Andrew Kuhn September 5, 2019

Why Are U.S. Banks More Profitable Than Banks in Other Countries?


A Focused Compounding member asked me this question:

“Have you any thoughts on why U.S. banks are so profitable (the better ones at least)? I’ve looked at banks in other countries (the U.K. and some of continental Europe) and banks there really struggle to earn the spreads and returns on equity of high quality U.S. banks. This is particularly strange as the U.S. banking market is actually pretty fragmented, certainly more so than the U.K. which is dominated by a few giant banks and has a pretty non competitive deposit market. But U.S. banks seem to earn far better returns.

I was discussing this recently with someone who is a consultant to banks advising them in regulatory capital (among other things) and he said that it was down to the regulatory capital requirements being looser in the U.S. I’m really not convinced by that explanation though.
Is this something you’ve thought about at all?”

I don’t know if the regulatory requirements are looser really. There’s one aspect of regulation nothing about outside the U.S.: fees. It may be that U.S. banks are better able to earn non-interest income on fees (not sufficient funds fees, charging monthly fees for accounts below a certain minimum, etc.) then banks in some countries, because there might be tougher consumer protection rules in some other countries. However, from what I know of U.S. regulatory rules as far as capital requirements versus banks in other countries – I don’t really agree. It’s not that common for me to feel a bank in another country is a lot safer than large U.S. banks. So, if it’s a regulation advantage. It doesn’t seem to be an advantage due to forcing banks in other countries to be too safe.

The U.S. has FDIC. I don’t know what programs in other countries are like. Obviously, the FDIC program in the U.S. – combined with some other rules – helps minimize rivalry for deposits. An unsafe bank shouldn’t be able to draw deposits away from a safe bank just by offering high interest rates on deposits. And depositors shouldn’t abandon a bank they like just because they learn it may be about to fail. Obviously, before the FDIC and other rules – those were concerns which could mean the weakest operators in the industry would lower profitability for the strongest operators through irrationally intense rivalry for deposits.

Negative rates are bad for banks. There could be cyclical reasons for why you are seeing poor results in parts of Europe, because of that.

However, I need to warn you that it’s NOT true that small U.S. banks are more profitable than banks in other countries. They aren’t. It’s ONLY banks with good economies of scale in the U.S. that even earn their cost of capital. My estimate when I looked at U.S. banks as a group is that since about World War Two, they haven’t earned their cost of capital and they have earned returns below the S&P 500. What I mean here is if the small banks were never acquired, merged, etc. In reality, banks that don’t have good ROEs are often taken over by other banks and so shareholders might get better returns in the stock than the bank’s own ROE. However, I’d be extremely cautious about buying U.S. banks with low profitability.
Generally, you’re correct that less fragmented markets seem to have higher returns. Of course, very few U.S. banks operate nationwide. So, most banks are only going to be in one state, county, etc. Investors overlook this fact. But, there are tiny banks with 25% to 50% deposit share in the major town in their county or across the whole county. These banks lack scale when paying for technology, when paying for the costs of being a public company, with corporate overhead, etc. But, they aren’t a small player in their tiny community. So, some tiny banks aren’t in fragmented local markets. But, a highly concentrated statewide banking industry is obviously a better place to be. I wrote a report about Bank of Hawaii (BOH). Hawaii’s banking market is probably as consolidated as national markets in other countries. There are also fewer outside the state competitors in Alaska too. So, if you are right that less fragmentation should mean better returns – I’d start by looking at Alaskan and Hawaiian banks. Several of them are public.

My explanation generally for why bigger U.S. banks (like the biggest banks in a state, region, or country) have higher returns than banks in other countries is because American businesses and households have a different attitude to deposits. However, this wouldn’t be true of the U.S. versus all countries. Like you mentioned the U.K. I don’t have data on this, but I believe home ownership rates in the U.K. are even higher than the U.S. I don’t know if the mutual fund industry is as big in the U.K. as the U.S. So, to be honest, this explanation doesn’t seem to work for why a U.K. bank would be less profitable than a U.S. bank. From the customer’s perspective, the decision should not be overly focused on rates paid on deposits in either the U.S. or U.K., because people in those countries have a lot of other places to put their net worth. Regardless, it is true that in MOST countries versus the U.S., the country you are comparing to the U.S. will have lower net worth in stocks (for example). More net worth may be kept in banks. There will be more of a focus on rates paid on deposits than other factors compared to when U.S. households shop for a bank. This is important, because to the extent a depositor treats deposits as a form of wealth and not just a transaction account – the worse the negotiating position of the bank taking the deposits will be. I’ve talked to some people about banking in Asia (excluding Japan) and this is clearly the problem. The depositors are too aware of the importance of interest rates paid on their deposits, banks are too sensitive to using interest rates to control their level of deposits – as a result, there is too much rivalry on the deposit side. This makes it hard for those banks to get their cost of funding down as low as U.S. banks.

You can check this when comparing any bank.
A bank’s returns basically depend on 5 things:
1) Non-interest income
2) Non-interest expense
3) Interest expense
4) Yield on assets
5) Assets / Equity
Numbers 1-3 can be stated in terms of a % of deposits.

If the consultant you spoke to is correct – this should show up ONLY in assets/equity. Or, it’s possible if he means “risk adjusted capital levels” that it would show up in yield on assets. However, I’m skeptical of that. I’ve looked at E.U. banks and they are allowed to treat very high risk bonds of EU member countries (for example, Greek debt) as being safer than a U.S. bank can treat the obligations of a state such as Texas (or a government sponsored enterprise presumably supported by the federal government). Texas bonds are safer than Greek bonds. While sovereign and municipal debt ratings aren’t perfectly comparable, Moody’s probably rates some Texas obligations AAA and some Greek debt B. But there are probably banks weighting the Greek debt as safer than the Texas debt. The logic here could be that the EU is presumed to guarantee the debt of Greece no matter what. But, by that logic, is it more likely the U.S. would allow Texas to default than that the EU would allow Greece to default. So, I think the answer is that the U.S. and EU don’t have the same capital requirements as far as risk – but, I’m not sure the EU rules are always the more demanding. I think banks in both countries can – while staying within regulatory rules – either be run very safely or very dangerously. In my Texas vs. Greece example – you have to hold $5 of Texas debt to be considered as safe as $4 of Greek debt. That’s the most extreme example I can think of. But, it shows that this regulatory stuff is based on rules (that work in general) but which sometimes give nonsensical answers like suggesting banks would be safer if they had $4.5 million in Greek debt than $5 million in state of Texas debt.

To be fair, this is not an EU rule that’s looser than a U.S. rule. It’s just that a U.S. bank is usually not going to buy a risky sovereign bond (not because the bankers are any safer in the U.S. – just because the bonds they’ll be buying are U.S. debt). In fact, off the top of my head, I can only think of one U.S. bank that seems heavily concentrated in foreign government bonds that have nothing to do with its business. I’m sure there are others. But, it’s notable that only one specific bank came to mind instantly. So, this isn’t a regulatory difference between the EU and U.S. It’s just a practical difference. In practice, American banks will end up buying safe government debt, because there’s a lot of safe government debt in their own country to buy.

So, if outperformance did have something to do with capital levels that’d be in the form of underperforming banks having a lower assets/equity ratio (lower leverage) and maybe higher yields (though that’s much tougher to judge).
A difference in the business model would show up on the cost side. There’s an easy calculation you can do to find a bank’s true cost of funding:

Take FY 2018 (Non-interest Expense – Non-interest income) + Interest Expense
This is “total operating costs”. Divide by the average of December 31st, 2017 and December 31st, 2018 total deposits.
You now have a number that shows you the bank’s “all-in” cost of deposits. In the case of Truxton it’d be something like 1.8% (if I remember right). Please note this is misleading. I’m using ALL of Truxton’s wealth management fees to offset the bank’s cost of deposits. Logically, this misrepresents the business model. However, banks don’t usually give us enough info about the way various segments work, customers overlap, etc. So, for predicting the earnings of the entire bank – it’s often best to use the calculation I’ve mentioned here.

This means Truxton’s pre-tax earning power will be: Yield on Assets – 1.8%

Today, the Fed Funds Rate is 2.25%. And 2.25% > 1.8%, so it would seem that bank won’t lose money regardless of whether it makes loans, buys bonds, etc. All of those things should yield more than the Fed Funds rate (if they’re not, a bank could deposit money with the Fed). Note that this only applies to NEW bond purchases and loans made. Truxton, unfortunately, can get in a situation where it made 15-year loans it kept on its books that yield less than its current costs. This would happen if the company made loans today and then short-term interest rates rose to like 6% Fed Funds Rate. I haven’t done the calculation for Truxton – usually, you can estimate a bank’s interest expense as a stable fraction of FFR. So, like you might know Wells tends to pay 0.7 times the Fed Funds Rate on its savings / money market accounts or whatever (I made that number up, but the fact it’s best modeled as a fraction between 0 and 1 times the FFR is correct – loans, by the way, are best modeled as FFR + x% instead of 1.x times FFR).

My point is that you can check banks in any country against banks in the U.S. If they are paying more for deposits when you factor in interest expense plus non-interest expense minus non-interest income – it’s a deposit gathering model problem. I’m not sure that’s fixable ever. If the yield on assets is very bad versus the cost of funding – that could be a cyclical problem where interest rates are too low in the country. Maybe that fixes itself over time. I don’t know.
If the assets/equity ratio is too low – maybe the regulators are too tough on the bank (or more likely, the bank is just very conservative and sees little way to grow).

In theory, if it was just a leverage issue then you should see banks with the same ROA or higher than U.S. banks.
I can think of U.S. banks with ROAs (at present) between 0.9% and 1.8%. If European banks don’t have ROAs that high – I think it could be more than a capital issue. I think it could be their cost of funding from gathering deposits is not low enough. Or – it could be that the yield on assets is too close to their costs.

Here, it’s worth mentioning why I don’t think net interest margin is as helpful as most people do.
Interest expense is less stable as a % of deposits than net non-interest expense. What matters to a bank’s profitability is all of its expenses combined / all of its deposits. News reports usually make it sound like a Japanese bank can lend at a low interest rate and make money because the bank can pay nothing on its deposits. But, that’s wrong. A bank does more than just pay interest. In fact, in the U.S. a bank’s interest paying is one of the less important factors compared to its number of branches, their location, its website, its app, all the processing/tech it pays for, its advertising, its customer service, its number of ATMs in the local area at convenience stores, supermarkets, gas stations, etc. This stuff is more fixed as an expense. But, it’s every bit as real as the interest expense.

When the Fed cuts rates – or rates fall or are cut by a central bank in some other country – that only gives limited relief to the bank. It does NOT lower the bank’s cost of funding by anywhere near the size of the cut. If you do the math, a lot of bank’s will ALWAYS have a 1% to 2% cost of funding that can NEVER be removed by even having the bank pay zero interest on its loans. Even if rates were negative – they’d have to be very negative to offset the overhead of operating a bank.
It’s like if you were a producer of wooden widgets and you carved these widgets from wood at your factory. The central wood authority cuts the price of wood to nothing. That doesn’t mean you will make a profit. It just means you will make a profit on your materials cost. But, the cost of overhead on your factory is not controlled by the wood widget authority.

For this reason, a central bank can’t lower rates to the point where they would ensure bank survival, growth, etc. It can’t be done. You need to have the yield on whatever assets the bank owns be higher than the fixed costs the organization has. This isn’t how central banks intervene. Their playbook is not to subsidize operating expenses at banks, or try to steepen the yield curve by pushing down rates paid on deposits while trying to keep rates on loans and bonds high. Because of that, I think it’s very possible that – especially for banks with inefficient cost structures on the deposit side – low yield present a problem for bank profitability no matter how little they pay out in interest.

This does lead to the weird macroeconomic conclusion that I’m not sure how a banking system expands credit at a decent rate once the yield on the assets held by the banking system approaches the (net non-interest) expenses of the system. If you are making loans at 3% and provisioning 1% for losses such that your yield after losses is 2%, then a banking system with 2% net non-interest expense as a percent of deposits can’t expand (since it can never add to retained earnings, it can never grow without increasing leverage which is unsustainable). Assuming “normal” by U.S. leverage standards, you would need rates to be like 50 basis points higher than that (or deposits to costs 50 basis points lower) to support even normal nominal GDP growth assuming you pay ZERO dividends. Given the kinds of payout ratios banks like to have, if they’re unwilling to cut their dividends you would need yields of 4% on assets with 1% loss provisioning if you had 2% net non-interest expense just to keep the banking system expanding at a normal nominal GDP rate. And that’s assuming ZERO taxes. Banks are usually taxed. So, I think a banking system with 2% net non-interest expense and 1% loss assumptions on loans would actually need about 4.5% gross yields on its assets to feel it could keep growing at the kind of past rates of nominal GDP growth people were used to in the 1900s. But, the cost structure of banks varies so much it’s hard to say. Like, in the U.S., Wells Fargo could keep credit expanding year after year even with much, much lower than 4% yields. While the bank wouldn’t be very profitable – Wells could keep growing even with rates very close to 0% on bonds, loans, etc. This isn’t true of small U.S. banks. There are probably some where yields on the assets they could hold are already too low for them to grow credit as fast as the potential GDP growth rate of their local community. In other words, yields are too low now for them to expand credit at a normal rate. That may be true at larger banks in some other countries. I don’t know non-U.S. banks well enough to judge.