What you will learn
- What makes banks different
- How to approach bank valuation
- How OSV can help
Why Banks are Hard to Value
Aswath Damodaran put it succinctly in his piece on Valuing Financial Services Firms:
“The problems with valuing financial service firms stem from two key
characteristics. The first is that the cash flows to a financial service firm cannot be easily estimated, since items like capital expenditures, working capital and debt are not clearly defined. The second is that most financial service firms operate under a regulatory framework that governs how they are capitalized, where they invest and how fast they can grow. Changes in the regulatory environment can create large shifts in value.”
To make the first point even more clear:
- “Debt” is a somewhat undefined concept for banks because it is more like a raw material than a source of capital.
- Reinvestment in financial services firms is not easily measured by capital expenditures and profit retention.
Implications for Bank Valuation
So, what does that mean for valuation?
First, I think this recent article about Wells Fargo is a great example of how to think through a bank’s valuation. Pay particular attention to the “Valuation” section.
Valuation is not just about the numbers. It’s about building a data-based story. Before you deep dive into a bank, though, you need to narrow down your prospects. Let’s look at how.
Use Book Value
Banks are required to mark their assets to market. Because most of their assets are tradable financial assets, these book values are a fairly reasonable estimate of their market value.
Of course, as value investors, we know that assets can be over- or under-valued for extended periods of time. It can take the market a while to catch up, so we can’t just blindly trust these values, but they are going to be a lot closer to reality than most other businesses.
Also, it’s even better if you use Tangible Book Value, which excludes the value of goodwill and other intangible assets, which are not going to be marked to market in the same way.
Second, if free cash flows are hard to estimate, it’s generally accepted that a very good proxy for banks is dividends. And in the day and age of large buybacks, you would include those, too. So, total shareholder return (dividends paid + net equity repurchases) is a good estimate of free cash available after reinvesting in the business. You can thus discount future dividends to approximate intrinsic value.
How to use Old School Value to Analyze a Bank
Screen For Attractive Banks
To screen for attractive banks where it would make sense to think more thoroughly about the valuation, here are filters I’d recommend using:
- Price to Tangible Book Value, TTM. This is still a great valuation metric for banks. I’d look for opportunities < 2.
- In OSV, Other Current Liabilities is a close proxy for deposits, which is the raw working capital of a bank. I’d screen for Other Current Liabilities 3-year growth > 0. The higher the better.
- I also like my banks returning capital to me. I’d screen for dividend yields > 2% (.02 in the screener), or higher. Higher yields suggest lower prices/better values, too.
- You could also screen for 3-year growth in dividends paid (from the cash flow statement). This doesn’t normalize for the number of shares outstanding, but if the amount they’re paying out is growing, this is good. (Also, since this is a cash outflow, you want to screen for this growing negatively).
- Since banks are also doing a lot of buybacks, I would screen for 3-year diluted shares outstanding growth < 0, too.
Here’s a screenshot of a possible screener:
You could probably also add P/E in there.
This shows WFC, BAC, JPM, and C are possibly attractive right now, along with about 2 dozen others.
Analyze Key Metrics
Once you’ve screened for high potential bank stocks, your first step should probably be to confirm that the bank is not relatively over-valued compared to its past history. Here is JPM’s historic P/TBV and P/E:
Neither of these screams “OVER VALUED,” but JPM also doesn’t look like a bargain.
Use a Dividend Discount Model for Bank Valuation
Next, you can try a simplistic dividend discount model to value the bank. Treat the sum of a bank’s dividends and share repurchases as a close proxy for its free cash flow to shareholders.
Sum those two numbers from the cash flow statement and look at their growth over the last few years (and/or look forward at what you think the likely future growth will be), then plug those into the DCF valuation model. That’ll basically then work like a dividend discount model.
So, for JPM as of early August 2019:
- TTM dividends of 11.5B with growth of 6-13% lately
- TTM net change in equity repurchases of 19.7B with some pretty wide fluctuations in the growth here
- Net income growth can help figure out a probable future growth rate, as without earnings growth, dividends and buybacks won’t grow; this, too, has been variable but is probably closer to 1-3% overall
Plug $31.2B in shareholder return and 2% growth into the DCF, and it looks fairly valued.
Note that this is using a discount rate of 7.5%, which is probably a little higher than JPM’s cost of equity. (Since dividends and buybacks are cash flows to shareholders, we’re talking about the firm’s cost of equity, not their overall cost of capital.) Generally, though, I typically don’t use a firm’s cost of capital as my discount rate; instead, I use my own investment hurdle rate given the risk and my other options, so 7.5% seems pretty fair for a large, stable bank in today’s environment.
That’s how we’d approach bank valuation using our tools. We wouldn’t recommend using Piotroski or Altman much to determine quality since both are pretty focused on Assets instead of Equity.