This is Part II on the introduction to bank investing by Eric, a ValueUncovered reader who specializes in community banks and thrifts. Check out Part 1 for additional commentary on Centrix Bank & Trust (CXBT.PK), a $43m market cap bank based in NH.

Comparables

To be fair, the New Hampshire economy/housing market hasn’t had the turmoil that the rest of the country has had. In fact, the New Hampshire housing market never bubbled up much and has held up well.

I haven’t looked much in depth at the economy/housing/foreclosure situation in Centrix’s area yet, but I have looked at it briefly and have looked at their competitors. Housing prices really didn’t run that far, and thus they haven’t fallen that far either.

For example in Dover (one of the cities with a CXBT branch – see part 1 for additional branch information) housing prices went up about 57% cumulatively from 2000-2007, and then fell less than 20% from 2008 until now. This is a good site for housing prices per city:

http://www.nhhfa.org/demographic_housing.cfm

Click on price trends, then town/city reports, then click one of the six towns Centrix has an office in.

The great thing about banking is that all public and private companies can be compared/analyzed, since they all have to file reports with the FDIC, and anybody can get their info online, unlike other industries. I’ve included some basic information on CXBT’s competition:

CXBT Comparables

I did this for just 5 competitors but there are more I just haven’t recorded yet. I included just some very basic numbers of similarly sized banks focused in the same markets. The numbers just give one a sense of how quickly they’re growing (or contracting), what their returns are, and the quality of assets/loans.

I was curious to know how much of Centrix’s growth/asset quality was due to their market or their competitive position in that market. It appears that the vast majority of their growth is due to their competitive postion, while when it comes to asset quality it’s more due to their market.

However, keep in mind that Centrix had more new money to lend during the time periods I looked at so there was surely more imperative to lower standards.

Looking at their competitors and then looking back to how Centrix has moved in to new cities and areas close to them, they really have dominated.

It appears they have something depositors really love, maybe it’s good service, maybe it’s advertising or community involvement, but whatever it is they have brought in deposits very well as the expanded their offices.

Earnings Power

Since CXBT is very small, they don’t disclose many financial measures that bigger banks may disclose, like FICO scores or loan to value (LTV) ratios of their loan book. An analysis of a larger bank with better disclosure would look different, but that is just something you have to live with in the world of small banks.

The following table represents some historical income/balance sheet figures I use when trying to figure out earning power.

I also omitted Preferred dividends for 2009 & 2010

Earnings power under good scenario:

Assuming $700 million in average assets and $490 million in total loans.

Assume PPPT ROAA of 1.75% and no net chargeoffs/provision for loan losses.

Assume 37.5% tax rate and no preferred dividends

= $7.65 million in net income

Earnings power under worse scenario (but not stress scenario):

Assuming $700 million in average assets and $490 million in total loans

Assume PPPT ROAA of 1.45% and net chargeoffs (and thus provisions for loan losses) of .5%.

Assume 37.5% tax rate and no preferred dividends

= $4.8 million in net income.

Earnings power is 4.8-7.65 million dollars. Current market cap is $48 million + $7.5 million in preferred stock= EV of $55.5 million.

So multiple is 7.25-11.5. Although it’s probably about 9.

Over a decade, most of a typical community bank’s loan losses will come in just 3 or 4 years, so you need to normalize loan loss reserves over the business cycle to get normal earnings.

Net chargeoffs have been low over the last 10 years, and amount to about a dozen basis points as a percentage of total loans on average over the last decade.

Of course, a lot of net chargeoffs are likely ahead of them, which will raise the average, so I have included a good and bad normalized loan loss provision amount in my earnings power assumptions.

This earnings multiple (about 9) is about average for a bank. The average bank stock should trade at about 2/3rds the multiple as an average non-financial stock. KBW bank index, a popular large bank index, over time trades at this discount and if it gets well above (below) 2/3rds SPY’s multiple, then it’s a good probability that large bank stocks will do poor (good) relative to their large non-financial counterparts.

This discount is because of bank’s inherent leverage, which increases risk and therefore makes valuations lower to account for this. However, much higher quality and safer banks will deservedly sell at higher multiples on average.

Their returns on average assets (ROAA) are rather average as well. Their efficiency ratio, an indicator of how lean the companies operations are by measuring non interest income as a percentage of non and net interest income, is rather average (about 55%) as well. Like most community banks, their NIM’s are correlated with interest rates, so Centrix will make more money if/when rates rise.

Since their NIM’s and interest rates on loans are in line or higher with all their competitors, they aren’t offering higher rates to depositors, so their core deposit gathering capabilities seem real.

Negatives

The biggest negative by far is their recent leverage strategy.

They took about $7.5 million in TARP money, that carries a higher dividend that they must pay, higher than the yield they earn on most loans. To offset this, they pursued a leverage strategy, which consisted of acquiring to date about $100 million of easily obtainable brokered deposits, and investing the proceeds in GNMA’s.

This is classic yield curve arbitrage.

The brokered deposits are shorter term and carry a lower interest rate, and the GNMA’s are longer term and carry a higher rate. They are also government guaranteed, so no credit risk. The problem is, there is plenty of interest rate risk.

If rates move up their government guaranteed securities will lose value, and the deposits will eventually have to be replaced.

Many banks rationale for pursuing similar leveraging strategies is usually to reduce asset sensitivity. Since their assets reprice faster than liabilities, NIM’s (and thus earnings) tend to move with interest rates.

This is especially true for banks with great low and no cost deposit bases, which usually consist of a high level of transaction accounts. Many community banks are asset sensitive, and Centrix is one of them, but they are only modestly sensitive.

However, Centrix’s reason for the leverage strategy is apparently to compensate for the large dividends it pays on the preferred.

I’m not entirely sure why they took TARP money at all. They have always been well reserved and decently capitalized. I could be missing something there, although it’s likely management is just being overly conservative.

This is not what I like to see, I don’t want to see my bank playing around with their balance sheet. From the looks of it they aren’t going to unwind the “trade” anytime soon.

If rates shoot up for whatever reason, they will have to take huge losses to unwind the leverage strategy.

Furthermore, brokered deposits are from anonymous depositors secured through a brokered network, and it’s not a stable source of funding, so the strategy will have to be unwound at some point.

Brokered deposits have given many banks troubles –  I remember reading that IndyMac’s deposit base was 37% brokered. If Centrix really needs the $7.5 million in capital, as a shareholder I’d rather just see them dilute and raise equity than ratchet up the interest rate risk to the balance sheet this much.

Additional risks but not as serious:

Growth is more difficult to predict in banks than more traditional companies. They are more of a black box and you have to rely trends of past numbers, which can have a driving by looking through the rear view mirror effect.

However, I do see plenty of new areas CXBT can expand into, and management has done a good job so far allocating money to good loans. Plus, no growth is priced in, so it’s not a risk to the downside.

There is always the risk that as deposits keep growing Centrix stretches and lowers its lending standards just to allocate the money, but lending history shows none of this.

Conclusion

Centrix is modestly valued and growing quickly. They are an average community bank, except for the fact that they are very good at expanding into new markets to gather new core deposits.

They appear to dominate their local New Hampshire markets when it comes to gathering deposits and I see no reason why this wouldn’t continue. They’re partly fortunate and partly able when it comes to asset quality, as they have been helped by a stronger local market.

However, the leverage strategy in my view is too risky and speculative. Given their good core deposit base, I don’t like to see them leveraging with brokered deposits, and I’d rather see them dilute to raise capital.

Disclosure

No positions.

I’ve mostly avoided financial services and banking stocks as the analysis is typically outside of my circle of competence. One of my readers focuses on this particular area, and was gracious enough to share his notes and due diligence process for investing in small community banks.

While this write-up isn’t a buy or sell recommendation, it does provide a nice introduction to the thought process behind investing in banks. Please welcome Eric, an individual full-time value investor, for his wonderful insight and very thorough analysis (hence the two part series).

Introduction to Bank Investing

First, a word about small community banks. There are many different kind of banks, ranging in size from the gigantic money centered banks, to the more mid sized regional banks, to the smaller and in some cases downright tiny community banks. I like to focus more on community banks and thrifts (although I’ll buy any size financial institution) because I feel like the most inefficiencies are there.

The community bank market is ridiculously fragmented, and rightly so. Unlike most industries, banks are largely uneffected by globalization, and actually experience more advantages by being small and local. A local bank can easily compete against a giant like Bank of America, as depositors are more likely to deposit their money with their local bankers, who may be their neighbors that they see around the community at church or local high school football games.

They are also much more likely to try harder to repay the loan they borrowed from their local bank, as opposed to some anonymous giant firm from New York city. Local branches and relationships are expensive and difficult to develop for the giant banks.

Social reciprocity and in group human behavior help create these diseconomies of scale, and shield community banks from the efficient, giant, and menacing goliaths that devour other industries, like the Wal Mart bloodbath that many small and local grocers have experienced.

For these reasons banking world will stay fragmented. Just how fragmented?

There are well over 1000 publicly traded banks with under $1 billion in assets in the US alone, and that’s after the serious consolidation of recent years. Most of these firms are listed on the pinks or OTCBB, have little trading volume, and are too small for most investment funds, especially the larger/more successful ones, to invest in. Furthermore, these companies can be more complex, require more research, have business models that are pretty boring, and that can be very cyclical.

So we have a huge number of stocks that few people are analyzing or investing in, that are also cyclical, complex, boring, illiquid, and obscure.

If that’s not a recipe for inefficiencies and misvaluations, I don’t know what is.

There is also another interesting dynamic, unlike most tiny companies that plod along on the pink sheets, banks are heavily regulated and have to report to various government offices if they want to stay a bank. Some pink sheet stocks don’t file with the SEC and are of course lacking exchange oversight, so information can be scarce, unreliable, or in some cases non existent.

However, banks are regularly audited and examined by the FDIC, and other government agencies like the OTS. This means that you get constant quarterly financial statement updates for even the smallest and quietest banks. It also makes fraud a lot less likely, and total fraud completely unlikely. For good info on any bank, one can go straight to the regulatory mouth, by doing a FDIC bank find.

Centrix Bank & Trust (CXBT.PK)

Centrix Bank & Trust (CXBT.PK) operates a good old fashioned community bank in Eastern New Hampshire. It got my attention, not because it is selling at a discount to book value or earnings, but because of their blistering growth.

Current market cap is around $43m.

Below is a list of Centrix’s physical offices (all located in NH), when they were opened and some balance sheet numbers at their time of opening for the bank:

As one can see they are clearly expanding consistently into adjacent towns. All these are close to each other in eastern NH. If you’re really interested you can type those cities into google maps and see where each one is.

Asset Quality

CXVT’s asset quality is good. NPA’s (non performing assets) are completely covered by loan loss reserves, almost exactly. Due to limited disclosure it’s tough to dig deeper into NPA’s to see if this is as conservative as it seems, but it likely is.

It appears that the company has had a policy of taking at a minimum 30-40 basis points of total loans into loan loss reserves every year even if they have no net chargeoffs. Because of this, reserves were built up very well before recession.

The worst year for chargeoffs was 2010, about 23 basis points of total loans were charged off. The worst year for chargeoffs is likely ahead of them, either this year or the next, as NPA’s have gone from half a million dollars before the recession to $6 million now, and that will be worked through.

There will be chargeoffs but they are covered by reserves.

PPPT income is about 200-250 basis points of total loans depending on where NIM’s and noninterest expenses are, so there is good earnings power to offset unseen loan problems, which is important. Bad loans oftentake a long time to cycle through until they are written or charged off. A company with a lot of earnings power can quickly earn their way out loan problems.

A poorly reserved bank with a ton of earnings flowing in can often deal with problem loans and chargeoffs better than a well reserved bank with little earnings.

Considering CXBT’s blistering growth over the last decade, It’s good to see that management has been able to grow loans at a pace fast enough to keep up with the deposit growth they’ve had, and yet haven’t sacrificed lending standards, at least when it comes to loans.

The great recession was really a terrific test.

Take the 5 year period from the end of 2002 to the end 2007, a period in which the housing market really took off and bubbled up, Centrix’s deposits went from about $91 million to about $250 million.

With all that cash coming in it must of been difficult to find conservative yet attractive loans to make given the competition/conditions. Centrix managed to put all that cash to work and then some, increasing total loans from $71 million to $309 million at the end of 2007. Safely loaning out that much money during a large housing bubble would be difficult, and the start of 2008 until now would seem like the ultimate test to see if lending was truly prudent.

Given the low level of NPA’s and net chargeoffs Centrix has had, both absolutely and comparitive (I’ll get to that more later) to banks in their area, they were clearly never taking excessive risks or lowering standards just to put money to work.

Centrix is well capitalized, with Tier 1 capital ratio of 9.6% and total capital ratio of 11.9%, both are considered well capitalized in the eyes of regulators. Their balance sheet looks more leveraged because of a leverage strategy they’ve been employing (I’ll get to this later), and if they were to unwind that they would be more modestly leveraged.

Check out Part II…

Disclosure

No positions.

After reviewing the rationale behind closing my investment in SYK, this is the second part of my series on recent stock sales.

iParty (IPT)

While the investment in SYK turned out well, iParty (IPT) is an example of how several errors can compound to change the attractiveness of a position.

Through the first two quarters of this past year, IPT had turned in better than expected results.

In the initial writeup on IPT, I thought that the company was poised for a big year. Same-store comps were up 1.3% and 1.4% on a quarterly basis, putting the company on a strong path going into its busiest time of year – the Halloween season and fourth quarter.

Third quarter results showed a boost in the number of temporary Halloween stores – from 4 to 11 – that caused a  $1.9m loss in the 3rd quarter compared to $1.4m in the same quarter in the prior year, setting back the company’s progress.

Then, fourth quarter results did not show the expected jump. While the balance sheet improved, the company reported full year earnings of only $254k compared to $1.1m in 2009.

I try to avoid putting too much attention on quarterly results, but do keep a close eye on investments and evaluate them when appropriate.

Investment Review

With IPT, I made a mistake in this analysis on two fronts:

1) In the DCF valuation, I placed a much heavier weight on 2009 results – a high water mark for the company – vs. using the long-term averages. My upper estimates were predicated on a growth rate that was not supported by the historically lumpy results (as opposed to a company like SYK, where the trend line has continued upwards more or less in a straight line)

2) I did not include the significant number of preferred shares in my original EV calculations. Not my finest moment, but one of the consequences of learning on the job.

With this new calculation using the preferred shares, total EV comes out to $27.5m. This translates into an EV/EBIT multiple of 30x (!) using 5 year averages.

EV/FCF is slightly better due to the large depreciation expenses, but still high at 21x.

While the company has the backing of top tier investors, continues to expand its retail footprint and clean up its balance sheet, it probably doesn’t qualify as a value investment at these levels.

Conclusion

Warren Buffett said,

“When a management with a reputation for brilliance tackles a business with a reputation for bad economics, it is the reputation of the business that remains intact.”

Retailing is a tough business, and IPT is in an especially tough part of the retail market. 5 year average ROE of 4.06%, CROIC of 9.25%, and ROIC of 2.54% reflect this challenge.

My investment methodology has gotten more conservative, with a heavier focus on long-term averages, which therefore changes the viewpoint of IPT as a viable value investment.

With this in mind, I closed out my position at $0.30 per share on Feb 24 – I’ll take the gain and move on.

Final Thoughts

To be clear, my preference for long-term investments hasn’t changed (despite the whirlwind of sales in recent months).

It is important to grow as an investor, and I feel like my approach has changed significantly from some of the early analysis on this blog. As my philosophy matures, I’ll continue to make adjustments to the portfolio.

Learning from mistakes is a big part of that process.

Disclosure

I still own a small remaining position in IPT (1%) in my personal account.