax,” big banks get enough benefits from government that they are better off big.

I do not believe that there are economies of scale in banking at such high levels. We do not see banks get large organically. They get large through mergers. The result is to create institutions with very high levels of operational risk, because senior management cannot possibly keep track of what the various units are doing. But if one of the units messes up, the taxpayers are there to provide a bailout.

A commenter writes,

1) In reality, it was the medium size banks or investment banks that were the worst offenders. not the largest one. In the case of the largest ones, only Citi really was that borderline survived if you assume BofA was sort shotgun to take Meryll Lynch.
2) Haven’t most long time businesses consolidated a lot the last 30 years. Grocery stores or airlines. The economy has naturally moved this direction with most long time businesses.
3) Historically the US has been the developed world oddball with a heavy local banking presence.
4) In terms of the S&L crisis there was bailouts and lots of failure of small institutions. In fact the government probably wrote a bigger check on that crisis than the 2008 Housing Crisis despite being significantly smaller.

Regarding point (4), the bailout cost the taxpayers $150 billion. Even adjusting for inflation, the bill for the financial crisis was quite a bit higher. Also, most of the cost of the S&L crisis could have been avoided had the regulators not relied on “extend and pretend.” The S&L business model was destroyed by inflation in the 1970s, but most of the clean-up did not take place until a decade later, by which time the losses had multiplied.

Regarding point (3), this is very much a plus for the United States in my opinion. We have the best developed stock markets and venture capital market in the world in part because until the 1980s our banking system was unusually fragmented. There is a decent argument (made most strongly by Calomiris and Haber) that our banking system was too fragmented historically, but by now I believe that our banking system is too concentrated.

Regarding point (2), there are industries where firms grew to dominate because of superior execution and/or economies of scale. That is definitely not how the top banks in the U.S. became large. They got big through mergers. And through all sorts of regulatory barriers to competition.

Moreover, in other industries there has been new entry. The competition in groceries in recent decades has been very intense. Airline competition has been more uneven, but still there has been much more entry than in banking, even though the fixed cost of getting an airline operating would appear to be much higher than the fixed cost of opening a bank.

Regarding point (1), I do not claim that big banks are inherently riskier or more poorly run than smaller institutions. My main problems with big banks are:

1. Their CEOs have immediate access to key politicians.

2. Money managers who lend to banks have to multiply the probability that the bank is/becomes insolvent times the probability that it will not be bailed out. For small banks, the second probability is not high, because the FDIC tries to resolve these banks with mergers, and that leads to excess willingness to lend to those banks. But for small banks, the second probability is exactly zero, and that distorts behavior even more.

3. In a crisis, the big banks are certain to receive bailouts. The CEO of a big bank can expect huge rewards for success, or even for mediocrity, with no accountability at all for failure. How could the rest of us not feel angry about that situation?

Ronald J. Mann wrote,

Wal-Mart’s application to form a bank ignited controversy among disparate groups, ranging from union backers to realtor’s groups to charitable organizations. The dominant voice, though, was that of independent bankers complaining that the big-box retailer would drive them out of business. Wal-Mart denied any interest in competing with local banks by opening branches, claiming that it was interested only in payments processing. Distrusting Wal-Mart, the independent bankers urged the FDIC to deny Wal-Mart’s request and lobbied state and federal lawmakers to block Wal-Mart’s plans through legislation. Ultimately, WalMart withdrew its application, concluding that it stood little chance of overcoming the opposition.

I was reminded of this by Lawrence J. White, who had written a piece in favor of allowing Wal-Mart into banking.

We are always told that we need regulation to protect consumers and make the financial system safer. That is the theory. The practice is that regulation very often gets used to limit competition.

This is an example of what I mean when I say that in any dispute between libertarians and statists, the libertarians are usually right.

A commenter writes,

Before the depression, the US heavily regulated banks and restricted the founding of branches; there lots of small banks tethered to local markets. In contrast, Canadian banks didn’t face such stringent regulations and were larger and more diversified. For this reason, in the US we had an epidemic of bank failures, and Canada did not.

My thoughts:

1. The best part of Canadian banking is their mortgage design: a five-year rollover, with recourse. Recourse means that if your house goes down in value, you cannot just turn the keys in to the lender and walk away. They can come back to you to make up their loss. Our 30-year, fixed-rate, no-recourse mortgage has lots of credit risk and interest-rate risk that sits with the lender, until stuff happens, and then it goes to the taxpayer. We got stuck with losses from interest-rate risk during the S&L crisis, and we got stuck with losses from credit risk during the 2008 crisis.

2. The worst part of Canadian banking is the high concentration in large banks. As in Europe, this goes along with a very stunted equity market. Firms raise capital using debt, and they owe that debt to big banks. The U.S. system, with its much more prominent stock markets, is better.

3. The worst part of the U.S. financial system is the political power of trade associations and large banks. The trade associations have leverage over Congress, and the large banks have leverage over everybody in Washington. In Canada, the big banks have to respect the regulators. Here, the bank executives can go over the heads of the regulators any time they need to.

4. Over the last thirty years, we have seen a decline in the relative importance of the stock market in the U.S.:fewer public firms, many fewer IPOs, and firms raising less of their capital in the stock market. At the same time, we have gone from a fragmented banking system to one that is very highly concentrated. Household wealth also has become more highly concentrated.

I think that if you don’t do something to limit the growth of big banks, you end up with big-bank-dominated corporate finance, meaning less active equity markets and a less democratized financial system. Also, given our political culture, you end up with the political system subservient to the CEOs of the biggest banks. In short, you combine the worst of Canada and the worst of the U.S.