I came to the U.S. in the fall of 2008. It was a turbulent time when I read about bankruptcies or about “too big to fail” banks almost every morning. As a new student of economics and value investing, I developed an ingrained prejudice against banks in this abnormal period.
Like most casual observers, I thought that banking is a risky business because of high leverage. Business cycle is inevitable and a small mistake can be magnified into big losses. So, I’d better stay away from banks.
Geoff asked me to look at several banks over the years. No bank was good enough to change my preconception. But some did recently. I spent the last 3 months researching banks. I realized that there are good banks and bad banks. Some have more stable funding sources than others. Some banks focus on inherently safe loans more than others. It’s not impossible to understand some banks.
In this post, I will discuss why some banks can be a great long-term investment. I will also discuss what I like and don’t like about Wells Fargo.
Good Banks Can Grow Profitably for Many Years
The banking industry is very durable and predictable. Deposit growth tends to follow GDP growth. Over the last 20 years, total deposits of all FDIC-insured institutions compounded 6.2% annually.
Good banks can do better than that. The long-term trend is that there are fewer and fewer banks. The total number of FDIC-insured institutions declined from 11,970 in 1995 to 6,509 in 2014. That means deposits per bank can grow faster than GDP growth. The group of banks with more than $10 billion in assets grew much faster than groups with fewer assets. That can be explained by acquisitions. But the fact is that in most local markets, the top 5, 10, or 15 banks as a group tend to increase total market share over time. So, a durable bank can have a high chance of growing more than 5-6% organically. There aren’t many businesses that have high single digit growth forever. But that’s very possible for good banks.
The banking industry has lower costs than other types of lenders. Banks get funding from noninterest-bearing deposits, interest-bearing deposits, and other interest-bearing liabilities. The rates banks pay for interest-bearing deposits and other liabilities is very stable as a fraction of the Federal Fund Rates (FFR). For example, the cost of interest-bearing deposits for most banks is between 60% and 90% of FFR. Some deposits are more expensive than others. For example, the cost of money market account can be 70% of FFR, and the cost of time deposits can be 90% of FFR. The total cost of funding for banks is usually lower than FFR. The estimated cost of funding as % of FFR for some banks I looked at is:
Commerce Bancshares: 48%
First Financial: 49%
Prosperity Bancshares: 53%
BOK Financial: 59%
Wells Fargo: 64%
U.S. Bancorp: 67%
The industry data isn’t available but these are banks with lower than average funding. We can assume that cost of funding for the industry isn’t far from 80% of FFR. So, if FFR is 3%, the cost of funding is 2.4%.
The average net operating cost of banks with $1 billion to $10 billion in assets is 1.8%, and of banks with more than $10 billion in assets is 1.1%. These two groups are representative of the industry because banks with $1 billion to $10 billion of assets hold 10% of total industry deposit, and banks with more than $10 billion in assets hold 80% of total industry deposit. So, the industry’s average net operating cost is less than 1.5%.
That means the industry’s total cost of money is less than 4% if FFR is 3%. That’s lower than risk-free rates and is surely lower than the required rate of return for most money market funds or bond funds.
So, the banking industry has lower cost than other lenders. But strong rivalry among banks forces yields to go up and down along with the cost of funding. Therefore, the industry’s net interest spread, defined as yield minus cost of funding, was very stable at about 3.5%, as shown in the following graph:
The U.S. banking industry’s net interest margin had a fairly stable tendency to be about 3.5% from 1999-2014.
Good banks can make 15-20% after-tax ROE. For example, Wells Fargo’s total cost is about 2.5% of total asset assuming 3% FFR. That means a very modest yield of 4% can result in 1.5% pre-tax return on earning assets. With 10x leverage, ROE will be 15% pre-tax and 10% after-tax. A normal yield of 6% results in 3.5% pre-tax return on earning assets or 23% after-tax ROE. FFR can go up and down, but Wells Fargo tends to maintain at least 1% cost advantage over the industry. That means Wells Fargo’s ROA is always at least 1% more than the industry’s average ROA. That guarantees an above average ROE.
Some Banks Have Low Risk
The main argument against banks is leverage. But not all liabilities are equal. Deposits are not debt. Deposits are more like insurance float. Even better, deposits tend to increase a lot in financial crises because people want to conserve cash. Some good banks are considered “safe havens” and thus gain deposit market share in bad times.
Short-term liabilities are what cause liquidity problems. Many banks rely on commercial paper or repo. According to a McKinsey report, deposits were just 49% of liabilities of U.S. banks in 2012. Commercial paper, repo, senior debt, and other liabilities made up 39% of liabilities. So, these banks can have big trouble in renewing short-term borrowings in a crisis. That’s very different from a bank like Frost whose deposits represent 96% of total liabilities.
So, the inherent risk in banking is liquidity. The key for investors is to avoid banks that rely on borrowings other than deposits. Regulators watch some capital ratios but banks with a lot of short-term borrowing could still face liquidity problems in a crisis even if they have a lot of tangible equity to assets. Quality of funding sources is more important than capital ratios. Banks that rely on deposits for most of their funding are very safe. They won’t have a liquidity problem even if they incur some loan losses.
However, banks have to maintain regulatory capital ratios. So, it’s important to avoid loan losses.
I found that loans have different risk profiles. For example, consumer loans tend to have more losses – offset by higher yield – than business loans. That’s perhaps because of government programs that encourage lending to home buyers or students. Among business loans, commercial real estate (CRE) loans tend to have more losses than commercial & industrial (C&I) loans. That’s because there is more speculation in real estate development than in other industry.
So, the types of loans a bank makes are as important as a conservative culture in minimizing losses. For example, it’s widely accepted that Wells Fargo has conservative lending. Yet, its average net charge-offs over the last 20 years was 1.09%. That’s way higher than Frost’s 0.48% or BOK Financial’s 0.27%. Frost and BOK make much fewer consumer loans than Wells Fargo.
Overall, banks that have stable funding sources and that make inherently lower risk loans can be very safe. I propose a checklist for good banks:
1. High deposits/liabilities (80% or more)
2. High noninterest-bearing deposits/total deposits (30% or more)
3. Low operating cost (1% or less)
4. High C&I loans/Total loans (60% or more)
My View on Wells Fargo
When I first looked at Wells Fargo, I thought that it can make $60 billion pre-tax earnings. The idea was simple. The median Net Interest Income/Earning Asset from 1991 to 2014 was 4.45%. Net Operating Cost was 0.57% in 2014. Net Operating Cost has been declining over time so it makes sense to use the latest number instead of the past median number. So, Wells Fargo can make 4.45% – 0.57% = 3.88% pre-tax return on earning assets (ROEA). As of the last quarter, Wells Fargo had $1.55 trillion in earning assets. $1.55 trillion times 3.88% equals to $60 billion.
Another approach gives a similar result. Wells Fargo’s cost of interest-bearing liabilities as a % of FFR was stable at about 88% with a variation of 0.20. Wells Fargo’s net interest spread, defined as yield minus cost of interest-bearing liabilities, was very stable at about 4.52% with a variation of 0.13. So, at a normal 3% FFR, cost of interest-bearing liabilities would be 3% * 88% = 2.64%, and yield would be 2.64% + 4.52% = 7.16%. Free funding sources are 27% of total earning assets so weighted average net interest spread is 5.23% (= 27% * 7.16% + 73% * 4.52%). Subtracting net interest spread by 0.84% average net charge-offs/earning assets and 0.57% net operating cost results in 3.82% pre-tax ROEA. That implies $59 billion pre-tax income.
Wells Fargo’s current market cap is $275 billion. So, Wells Fargo is trading at only 4.6 times normal pre-tax earnings.
That makes Wells Fargo the cheapest bank I’ve ever seen.
Even if it takes 10 years for Wells Fargo to make a normal 3.82% ROEA, investors can still make 13% annual return based on today’s price. Assuming 5% growth, Wells Fargo would have $2,520 billion earning assets in 2025. Applying 3.82% ROEA results in $96 billion pre-tax earnings. That’s 10.6% annual growth from last year’s $35 billion pre-tax earnings. Adding 2.8% dividend yields results in 13.4% annual return. There can be also some multiple expansion and share buyback.
Wells Fargo has characteristics that I like.
It has a low cost of funding. Noninterest-bearing deposits are 32% of total deposits. Noninterest-bearing deposits are only 20-25% of total deposits at most banks.
It has an extremely low net operating cost of 0.57%. That’s the lowest I’ve ever seen. I looked at about 50 publicly traded regional banks and most have between 1.6% and 3.3% net operating cost. Wells Fargo has low net operating costs thanks to cross-selling. For example, its retail bank cross sells on average 6.17 products per household. That generates a lot of fee income.
Wells Fargo has strong consumer brand awareness. All national banks have this advantage but Wells Fargo is better at cross-selling than any other bank. It’s likely that Wells Fargo can gain market share over time. In other words, it can grow deposits by more than 5% annually.
Wells Fargo is conservative. For example, it exited subprime lending in 2004. Most of its problems during the Great Recession were related to the loans acquired from Wachovia. Its lowest pre-provision earnings before tax (2.43% in 2014) covers 1.4 times the highest net charge-offs (1.76% in 2010.)
But there are things I don’t like about Wells Fargo.
Wells Fargo is very big in mortgages. The mortgage origination business is okay. This business requires scale and Wells Fargo funds one out of every three home loans in the U.S. The problem is that it retains about 10% of the loans it makes. Mortgage loans are 21% of earning assets. I’m not comfortable with these loans.
There are some adjustable-rate mortgages (ARM). Can customers afford interest expenses of these loans if interest rates increase? It’s likely that loans made after 2008 are very safe. Mortgage loans have been so restricted after The Great Recession. For example, total mortgage credit availability index was just 125.5 in July 2015 compared to almost 900 in July 2006. Regardless, a higher interest rate is still a concern for ARM.
Fixed mortgage loan can limit upside. Consumers took advantage of the low interest rates to refinance mortgage loans. So, Wells Fargo may have to carry a low yield portfolio for a while. For example, the yield on 1-4 family first mortgages was 4.19% in 2014, which is way lower than 7.27% in 2006.
The securities portfolio is another concern. Securities totaled $322 billion or about 21% of total earning assets. Mortgage-backed securities (MBS) totaled $124 billion. Wells Fargo expects the value of MBS to decline by $8.2 billion if interest rates increase by 2%. It doesn’t give the same expectation for the whole securities portfolio. But it can possibly lose over $35 billion if interest rates increase by 3%.
This is just a paper loss. The weighted-average expected maturity of securities available for sale was just 6.6 years. So, Wells Fargo can hold these securities until maturity and incur no actual loss.
This potential loss is only relevant because of regulatory capital ratios. Basel III requires banks with more than $250 billion assets to include marked-to-market gains or losses in the calculation of capital ratio. Assuming a $40 billion loss, tier I ratio would be 7.6%, which is well above the requirement of 6%. Supplementary leverage ratio, which takes into account off-balance sheet exposures, would be 5.9%, which is higher than the required 5%. So, Wells Fargo passed my stress test.
My last concern is that Wells Fargo is a “too big to fail” bank. Even if it will never have a liquidity issue, it can be forced by regulators to get a cash injection, resulting in share dilution. The book “Too Big to Fail” by Andrew Ross Sorkin tells the story about the protest of Wells Fargo’s CEO Kovacevich against the TARP:
“Dick Kovacevich, for one, was obviously not pleased to have been given this ultimatum. He had had to get on a flight—a commercial flight, no less—to Washington, a place he had always found contemptible, only to be told he would have to take money he thought he didn’t need from the government, in some godforsaken effort to save all these other cowboys?
“I’m not one of you New York guys with your fancy products. Why am I in this room, talking about bailing you out?” he asked derisively.”
And Henry Paulson responded with a threat of a regulatory crackdown:
“You’re going to get a call tomorrow telling you you’re undercapitalized and that you won’t be able to raise money in the private markets.”
Well Fargo is a good bank. It’s very cheap. But it’s too big. Its loan portfolio is too complicated. And it gets too much attention from regulators. Many people follow the stock. They may know things that I don’t know about Wells Fargo. So, I prefer smaller banks that have a simpler balance sheet.