Geoff Gannon October 10, 2020

How is a Bank Like a Railroad? – And Other Crazy Ideas Geoff Has About Investing In “Efficiency Driven Businesses”

Someone sent in this question:

When deciding to look for a bank, where do you start? What is your initial approach when finding a bank?

So, I don’t really have one specific thing I look for in a bank stock. The way it usually works is that I read about a bank somewhere. I’ve gotten a couple good bank stock ideas off the Corner of Berkshire and Fairfax “investment ideas” thread (I read every post in that thread). Now, the truth is that it’s never been a very long thread about the bank. In fact, there are like 3 banks in all the time I’ve been reading Corner of Berkshire and Fairfax’s “investment ideas” thread that jumped out at me. The numbers someone cited just jumped out at me immediately as unusual. Now, unusual doesn’t necessarily mean good. But, it does mean do more research on it. So, the basic numbers where being especially high or low or whatever for the bank would make it really, really interesting would probably be:

– ROE (higher is better)
– ROA (higher is better)
– Efficiency ratio (lower is better)
– Deposits/branches (higher is better)
– Leverage (lower is better)
– Dividend payout ratio (lower is better)

Many banks are going to look awfully similar on these measures. That’s because banking is in some sense a commodity business. To me, banking is most similar to insurance. It also – in some ways – can be a little similar to things like railroads too. I know that’s hard to believe. But, I would say that there can be potentially some similarities between things like: banking, insurance, telecom, power, water, and railroads. Why?

These businesses work really, really differently from most businesses investors are used to analyzing. Most investors are used to analyzing big tech, media, restaurants, retailers, consumer brands, etc. Capital is relatively unimportant in those industries. Intangibles are important. Competition is fierce. The industry changes quickly. So, competitive position can be very, very important in those industries. Whereas capital is less important. And then oddly efficiency is less important as a differentiator. Efficiency becomes very important as a differentiator if you are in a more capital intensive and less competitive business. Like, a monopoly cable company can have all sorts of different returns depending on who is running it. John Malone runs it – it’s a great
business. Average family running it in the early days of cable – it was an okay business. It grew fast and such, but it may not have been very highly leveraged (so it paid taxes) and it was probably spending more than it needed to in terms of expenses (so EBITDA margins were worse) and so on. Businesses like banking, insurance, railroads etc. are very efficiency driven because the return accruing to owners is further removed from the value you can extract from the customer. If you have a great brand – it’s going to either be a good business or bad business primarily based on the product economics rather than the operational efficiencies. Basically, how many units can you sell at the same sort of prices as your competitors and then also how much pricing power do you have. So, I don’t think it makes a huge difference who is running a company once a brand like Tide or Reese’s or Tabasco is established.

I think it makes a huge difference in businesses like banking and insurance. In fact, when I’ve been asked about like how important the “jockey” is – I would say I’d only ever make a pure jockey bet in 3 situations: M&A roll-up business, insurer, or bank. That’s it. So, if you have a guy run a bank or an insurer or roll-up an industry for 20 years and then he leaves to start or takeover a brand new business in the same industry – if the price was right, I’d be willing to follow him. In fact, in one case of the banks mentioned at Corner of Berkshire and Fairfax – that’s what happened. I noticed that one of the banks mentioned was run by a guy who had like a long-term record at another bank that he left (“left” is putting it politely, they fired him) where you could analyze the past record.

So, my #1 thought when looking at an efficiency driven business (an “efficiency” driven business to me is one where capital requirements and fixed expenses are high and competition is low) is the management. Sometimes, “management” could mean culture. Like, if the business was founded by penny pinchers and all top management since then have been insiders promoted from the bottom up through to the top of the company – management is probably cheap. This would honestly be my focus for banks, for insurers, for telecom, for power companies, for water companies, and for railroads. Who is running the place? Who is the CEO? Who is the M&A guy? Who is the CFO? What is the culture? You want the key people and the culture to be one focused on compounding shareholder value and on keeping operating expenses very low. At different companies, the key person can sometimes be different. I know a bank where the CFO was probably more key than the CEO. Another where the Chairman (he’s a huge shareholder who basically bailed out the bank personally and took it over decades ago) is the key person. And obviously there are many where it’s the CEO. You can also think of situations – the building up of Citigroup is the most famous example, but I’ve seen other modern ones – where it’s really like a two person team. And then sometimes it’s the culture. Like, I know of one bank where everyone is from the same town and went to the same college and all that and basically was hand picked by the same guy who founded the bank. So, whoever ends up running that bank years from now – they probably have the culture created by that founder. Then there’s another one I know of where basically everyone who matters was taken from another bigger bank. So, like a bank got very big and successful and then some of the more independent minded people at that bank got tired of it getting slower moving and more committee driven over time and they joined another bank that was kind of like a re-founding of that successful big bank. You see this sometimes.

So, what is it that I want to see in terms of management/culture?

I think you want a very, very stingy management team. And stinginess to me can be scored on two points:

1) You want them to be super stingy when it comes to operating expenses. Do they talk about things like “other” expenses (postage and letterhead and other waste) and “occupancy cost” (rent) and headcount growth relative to asset growth and stuff like that.

2) You want them to be super stingy when it comes to using owner’s capital. Owner’s capital is equity. So, you don’t want a bank to think it’s okay to operate at 6x assets/equity or 12x assets/equity. At 6x assets/equity – they need to start talking about how they need to find an acquisition to do, or a share buyback to start, or a regular dividend to raise, or a special dividend to pay or something like that. Some bankers will say they have a “fortress balance sheet” and stuff like that when saying they are glad to be operating at half the leverage of other banks. But, this is wrong. Talking about ROA but not ROE when you have lower than normal leverage is a bad sign. Now, the reverse – if a bank says we have 1% ROA and 12% ROE but some of our competitors have 1.1% ROA and 11% ROE (using less leverage than us) so don’t give us too much credit for just using more leverage – that’s fine. I’m not saying a bank should only focus on ROE instead of ROA. It shouldn’t. But, it should be reluctant to issue shares. It should be reluctant to allow itself to operate permanently at lower leverage than it knows to be safe and so on. What is safe is going to vary by bank and management team. Some will say 15x assets/equity is correct and some will say 10x assets/equity is correct. Either answer could be right. I don’t think 5x or 25x is going to be right though. Actually, 25x should be basically impossible for a normal bank to do now. But, you get my point. You don’t want someone to do the equivalent of taking your money as a mutual fund manager and going 100% into short-term bonds. They’ll never lose you money. But, they’ll never make you any either. If you wanted to invest long-term with someone in stock and bond markets, you’d probably want them to invest as close to 100% in stock whenever they could find decent stocks at decent prices. But, you wouldn’t want them to use leverage. And you might be okay with them holding some bonds in like 1929, 1965, 1999, or today. In other words, if a banker says – our leverage has dropped from 12x to 8x over the last 5 years, because we just can’t find loans in areas we understand at decent rates – then, that’s fine. They just need to have a very clear understanding that it costs something to let owner capital sit idle and not be leveraged up fully as the laws and such allow. Saying we can’t find intelligent things to do right now is a good excuse.

The idea here is similar to insurance. I have no problem if an insurer normally writes at 2x premiums / statutory surplus (usually but not always similar to equity) when a regulator / A.M. Best would be fine with 3x premiums / statutory surplus – if they explain why. I’ve seen insurers that write 1x premiums / surplus and give no explanation as to why it’s not 2x premiums / surplus. They need to explain their leaving owner’s capital unleveraged in terms of betting of some kind. Like they have such small betting because the market is too loose and pricing is wrong and so on. They can’t just be so conservative that they don’t use appropriate leverage. Leverage is a normal part of banking.

So, I would want a management team obsessed with keeping expenses low and keeping the amount of owner’s money in the business low. This doesn’t mean I want them overleveraged. I wouldn’t necessarily complain if a bank operated at 10x assets/equity if it was always 10x assets/equity. They might have an ROA of 1.5% and ROE of 15%. I’m not sure that needs to be pushed to get the ROE to 22.5% or whatever if you leverage as much as some peers. But, I need some sense that there is a point where you are just hurting your shareholders by using less leverage than you know would be safe.

Then, I’d like a “niche”.

So, banks are “network” businesses in a sense. But, many banks don’t possess strengths on both sides of the network. So, a bank works like a money middleman. It takes money on the deposit side and lends it out on the loan side. What you’ll find is that the vast majority of banks are commodity like on both sides. They have no competitive advantage worth talking about as either a source of money (deposits) or a supplier of money (loans). Then, there’s a smaller subset of banks (less than half) that have some competitive advantage on either the deposit side or lending side. Very few banks have an advantage on both. Usually, if a bank has a double-sided true network advantage it’s going to be: a private bank, a business bank, or a specialized bank. In fact, specialization toward like one industry for a business bank or toward clients at a private bank that are also more focused on one industry is going to be more likely. Most banks I’ve found that are like this are not very big. So, it’d be hard to give examples people have heard of. I’d say that SVB Financial (SIVB) and First Republic (FRC) are two examples of size people have heard of. But, there are lots of bigger banks that have some segment of their bank that might have network type advantages tied to some specific industries. SIVB is somewhat tailored to venture capitalists and FRC is somewhat tailored to hedge fund managers. If you just look at the 10-year financials of those banks – you wouldn’t see that. But, if you go browse their websites, read their investor presentations, etc. – you’ll see their strategies involve more of a focus on some specific individuals / entities (often in banking, you’re going to find the bank does some stuff both on behalf of a firm and also for the key people at the firm – this is a good way to get sticky business and referrals).

But, this doesn’t happen much. Most banks you’ll find will not have strengths on both sides of the deposit taking / loan making side of
things. For example, I’ve talked about Frost before. Frost has strengths in deposit gathering. It doesn’t really have strengths in lending. They aren’t stupid, reckless, etc. Which you need to watch out for. But, an investment in Frost would basically be a bet they can do smart things on the deposit gathering side and then just not dumb things on the lending side.

A reverse situation is Hingham (HIFS). This bank doesn’t have strengths historically in deposit taking. In fact, this is very obvious by the fact they are lending > 100% of their deposits and they have a group called “specialized deposit group” inside the bank. Everything about HIFS gives you the impression they find it easier to make loans than to have deposits to fund those loans. Everything about Frost gives you the impression they find it easier to take in deposits than to have loans to make with those deposits.

Does that make HIFS a bad bank?

I don’t think so. And I know this confuses people, because I talk about the importance of having a low cost of funding (like at Frost) and then I turn around and write-up a stock like Hingham that does not have especially low interest costs and does not fund itself entirely with sticky customer deposits. Hingham uses borrowings. It uses what I’d call wholesale money (some is officially wholesale, but also just any money coming in for the rates is really hotter money than non-interest type deposits).

So, what does Hingham do right?

First of all, I don’t care what a bank’s non-interest expenses are. This is irrelevant. I also don’t care what a bank’s interest expenses are. These too are irrelevant.

There’s only ever one cost that matters.

It’s non-interest costs PLUS interest costs.

If you could build a bank that paid ZERO INTEREST to any of your depositors but provided concierge levels of service, no fees for anything, blah, blah, blah – you may actually have built a good bank. If you can go and lend money at 3.5%+ and leverage it up 10x+ – then, honestly you can afford to spend 2-2.5% or your assets each year providing amazing customer service as long as (in exchange) your depositors are happy to receive no interest.

The reverse is also true. If you could operate a bank – let’s say over the internet or something like that – which took in only massive deposits that were always happy to stick around, you could pay interest at like 2-2.5%.

The above was just for illustration. There are two realities to keep in mind.

One, all banks will have some non-interest expense and some interest expense. So, Hingham is efficient enough to be running at 0.8-0.9% of assets in terms of non-interest expense. Some banks can operate about that efficiently on a NET non-interest basis, but they do this through charging high fees. HIFS has like almost no fees at all. So, it is just exceptionally low expense versus the size of its assets. Frost, meanwhile, has a lot of non-interest paying deposits. But, on an interest plus net non-interest expense basis it is only very good – it doesn’t come in a lot lower than a super low expense bank that has almost all interest bearing accounts.

So, it’s always the combination of these two things. The cost of funding for a bank should only ever be evaluated on an “all in” basis. The “combined ratio” for a bank is the non-interest expense and the interest expense taken together. Never separately. Don’t let a bank fool you by stressing how low their non-interest expenses are and then never talking interest costs or by stressing how they pay almost nothing in interest but never admit they are a very high non-interest expense bank. This would be like listening to an insurer talk about their low expense ratio without mentioning their high loss ratio or vice versa.

The goal for any bank should be the same. It has to be to aim to become the lowest cost producer of money around. The low cost producer is the guy with the lowest combination of non-interest and interest expenses relative to assets.

And then secondly, I excluded charge-offs from the above. So, spreads have to be wider before charge-offs than I talked about above. For some banks, this matters very little. For others, it matters a lot. If you are making loans at 5.5% and charging off 1.5% per year on average over a full cycle – this is no different than making loans yielding 4% with zero charge-offs.

I can’t stress this enough. A wide net interest margin is not necessarily a sign of a great bank.

Often, it’s going to be the combination of two factors that’ll work better:

1) A low “all in” cost of funding (non-interest expense plus interest expense divided by total deposits is very low)
2) An almost non-existent charge-off rate

Why does this pattern work well?

If you have a low cost of funding, you can make low interest loans and still turn a profit.

If you have low charge offs, you can make low interest loans and still turn a profit.

Low interest loans are often going to be easier to make, easier to monitor, get you repeat business, etc.

A bank that has to make high interest loans all the time is just going to have a tougher time in a lot of ways.

So, say our goal was to survive on making 4% loans.

Well, if our “all-in” cost of capital is 2.5% or less and our charge-offs are basically zero percent we could have a spread there of like 4-2.5% = 1.5%. If we could make a huge amount of these 4% loans, we could have almost our entire balance sheet in loans instead of securities. With normal leverage a bank might use, we’d have a double-digit ROE despite investing in what are really quite low interest and presumably safe loans.

This brings me to the last point I’d look for in a bank.

I definitely like a specialty.

This isn’t required. I mean, a bank with a very low cost of capital could – in some interest rate environments – park money at the Fed, hold municipal bonds, etc. and get results that were anywhere from somewhat profitable to higher ROE than most other banks. If money costs you almost nothing on an “all in” basis – you really don’t need to actually lend it out except as required by your regulators and such.

But, most banks are going to lend out their deposits. And almost all bank failures are going to be the result of lending mistakes.
So, the lending side of a bank is important.

I would not focus on finding banks that can get higher yields on their loans than their peers can. I think that’s too difficult for an outside investor to evaluate.

Instead, I’d look for banks that do 3 things:

1) Aim to have very low charge-offs on their loans (I think it’s more achievable to eke out an extra 1% from charge-off reduction to basically nil than it is to eke out 1% by making 5.5% yield loans instead of 4.5% yield loans)
2) Specialize
3) Diversify

I know #2 and #3 sound like they are at odds.

Personally, I want as high a proportion of loans as is prudent to be made in as narrow a specialty as possible. This may sound like I’m suggesting making 50% credit card loans or 50% energy loans or 50% art loans or something. But, I’m really not. What I’m saying is that it is best to focus on making loans in your area of expertise. You might call this your circle of competence. You might call it loans where you have an edge. Whatever it is – you don’t want to make loans where you don’t believe you aren’t at least somewhat more skilled than other lenders. So, the average bank should not be making 5% energy loans and think this is safe. Banks that specialize in energy loans should be making 15% of their loans in that area. For a half dozen or so banks in the country – a 15% allocation to energy loans is appropriate. For 1,000 banks in the country: a 1.5% allocation would be inappropriate. Investors understand this idea better in insurance. In insurance, you want to invest in “niche” insurers whenever possible. You want to buy into the insurer that has the most past experience with losses in the industry, that has the most data, that has the most trained people in terms of the risks they’re covering. You want to avoid insurers that are writing some line of business for the first time. Investors in insurers get this wrong all the time. They buy into an insurer with a ton of growth in premiums, a low price-to-book, etc. – but that “book” is literally a book of business which is fairly recently written risks that the company has no prior experience writing and that was obtained on the basis of sometimes quoting a lower rate than a competitor who had been writing those risks for the last 30 years. Generally, I’ll admit, the risks in lending for banks are going to be really muted versus the risks in insurance. If your lenders are really level-headed emotionally and risk averse – I don’t think they’ll blow up a bank just because they don’t have any experience or data making these kinds of loans. I think that will happen more frequently in insurance. But, it’s still an analogy I’m comfortable with. Think of a lender like an insurer. You’d prefer they have a niche.

But, they have to be somewhat diversified.

The best way to do this is through having a low correlation between the two or more risks you are taking. I talked about Hingham in my write-up (on the Focused Compounding premium site). In that case, the bank is lending a lot against two different kinds of collateral (both in the Boston area). One is single-family houses. The other is multi-family buildings. The good news is these are both fairly low risk categories. The better news is they aren’t that closely correlated. In a housing bubble, you rarely have an apartment bubble. In an apartment bubble, you rarely have a housing bubble. The thing you’d want to avoid is like having two equal categories: housing and construction. That’s bad because housing is fairly low risk. Construction is fairly high risk. But, construction and housing are too highly correlated. The moment when you could end up foreclosing on some land bank that’s totally undeveloped by some now bankrupt homebuilder is the same moment when your loans on existing homes are also going bad.

So, you want two or more specialized loan categories that aren’t overly correlated with each other.

Will you find this?

Very rarely.

Banks just aren’t run that way very much. In my experience, you’ll be more likely to find a good bank that is just very strong on the deposit side and then totally mediocre and generic on the lending side. I’ve found it very hard to identify banks with strong lending side attributes, because almost all banks are run conventionally on the lending side. They are diversified beyond their points of potential specialization. As a result, they tend to take fairly broad and normal cyclical charge-offs based on the area of the country they are in and a mixed loan portfolio. Unfortunately, they are sometimes overweight enough to some high risk loan category to manage unusual losses in that area during something like a bubble.

So, generally, you aren’t really looking for “good” lenders. You’re just looking for “not bad” lenders who are good deposit gathers. So, like you identify a good deposit gather and then you check to see if they are making loans in a state they haven’t before, in a loan category they haven’t before, in something you suspect to be a bubble, etc. The tough part for investors is that they usually focus on the size of the loan category instead of the size of the potential losses. So, like they’ll see that residential is 50% of the bank and RV loans are only 15% of the balance sheet. Focus on the RV loans instead of the home loans. It is easier to blindly do something dumb that will risk the bank as an overall entity by putting 15% into RV loans than it is by putting 50% into home loans. I mean, as long as you are in a part of the country where land values are not overinflating – you’d have to be dumb in pretty creative ways to sink a bank with 50% in home loans. They are just such a “general” type of loan that it shouldn’t be possible to do something that destructive unless there’s a bubble that you deny.

Bubble denial could cause a wipe-out of almost any bank. So, you want to watch out for that. If I was invested in a bank making energy loans and they said $110 a barrel was the new normal in oil and they just look at the market price and so on – I’d get out of that stock. If instead they mentioned that although oil is at $110 a barrel, we keep making our loans on an assumption of $60 a barrel – then, they might be right or wrong but they aren’t just blindly denying a bubble.

I’d say the biggest risk to a bank that is actually skilled in lending is that a bubble develops in a category that is large for them. This could sink any bank. I’ve mentioned this with Farmer Mac (AGM). They are an agricultural bank (not really a bank – they’re a GSE, but we’ll pretend they are “banklike” for this discussion). I honestly believe a true farm bubble like we saw twice in the last 100 years would basically bankrupt AGM. I don’t think any management team could stay prudent long enough at a GSE like that not to be wiped out by a bubble. Farm bubbles just get too big for it to be safe to have 100% of your assets tied to them and use the leverage Farmer Mac uses.

I feel that way about other banks too.

So, in terms of lending…

1) Are they prudent?
2) Are they specialized?
3) Are they diversified?
4) Are they trapped in a bubble?

If a good lender is trapped in a bubble – honestly, I’d probably just bail on the stock. This sounds unfair to management. And perhaps it is. But, the risks of large lending losses outside of bubbles are just so small compared to the risks inside a bubble – that I’d just avoid bubbles altogether.

I’ve mentioned this before with other lenders. Like, I did a write-up of Car-Mart (CRMT) where I said how much I liked CRMT but how little I liked its competitors. If you feel the competitors are making such bad loans that it’s hard for anyone in the industry to stay prudent without their market share vanishing to nothing – it may be right to just avoid anyone in that loan category altogether.