Evaluating Commercial Banks - Earnings (part 2)

Evaluating Commercial Banks – Earnings (Part 2 of series)
This post is the continuation of a series I started with my introductory post on commercial banks, which can be found here.

This post is going to start the analysis part of the series. I'm basically just going to look at a bank's 10k and talk through what I'm look at and my thoughts on it. We’re going to start by taking a look at bank earnings. As I stated in my first post, this is going to focus on commercial banks (deposit-taking banks). This is to be distinguished from so-called “universal banks”, which have broker/dealers, insurance companies, or other entities, in addition to commercial banks, that can cloud the analysis, or other types of specialized banks (purely credit card banks, custodian banks, etc).

It’s important to understand a bank’s corporate structure before starting the earnings analysis. Almost all US banks are structured with a holding company (whose shares are listed on the stock exchange, but doesn’t really do anything) that owns shares of the commercial bank and maybe a few other subsidiaries (in the case of JPM, C, etc, this can be many other subsidiaries). The 10K is going to present the info on a consolidated basis and it’s often worth checking the bank’s call report (https://cdr.ffiec.gov/public/) from the FDIC to make sure the other subsidiaries aren’t too big. Foreign banks can be a lot trickier, but I won’t really get into that.

The bank I’ve chosen to look at is KeyCorp (NYSE: KEY), whose main bank subsidiary is KeyBank NA. They do have a broker/dealer (KeyBanc Capital Markets) but this is pretty small (about 3% of assets) and shouldn’t make analyzing the 10K much different than using just the FDIC call report on the bank. They are a pretty decent sized regional, but this analysis is going to be the same basically for anything with just a few branches up to a large regional. I haven’t really looked at this bank before, so I’m basically going to go through what I’m looking at here to start. Key’s 10K is here: http://investor.key.com/docs.aspx?iid=100334 and I might also refer to some information from their FDIC call report from 12/31/12 (I’ll be using SNL for this – a data subscription service – but you can see this at the site I linked to above if you’d like).

First, it’s to page 110 to check the auditor’s note (looks good). Before I look at the I/S, I need to get a picture about Key so I know what to expect. Key’s a large regional bank ($89.2Bn in assets) that operates 1,088 branches (pg4) in 14 states. I note from their website the bank is rated A- by all 3 agencies (holding company ratings are usually a step lower), which tells me that I probably shouldn’t expect too much out of the ordinary, may be 1 or 2 minor issues (a bank this size can probably get A+ or A max). I’ll look at these if I’m doing an analysis at work, but will avoid them for this.

On page 5, it looks like the majority of this is in the Northwest (1/3rd of deposits) and in the Northeast/Great Lakes (2/3rd). They are based in Cleveland. Just by using google maps (confirmed with SNL), I can see that they have most of their branches in CO, ID, UT, OR, WA, OH, IN, NY, VT, and ME. Just based on this, I already know that I should expect their asset quality / provisioning to be a bit more stable than if they were located in the South or CA. For the future, they are exposed to a pretty good set of markets for growth. The Northwest is seeing pretty favorable population increases and the NY/OH markets are exposed to the shale gas boom that is helping some real estate and corporate growth. The Northeast/Mid-Atlantic is also well known to have a bit more stable property markets, though banking competition is high. The Northwest is less heavily banked, and Key can probably make some good headway there.

Going to skip down through the 10K a bit, the risk section for banks is basically identical bank to bank.

I want to see what their average asset/liability structure looks like so I can form some expectation about earnings. I notice on pg40 they give some goals – I’ll talk about some of this later. Pg 49 is where the goodies are at (yield analysis chart / ave bal sheet). Net interest margin (the difference between the yield on assets vs. liabilities) is slightly up from last year (3.21% from 3.16%), which is a bit against industry trends, so let's see why. This is also a bit low (would expect more in the range of 4%). NIM has been tough for banks b/c liabilities (deposits) are basically as cheap as they can get right now and reinvestment on the asset side keeps going lower (low rates on Ts/MBS and high competition for loans). Looks like part of the reason for their low NIM is they have a decent number of time deposits and CDs that are pretty darn expensive compared to their other deposits (deposits are the best / cheapest type of funding). They are winding them down, but it's kind of silly that they offered these in the first place. Going to guess they had some issues attracting funds in the 08-09 range and resorted to some of this. They are also reducing debt and upping checking/savings deposits, which is good. Probably a bit more upside on the liabilities side for better int expense levels moving forward. I'll talk about this more in the liquidity/funding part, but you typically want to see banks fund themselves with deposits and a little bit of long term debt. Other types of funding are more expensive and/or less stable than these sources. They also have 11.8% equity to assets, which is going to make earning their cost of capital a bit more difficult (7-10% or so is more typical).

On the asset side, 20% of their assets are securities, which is maybe a bit high (15% is a bit more typical), and can somewhat explain the lower than expected NIM. Asset choices like this are a choice between liquidity and yield. Securities yield less but can be sold easier than loans (in Key's case they get a bit less than 3% on secs vs 4.3% on loans). Looking at the loan side, 70% commercial loans is kind of high (this can vary but I usually expect 50-50 or 60-40 comm to consumer for a bank like this). The upside to this is that it can help you do non-interest income business easier. Companies tend to give this kind of business (treas management / trade / custody / AM / etc) to banks that lend to them. Looks like they do a decent amount of comm real estate which can help the yields on comm loans, but you can plainly see the 4% they get on comm vs. 5.2% on consumer is holding NIM back. Hardly any residential mortgages too, which either means we're going to see a decent amount of gain on sale in non-interest income, or they just don't make many mortgages (which would be strange with 1100 branches). Also looks like they are winding down some areas of lending (marine/construction) which probably got them into trouble in the crisis. Overall, seems like the lower NIM (3.21) makes sense – little bit less than ideal liab structure, corp over consumer, and a bit high on securities. Net int income was dead flat – assets falling a bit offset by the NIM gain. Also looks like one of their goals is a 3.5% NIM, which might be obtainable over time.

Heading to the I/S on pg 38, can see they do make a pretty nice amount of their income from non-interest (almost 47% of total revenue (int+nonint excl provisions). This is good because non-interest income sources tend to be a bit more stable than interest, don't require you to take credit risk, and don't require as much capital. Levels for non-int/revenue are roughly 50%-great, 40%-good, 30%-meh, 20%-wtf.

Moving to pg52 to take a look at non-int income, looks like their corp focus in lending helps them do some pretty decent letter of credit business (trade finance) and a bit of IB/CM. They also do okay on service charges (overdrafts, checking acct fees, etc). AM/Custody from their note on 53 seems to be hurting a bit. Fees have only been down a little, but the sale of the MMF business and only slight increases in the other areas (when markets have been good) make me think they aren't doing too hot. Not too surprised to see they are getting out of sec lending (custody-related business). Custody is a tough business for any not-huge bank to do, as it's a scale business that's low margin. Might see declines in fees here moving forward. Looks like they are winding down leasing, which will also remove some expense, so net-net probably not too big. Hardly any gains on securities sales (which I would typically adjust for when looking at non-int / rev & efficiency), which is kind of strange b/c fixed income has been doing well. Looking at OCI, they went from a bit of an unrealized gain to a loss, so they kind of missed their chance to take some gains (though this could hurt int income so it's kind of tough to say). I was expecting to see a lot more than 150 from loan sales (mortgages they originate then sell to Fannie/Freddie). On pg69, they only sold off $1.8bil of res mortgages, which is really low for a bank that has 1100 branches ($1.6mil per branch in mortgages). Weird considering mortgage business was awesome in 2012 for most banks and it's a good business to be in generally, because the pay is basically all commission (variable). Not a fan of this. Commercial RE looks better at least and they hang on to some of these on BS. Overall, might see some reductions here moving forward due to lower loan sales and some lost AM/AC business, but non-interest inc should continue to be okay moving forward, especially if the economy picks up a bit more and they get some more trade business / IB/CM.

Provisions taken this year (229) are low (5% of rev) and we need to see if this is justified or not (provision levels are easy to manipulate by management to boost earnings artificially). To do this, I take a look at the reserves on the BS, which are built by provisions (pg112 – 888mil) and compare it to their non-performing assets (pg96) which is their non-performing loans (90 or 120 past due depending on type), their foreclosed RE (OREO), and a bit of other stuff. In addition to the 735 they give, I'm going to add the additional 71 of restructured loans not counted here, because those have already gone bad once, and might easily go bad again. So, 888/806 gives 110% coverage, which is consistent with where they've been historically for coverage (usually 100-110%) and is above 100%. 100% is usually where we like to see reserve coverage because it means they have the ability to write everything off that's gone bad so far without hitting capital any further. In some cases, this can be lower (if they have a lot of collateralized loans) but for a bank like Key that doesn't do a ton of RE lending, 100% is probably where it should be.

Moving to expenses, everyone always does non-interest expense to revenue (int+non-int inc not counting provisions) – called the efficiency ratio. 50% is great, 55% is quite good, 60% is about average, 65% is pretty bad, 70%+ is bad. This sort of depends on the bank and their businesses (typically things like AM/custody can make this go higher), but Key should probably be around 60%. Looking at their I/S though, they've been at 67% for the past few years, which is pretty gross. Their goal is to get to 60-65%, which is good to see, but they've got work to do. Looking at pg55 kinda hard to tell why it's so bad. Not like they are using a ton of consultants or have major legal fees (other is kind of high, but not sure what they are putting here). They basically just need to do better on comp and get more with less. Not impressed here at all though.

Putting it all together, return on ave assets was 1% (net inc / ave assets), which is pretty good but maybe only has a bit of upside if they can expand NIM a bit. Provisions can't really go any lower. You'd hope they can boost this via cost control, but due to their bad track record here, I'm skeptical. ROAE is 8.5%, which is probably below their cost of capital (usually around 10-12% for banks). Again, their high-ish level of equity isn't helping. We'll see later if this is totally necessary or not. The market is putting them at 0.86 price to book equity value. Pretty typical for a bank that doesn't earn their cost of capital w/o a ton of upside for better earnings. Heading forward, I'd expect to see low credit costs (though we'll talk more about asset quality next post), pretty stable non-int, maybe a bit of upside in int inc, especially if they can grow loans, but the main key is expense control here.

TL;DR / Earnings Checklist
-Where is the bank / what does this imply about their franchise?
-Any Reg Enforcement Action / Credit Rating Red Flags / Litigation?
-Any recent changes in strategy / M&A?
-What's driving NIM / does it make sense with their asset / liab mix?
-How is non-int income doing / is it stable (trading gains or long term business)?
-Is the provision level justified? (forward-looking about the provision involved asset quality evaluation)
-How's efficiency / expenses?
-What's management's strategy if any of this isn't going so well?
-How are the ultimate returns / are they meeting capital costs?
-Growth opportunities?

Feel free to post questions.