With some of the banks on the JSE having recently reported and relatively large movements in their share prices, there has been a raft of articles discussing their different merits and valuations.
For example, despite the fact that Capitec put out strong numbers, the Capitec stock is trading on a valuation of 7.7x its book value relative to the rest of the JSE-listed banking stocks that are trading on a 1.1x average Price/Book.
Surely, Capitec is expensive then?
Well, first let me explain what a bank actually is:
If you look beyond the flashy branding and big head offices, the value of a banking license is really only two things:
- Accept deposits: It allows a bank to accept deposits from clients which it can decide what to pay these deposit holders in terms of interest for their capital, &
- Exclusive, cheap credit from the Central Bank: It gives a bank near-infinite access to an effective “overdraft facility” at the South African Reserve Bank where you can lend large amounts of capital quickly and at extremely low interest rates.
Both (1) and (2) are actually only one single benefit: access to exceptionally cheap capital.
What the bank does with this cheap capital really is what differentiates it.
The most common use of this “cheap capital” is to lend it out at a higher rate than the bank has borrowed it at. By earning a spread between what it borrows at and what it lends out at, the bank magically creates profits which requires (basically) no equity funding.
(That said, there are also services and other aspects that can wrapped into a bank’s business and these can also be extremely profitable.)
In a simplistic way, a “bank” is really just a balance sheet where it has cheaply-funded liabilities that it employs to generate the highest possible (risk-adjusted) return.
And, thus, valuing a bank on a Price/Book approach makes sense, but only if we look at how the bank uses its “book” to generate returns.
Is Capitec expensive on a 7.7x multiple of its book? Is ABSA cheap on a 0.8x multiple of its book?
Well, a scatterplot tracking Return on Assets (ROA) versus Price/Book of our local banking stocks is massively revealing:
Once we place in a “line-of-best-fit” (otherwise known as a linear regression), well, all our banking stocks are merely fairly valued relative to their relative effectiveness in deploying their balance sheets.
(Note: Investec is a tricky, as half of the Group lies in the UK where UK-relative returns are better comparisons and, thus, perhaps this is unfair to Investec without splitting out that portion of the bank. Firstrand, too, has a big UK component while Standard Bank has a pan-African earnings contribution.)
What could change this rather efficiently priced set of banking stocks?
Well, if Capitec’s profitability starts sagging, its valuation versus book should follow. If the less profitable banks find efficiencies and start generating higher returns, their valuations should re-rate upwards too.
Finally, if the “environment” for banking changes (e.g. interest rates, economic growth or credit risk in South Africa), then this entire curve could rise or fall. I.e. Banks as a whole do better or worse and the sector re-rates.
In other words, our banking stocks are currently fairly valued looking backwards, but are our banking stocks fairly valued looking forwards? Leave your comments below.
ORIGINALLY PUBLISHED ON MONEYWEB.