Bank investors: what do you look for in an investment?

[oddballstocks]

I've recently been thinking about how I approach bank investing.  I look for banks cheap on a P/TBV basis and work out from there.  Once I determine something is cheap enough to my satisfaction I look at other factors, but cheap on a P/TBV is always my first step.

 

What are other approaches used?  For example is there anyone looking for growing banks, or improving banks?

 

Just curious as to other approaches.

[Tim Eriksen]

I've recently been thinking about how I approach bank investing.  I look for banks cheap on a P/TBV basis and work out from there.  Once I determine something is cheap enough to my satisfaction I look at other factors, but cheap on a P/TBV is always my first step.

 

What are other approaches used?  For example is there anyone looking for growing banks, or improving banks?

 

Just curious as to other approaches.

 

You approach banks the same way you other companies.  It makes sense and works for you.

 

As more of an EPV (earnings power value) investor I approach banks the same way as I do other stocks.  I believe that book value should not be the focus, rather earnings power.  I want growing earnings at a low price.  The goal is to have share price appreciation capture both the increased earnings along with the PE multiple expansion. 

 

I like to look at the efficiency ratio and look for banks that are growing and have true earnings hidden due to loan loss provision.  Efficiency ratio captures cost control, or to say it another way, gross margins.  I prefer high gross margins to lower.  There are tons of inefficient banks trading  below book, but are hampered with high costs, either operational or on the deposit side, or hampered with lower revenue opportunities. 

 

With loan loss allowance, many banks add 1% to their provision for new loans which in a low rate environment more than offsets the net interest income in that initial quarter since few are earning 4% net interest margins.  So I calculate run rate without the additional provision related to new loans to reflect true earnings. 

 

Having said that I own no banks right now because I haven't found any that fit that criteria at an attractive price.  I am sure there are some, I just have been focused on other areas of the market.  One reason is that long term I believe eventual rising rates will cause a lot of problems for banks with fixed rate loans.  We may have a bit of a repeat of the late 1970's where the cost of new deposits could be higher than older loans on the books. 

 

[frommi]

I really don`t know if my approach works in the long term ( and i am a fighter for spreadsheet investing ;D ), but i look at P/B and what could be a "fair" P/B giving the history of the bank and their current RoA. Then i look where book value can be in 4 years given their past growth (always thinking about if history can repeat here) and add the dividend yield. Out of the three components (diff to fair P/B+dividend yield+book value growth) i get my forward rate of return, which i compare to all other possible investments.

I value most non-moat smallcap investments the same way.

[Fat Pitch]

It depends on what you are trying to accomplish. Are you trying to buy cheap liquid assets or cheap future cash flows?

 

When it comes to banks looking at ROA is a great indicator of how effective management is at converting assets into earnings. Anything above 1% ROA is worth looking at, but be sure to adjust for one-time events.

 

Buying cheap liquid assets can work if there’s a catalyst to liquidate the bank or a plan to replace the incompetent management. If no catalyst is present then cheap assets can stay cheap for a really long time.

 

I find buying cheap liquid assets interesting, in a sense it provides you a large margin of safety (buying  $1 for 50 cents), but if the earnings power of the bank is terrible then that $1 stays roughly $1 year after year once you factor in ~4-5% inflation. It has the illusion of “safety”, but you are running against the clock to off load your holdings at a higher price to realize the gain.

 

Buying reliable future cash flows at a reasonable discount will result in that $1 growing larger and larger year after year without you having to sell. In this case time is on your side. Both methods work, but you just gotta go with how your brain is wired to think.

 

[oddballstocks]

Great responses so far, thanks!

 

Tim, your approach makes sense.

 

frommi - What discount factor do you use for years 1-4?  Do you discount years at different values or add in a terminal value?

 

Fat Pitch - You look for banks about 1% ROA, then do you look for anything else?  I followed the link to your site and read your writeup on Chesapeake Financial, do you usually use a DCF model to value banks?  What do you use for cash flow?  What's your discount rate?

[frommi]

No discount factor and no terminal value, that makes it easier. But just 4 years, because that way the difference to fair value still has a big enough impact and that is my intended holding period.

Something at fair value growing at 19% has for me the same forward return than something that has 100% upside but no growth/dividend. The logic for me is that after 4 years both have returned 100% when it plays out, no matter how that result was achieved. In my list the stocks that have good growth and a distance to fair value >50% rank highest.

[racemize]

I tend to use Tangible book value as an indicator of potential earnings power.  e.g., with C/BofA, a mediocre or average bank should be able to make >10% (e.g., 15% or more, really) on tangible book, so buying near tangible or below should provide upside, presuming they can return to average banking ability.

[oddballstocks]

I tend to use Tangible book value as an indicator of potential earnings power.  e.g., with C/BofA, a mediocre or average bank should be able to make >10% (e.g., 15% or more, really) on tangible book, so buying near tangible or below should provide upside, presuming they can return to average banking ability.

 

I have a question regarding this, I've heard 10% thrown around often.  But when I look at the average or mean for all banking indexes (KBW/Dow/NASDAQ/others) the average ROE is between 8-9% and ROA less than 1%.  The indexes are biased towards larger banks, if we looked at all banks in the us the average ROE is 7% and ROA is .84%.  My question is where does 10% come from?  Is that what a bank should earn in some normalized environment?

 

A bank earning 10% ROE is in the top 1/3rd of banks in the US, which is better than middling along.  Anyways curious about the 10% number, that seems to be the anchor metric in many people's minds, I was wondering where it comes from.

 

 

[KCLarkin]

This blog post has some good perspective:

http://basehitinvesting.com/wells-fargo-vs-small-community-banks/

 

 

[racemize]

I tend to use Tangible book value as an indicator of potential earnings power.  e.g., with C/BofA, a mediocre or average bank should be able to make >10% (e.g., 15% or more, really) on tangible book, so buying near tangible or below should provide upside, presuming they can return to average banking ability.

 

I have a question regarding this, I've heard 10% thrown around often.  But when I look at the average or mean for all banking indexes (KBW/Dow/NASDAQ/others) the average ROE is between 8-9% and ROA less than 1%.  The indexes are biased towards larger banks, if we looked at all banks in the us the average ROE is 7% and ROA is .84%.  My question is where does 10% come from?  Is that what a bank should earn in some normalized environment?

 

A bank earning 10% ROE is in the top 1/3rd of banks in the US, which is better than middling along.  Anyways curious about the 10% number, that seems to be the anchor metric in many people's minds, I was wondering where it comes from.

 

Well, perhaps I am misinformed, but I thought it was a decent estimate of long term averages.  Moreover, if great banks can make ~20%, I would imagine banks without any huge issues could make half of it.

[yadayada]

in general banks where this guy is not included

http://static1.businessinsider.com/image/53469a4b69beddc917c62d84-1300-/joseph-stilwell-letter-to-shareholders.png

[TwoCitiesCapital]

I've recently been thinking about how I approach bank investing.  I look for banks cheap on a P/TBV basis and work out from there.  Once I determine something is cheap enough to my satisfaction I look at other factors, but cheap on a P/TBV is always my first step.

 

What are other approaches used?  For example is there anyone looking for growing banks, or improving banks?

 

Just curious as to other approaches.

 

Great question. I was thinking about this the other day because SAN is one of my largest holdings at this point. I've been adapting my investment approach to compensate for behavioral issues. I'm trying to limit what I consider important to thesis and largely ignore everything else. Research suggests the more info you have, the more confident you are but you dont improve accuracy of forecasts. I wanted to pick 5-6 data points to synthesize from and ignore everything else. I think the points will largely change based on your thesis for the co.

 

I decided that my thesis for SAN is that it's a recovering bank with a global competitive advantage and that earnings power will greatly improve as loans losses come down. I chose the following metrics to pay attention to.

 

Is thesis true? Whats risk of being wrong?

1) Solvency - Texas ratio

2) leverage - assets/equity

3) loan quality - NPL/loans

 

Earnings Power and expected value

4) NIM (how much "revenue")

5) efficiency ratio (margins on that revenue and competitive advantage)

6) deposits (can earnings grow at low cost)

 

I'm thinking I'm only going to pay attention to the above 6 when evaluating the holding over time

[wknecht]

I usually look at book value except with deposits adjusted for their cost relative to market rates for money. Where cost is the "all in" cost reflecting operational costs. I also try to understand if liabilities are understated (reserve adequacy, pension), and guage riskiness on the asset side. 

 

Geoff Gannon did a writeup about this approach and it's also somewhat analagous to Alice Schroeder's cost of float valuation of Berkshire. It just makes a lot of sense to me and I think hits on a lot of key things (operating efficiency, potential earning power).

[Fat Pitch]

Fat Pitch - You look for banks about 1% ROA, then do you look for anything else?  I followed the link to your site and read your writeup on Chesapeake Financial, do you usually use a DCF model to value banks?  What do you use for cash flow?  What's your discount rate?

 

Once I identify a bank earning above 1% ROA I look into the sources of the revenue stream. Every bank can take in deposits and lend it out to earn that margin spread. I try to look for banks with other sources of reoccurring revenue such as asset management on the side or other fee based transactions.

 

Chesapeake Financial is a great example of a bank earning a decent portion of their revenue from fee based transaction. Yeah the efficiency ratio isn't so great, but the earnings has been steady in its growth over the last decade.

 

There's another bank, Trinity Bank of Fort Worth Texas. This bank only lends out money, but Jeffrey Harp is one heck of a banker. I think their ROA is around 1.5% and that's only on margin spreads in this depressed interest rate cycle. I can see the ROA increasing to 2% when we normalize down the road. Check out their annual letters, they are a good read. The stock is not exactly cheap, but you may get some motivated sellers down the road.

 

The discount rate is an arbitrary number. I just happen to use 10%. This number does not change based on the investment. The only number that changes is how confident I am in the business when I handicap the NPV of the DCF.  As for assigning growth rates on businesses, I am very hesitant in assigning anything above 8%. I usually do half the 10yr growth rates. Basically I try to get my returns on a high FCF yield.. any growth is just a hedge against inflation.

[Rabbitisrich]

I start with revenue to tangible assets and non-interest expenses to tangible assets.

[racemize]

I usually look at book value except with deposits adjusted for their cost relative to market rates for money. Where cost is the "all in" cost reflecting operational costs. I also try to understand if liabilities are understated (reserve adequacy, pension), and guage riskiness on the asset side. 

 

 

Can you give an example of your deposit adjustment?

[oddballstocks]

I usually look at book value except with deposits adjusted for their cost relative to market rates for money. Where cost is the "all in" cost reflecting operational costs. I also try to understand if liabilities are understated (reserve adequacy, pension), and guage riskiness on the asset side. 

 

 

Can you give an example of your deposit adjustment?

 

I'm not wknecht but a typical way to calculate deposit premium is as follows:

 

Cost of FHLB borrowing minus deposit cost to get the spread.  You calculate that forward 4-5 years and discount it back at the 10yr treasury rate.  That's in effect the deposit premium one should pay for the bank franchise.

 

In a transaction this deposit premium is usually called out in a news release.  You take the premium paid over book and divide it by deposits and you have the deposit premium.

[racemize]

I usually look at book value except with deposits adjusted for their cost relative to market rates for money. Where cost is the "all in" cost reflecting operational costs. I also try to understand if liabilities are understated (reserve adequacy, pension), and guage riskiness on the asset side. 

 

 

Can you give an example of your deposit adjustment?

 

I'm not wknecht but a typical way to calculate deposit premium is as follows:

 

Cost of FHLB borrowing minus deposit cost to get the spread.  You calculate that forward 4-5 years and discount it back at the 10yr treasury rate.  That's in effect the deposit premium one should pay for the bank franchise.

 

In a transaction this deposit premium is usually called out in a news release.  You take the premium paid over book and divide it by deposits and you have the deposit premium.

 

So, the adjustment he was referring to would be up from book value using the premium? 

[wknecht]

I usually look at book value except with deposits adjusted for their cost relative to market rates for money. Where cost is the "all in" cost reflecting operational costs. I also try to understand if liabilities are understated (reserve adequacy, pension), and guage riskiness on the asset side. 

 

 

Can you give an example of your deposit adjustment?

Conceptually, it's pretty much the same as oddballstocks described it. The idea being that everyone else has to pay for money but banks get it handed to them at low (relative to market unsecured borrowing rates) deposit or savings rates. But in addition to the interest they pay rent, teller salaries, loan underwriter salaries etc to gather them. Dividing these deposit costs (deposit interest expense and net non interest expense related to deposit gathering) by average deposits gives their cost rate. Can look at this over multiple years.

 

Mechanically, you can then make the adjustment as a spread calculation the same as what oddball mentioned, or think of it as marking a bond to market (where the "coupon" is this all in cost). Can also consider deposit growth (I'll usually look at the rate the last 5 years and take 1/3rd).

 

The adjustment will lower liabilities if their cost is less than market rates and raise liabilities in the other case.

 

Oddball, what's the rationale for using FHLB rates versus say, treasuries or highly rated bond yields (assuming the deposits are sticky)? Capital raising constraints?

[topofeaturellc]

I think that's a pretty thoughtful way to do things, but I think if you believe that non-deposit banking is a CoC biz on marginal capital invested the spread of ROTE over your CoE is already including the value of that

 

Basically I think you'd find that = adjusted book*market ROE = tangible book*historic company ROTE

 

I wouldn't expect true deposit funded banks to have really different costs of funds over the long-term. Where there might be difference is in deposit growth rates - although I suspect the banks know that better for us and are more competitive with one another in markets with population growth vs markets that don't really grow - so it nets out wrt to valuing future growth.

 

I'd generally say I'm a normal earnings power guy, but I think the case of banks the earnings power is derived from the balance sheet and especially the funding side of the balance sheet, so here I think Book is a pretty good way to think about valuation. 

 

I also think you need to underwrite the asset side of the balance sheet pretty thoroughly to figure out what magnitude of loss can permanently impair the equity. 

 

At the end of the day I think deposit banks have show a remarkable ability to reprice their assets and liabilities to earn a cost of capital return on marginal capital.

 

[Spekulatius]

I look at asset quality and asset quality trends first. i won't buy anything with a high Texas ratio. Typically a 40% Texas ratio is my limit. I look for a combination of low PE and low price/tangible book ratio. I look at pot. Vulnerable assets like construction loans, how assets held up during the great recession and generally I like to see the potential to earn at least a 1 % ROA, without risky endeavors. I think nowadays, one should look at the duration of their loan and securities portfolio as well (interest rate risk).

[Junto]

Outside of the common metrics, price to tangible book, price to earnings, margins, asset quality, etc.. I look for these specific things in regional to smaller banks:

- Niche Lending Focuses (Think NYCB, TAXI, CBSH, )

- Growing Economic Areas (Banks are generally a reflection of the strength of their market area, the better local market place, the higher chance of better returns)

- Deposit mix is of particular interest right now forecasting into a rising rate environment

- Dividend Yield and historical consistency to maintain or grow it, speaks to accountability of management in my eyes

- Management (what can you find about their specific knowledge, team, and performance in the market area they serve. There are a lot of sleepy managers out there that appear to slowly float to the top just like in other industries)

 

I also like to look for characteristics that could lead to it to be an acquisition candidate or alternatively, I like buying proven acquirers (RNST, OZRK) who have demonstrated proficiency in consolidating operations while growing loan demand in the new markets. I also like teams like SBNY has established whereby they can solicit and sign on teams of relationship managers to move business (note I like their strategy but have not been an investor in their stock).

 

[BG2008]

TAXI is interesting.  It is a bet on NYC Taxi/limo commission's power versus Uber and Lyft car rides.  NYC taxi medallions are known to trade for over $1mm each.  If Uber and Lyft are allowed to roam free in NYC, those medallions will be worth a lot less.  TAXI's collateral may collapse in value and the whole company can go under.  It's amazing that three years ago, taxi medallions are viewed as recession proof assets. 

[Junto]

TAXI is interesting.  It is a bet on NYC Taxi/limo commission's power versus Uber and Lyft car rides.  NYC taxi medallions are known to trade for over $1mm each.  If Uber and Lyft are allowed to roam free in NYC, those medallions will be worth a lot less.  TAXI's collateral may collapse in value and the whole company can go under.  It's amazing that three years ago, taxi medallions are viewed as recession proof assets.

 

I think arguing the company could go under is a little aggressive. The values may reduce but the company itself is not highly levered against the medallions and the management has been proven operators in the space. Here is an article of interest. I went back into TAXI after selling a couple years back today after earnings.

 

http://seekingalpha.com/article/2352885-survey-shows-the-uber-story-is-overblown-taxi-remains-dominant