Wednesday, April 17, 2013

First Citizens Bancorporation, Inc (FCBN)

Having gone through the majority of the banks traded on the Pink sheets, one thing I've learned is that community and smaller regional banks are not very good businesses. Granted, I have only looked at banks with readily available financials.

The issue I take with most of these smaller banks is that they're just too small. The economic value of a bank is determined in large part by the relative cost of its deposits. A small deposit base leads to a high cost of deposits (salaries, overhead, etc. spread over fewer deposits raises the cost of those deposits).  That may be obvious to most investors, but it became clear to me after seeing many banks with cost of deposits consistently in the 4-5% range with long-term treasuries and AAA bonds yielding 3.2% and 4.0% respectively. At current interest rates, a bank with deposits costing 4%+ relies on extending further along the "risk curve" or reduced capital levels to make any money at all. It's hard to come up with a conservative estimate of intrinsic value for such a bank. At least these are my current views.

That being said, I did come across a number of strong banks - such as NACB, FBTT, CNBA, JTNB, CLDB - but these looked richly priced. I also came across FCBN, which looked like both a strong bank with a reasonable price tag (recently $625).

Although I did not eventually get enough comfort to establish a position, I post my thoughts/analysis as a means to flush out my thoughts and to keep a record of my reasoning and also to hopefully learn to prevent future errors in judgement and analysis.

The analysis below is organized according three key areas of focus for analyzing banks:
  • Cost of liabilities
  • Return on assets
  • Asset quality and reserve adequacy
Cost of Liabilities
The higher the cost of liabilities, the fewer options a bank has to safely earn a profit and the more speculative the situation generally becomes. The main liability on a bank's balance sheet are the deposits. So keeping the cost of deposits low is the key to making money in banking.

Other liabilities are of course important also, but generally this money is provided to the bank at market rates so in a sense is an economic "wash", neither adding nor detracting directly to the value of the equity (an interesting discussion of this and bank valuations is given by Geoff Gannon here). This logic largely appears to hold for FCBN (i.e., although some of the rates on the debt and preferreds appear a little above current market, overall adjustments to equity would not be significant).

Below is a table showing my computation of "all-in" cost of deposits, as well as some other interesting statistics.


Dec-12 Dec-11 Dec-10 Dec-09 Dec-08 Dec-07 Dec-06 Dec-05 Dec-04 Dec-03 Dec-02
Avg core deposits 6,963.1 6,985.0 7,102.3 6,349.4 5,261.5 4,844.2 4,584.0 4,118.7 3,706.8 3,375.9 3,156.4
Non-interest income 119.8 110.8 153.7 221.2 88.7 80.4 69.0 66.3 58.4 57.7 52.6
Non-interest expense 253.5 246.9 242.6 218.4 196.2 181.6 168.5 152.8 144.3 131.8 125.9
Net-non-interest expense 133.7 136.1 88.9 -2.8 107.5 101.2 99.5 86.5 85.9 74.1 73.3
Interest expense on deposits 16.5 33.6 65.6 76.8 107.6 129.0 98.6 58.5 37.3 40.7 53.3
Adjustments (1) (2) 14.3 53.5 50.2 126.8 2.2 -1.2 0.3 1.2 1.7 1.7 1.7
Cost of deposits 164.4 223.2 204.7 200.8 217.3 229.0 198.3 146.2 124.9 116.5 128.4
Cost of deposits % 2.36% 3.19% 2.88% 3.16% 4.13% 4.73% 4.33% 3.55% 3.37% 3.45% 4.07%
AAA bond yield 3.80% 3.85% 5.04% 5.26% 5.05% 5.33% 5.40% 5.29% 5.36% 5.54% 6.17%
Spread 1.44% 0.66% 2.16% 2.10% 0.92% 0.60% 1.07% 1.74% 1.99% 2.09% 2.10%
Demand Deposits/Total Deposits 12.7 11.2 9.6 9.3 8.9 8.8 9.4 9.7 9.8 9.5 9.4
Peer 9.4 8.6 7.4 7.3 6.9 6.9 7.4 8.9 10.1 10.2 10.5
FCBN Percentile 71 72 67 67 72 69 68 55 49 50 45
Total Overhead/Total Assets 3.1 3.0 2.9 3.0 3.1 3.1 3.1 3.1 3.4 3.5 3.6
Peer 2.8 2.8 2.8 2.8 2.9 2.7 2.6 2.7 2.9 3.0 3.1
FCBN Percentile 68 63 59 63 61 73 74 69 73 73 71

(1) During and after the financial crisis, FCBN purchased a number of failed banks in FDIC assisted transactions, and simultaneously entered into loss sharing arrangements with the FDIC. Losses in excess of those expected on the date of purchase create a receivable from the FDIC which is offset by non-interest income. Also, FCBN booked a large one-time discount purchase gain in 2009 associated with the purchase of a failed bank from the FDIC. Despite many years remaining on these loss-sharing contracts, I consider these items to be non-recurring in nature and exclude them in my estimation of the bank's earning power.
(2) Removes realized gains/losses on sale of securities from non-interest income. In recent years FCBN has posted realized gains, which is not surprising given the generally downward trend of interest rates over the last decade. However over time I expect these gains to offset by losses when rates increase, and so I do not view this income to be indicative of future earning power. These are approximate for 2003 and 2002 because the figures were not available.

Comments:
  • Despite being skewed by the loss sharing arrangements with the FDIC, the cost of FCBN's deposits have consistently been favorable relative to market rates for money proxied here by long-term AAA bond yields
  • Observing the overhead/assets ratio, the relatively low cost of deposits seems to stem from the high high percentage of demand deposits, rather than exceptional operating efficiency. Non-interest income has also benefited from a modest amount of mortgage banking income (gain on sale of loan securitizations).
  • Including deposits acquired, average core deposits have grown at a healthy annually compounded rate of 8% over the previous 10 years; there is also not a reliance in any way on brokered deposits
  • As a percentage of total deposits, demand deposits have steadily increased in absolute terms and relative to peers
Return on Assets:

On the asset side of the balance sheet:
Dec-12 Dec-11 Dec-10 Dec-09 Dec-08 Dec-07 Dec-06 Dec-05 Dec-04 Dec-03 Dec-02
ROAA 0.75 0.71 0.90 1.54 0.82 1.03 1.05 1.09 0.90 1.08 1.12
Peer ROAA 1.02 0.87 0.49 -0.24 -0.16 0.97 1.24 1.29 1.28 1.26 1.28
FCBN Percentile 29 36 64 90 69 52 30 29 22 31 35
NIM 3.18 3.39 3.51 3.67 3.75 3.86 3.97 3.94 3.77 4.11 4.44
Peer 3.62 3.67 3.57 3.29 3.42 3.51 3.55 3.58 3.50 3.52 3.84
FCBN Percentile 24 32 41 66 62 65 67 63 57 72 73
Loan-to-Deposit Ratio 61.11 65.53 67.22 76.18 85.79 82.79 78.64 79.75 79.24 78.21 72.66
Peer 77.41 78.19 79.86 84.28 92.77 93.33 89.94 88.53 87.53 86.26 85.00
FCBN Percentile 23 23 25 32 30 25 24 33 33 34 26
Avg loans 4,426.7 4,662.3 5,138.4 5,165.0 4,478.7 3,931.7 3,652.8 3,308.3 2,956.4 2,586.3 2,319.5
Total int income 252.4 286.1 330.4 321.9 329.2 352.6 309.7 244.9 189.1 190.7 200.6
Loan losses 20.1 23.5 56.7 72.3 30.0 9.4 5.6 4.1 0.7 9.6 9.9
Net int income 232.3 262.6 273.8 249.6 299.2 343.2 304.1 240.8 188.4 181.1 190.7
NII/Avg Deposits 3.34% 3.76% 3.85% 3.93% 5.69% 7.08% 6.63% 5.85% 5.08% 5.37% 6.04%

Comments:
  • FCBN's ROA has generally been in the lower 1/3rd of its peer group
  • Based on the loan-to-deposit ratios, FCBN seems to have been consistently underloaned
  • Avg gross loans have grown 6% annually
  • FCBN has been consistently profitable through the financial crisis 
Asset Quality and Reserve Adequacy

Like getting comfortable with an insurance company's reserves, getting comfortable with a bank's reserve adequacy and credit culture is critical to evaluating a bank for investment purposes. This is tough, particularly because FCBN's accounting is somewhat muddled by the deals with the FDIC. Also, so far I've relied purely on financial statements and UBPR regulatory reports. This is not ideal for understanding the credit culture (better would be to talk to employees and competitors).

The table below contains the percent of total for of FCBN's loan book.

Loan mix % (including noncovered loans) Dec-12 Dec-11 Dec-10 Dec-09 Dec-08
Construction & Development 6.9 9.0 11.9 12.1 11.6
1-4 Family Residential 43.5 41.7 39.9 41.9 44.0
    Home Equity Loans 12.8 13.5 13.4 14.1 14.4
Other Real Estate Loans 25.5 25.6 23.5 21.7 19.7
    Owner Occupied Non-Farm Non-Residential 19.8 18.9 17.2 16.2 15.0
Commercial & Industrial Loans 6.4 7.2 8.8 8.0 6.7
Loans to Individuals 12.2 11.0 10.3 11.0 12.6

The higher than average concentration in riskier construction and development and home equity loans is noteworthy.

A few (certainly not comprehensive) credit quality statistics follow:

Risk Grades - Not covered loans Dec-12 Dec-11 Dec-10
Pass 3,747.06 3,750.70 3,933.53
Special Mention 145.94 156.69 161.04
Substandard 186.58 226.72 291.66
Doubtful 0.00 0.03 0.01
Loss 0.00 0.21 0.14
Total 4,080 4,134 4,386
Not Pass % of total 8.2% 9.3% 10.3%

Aging Analysis - Not covered loans
Dec-12
Dec-11
Dec-10
30-89 Days
Past Due
1.18% 1.06% 0.96%
90 Days or
More
Past Due
1.55% 2.15% 2.90%
Total Past
Due
2.72% 3.21% 3.86%
Current 97.28% 96.79% 96.14%
90 Days Or
More Past
Due and Still
Accruing
0.27% 0.28% 0.36%

Reserves - Not covered loans Dec-12 Dec-11 Dec-10
Gross noncovered loans 4,080 4,134 4,386
Nonaccrual loans 96 126 111
Avg impaired loans 137 147 107
ALL 63 71 82
Charge-off ratio 0.65% 0.76% 0.86%
ALL/Nonaccrual loans 66% 57% 74%
ALL/Avg impaired 46% 48% 76%
ALL/Gross loans 1.54% 1.72% 1.87%

Comments:
  • The risk grade and loan aging analysis both indicate a somewhat steady improvement in credit quality
  • Overall, the reserves ratios seem low relative to peers. 
    • For example, usually ALL/Nonaccrual loans is at least 100%. It's possible that this is well less than 100% because the bank - in conservative fashion - is quick to charge off loans. 
    • However, charge-offs are consistently lower than peers suggesting either better than average credit quality or slow recognition of credit deterioration
    • Further, ALL/Net Losses is also lower than most peers
    • For conservatism's sake, raising the reserves to 100%-150% of nonaccrual loans would reduce book value $40-100 a share.
  • The above comments combined with the relatively high concentration in riskier construction & development, and home equity loans is a little concerning to me
Other general comments
  • FCBN does not disclose an interest rate sensitivity table, making it difficult to assess the impact of a rise in interest rates. It is likely that the mortgage banking activities would be reduced.
  • The assumptions FCBN uses in its pension accounting seem quite optimistic. 
    • That is, they expect to earn an 8% after-tax return on plan assets with the 59.8%, 29.5%, and 11.6% of plan assets in equities, bonds, and cash respectively. 
    • Looking prospectively: with long term AAA bond yielding ~3.8%, and 3-month LIBOR ~0.3%, they're banking on an after-tax return of close to 12% out of their equity portfolio (generously assuming a 15% tax rate).
    • Looking historically: they have earned 4.8% annually compounded:
2012 2011 2010 2009 2008 2007 2006 2005 2004 2003 2002 2001
Beg of Year FV 104.5 99.6 83.3 71.2 82.4 75.7 65.7 55.4 49.9 42.6 41.2 39.1
Return 7.6 3.4 9.5 9.8 -13.7 4.1 5.2 3.4 3.8 5.4 0.1 0.9
Return % 7.3% 3.4% 11.4% 13.7% -16.6% 5.4% 7.9% 6.2% 7.6% 12.7% 0.3% 2.4%
Annual Return 4.8%
    • Berkshire Hathaway assumes an expected rate of return of 6.6%, with much of this capital being allocated by arguably the best investor that ever lived (and his protege).
    • Even though they're not allowing additional plan participants, their pension plan seems big enough that over long periods I fear these assumptions could have a pretty significant impact in terms of present value. Anther natural concern is that this optimism extends into other facets of their accounting or possibly their credit culture.
    • Their are probably more accurate ways to adjust for this, but my line of thinking is: I want to know how much in additional assets must be contributed today to reach the same future value under each set of assumptions. Assuming (1) an expected rate of return equal to their historical returns (a still generous 4.8%), and (2) a 20-year time horizon, this amounts to nearly $115 per share of assets today.
Summary and Overall comments

Overall FCBN has a strong deposit base and cheap funding, but has a few offsetting negatives including
  • Overcapitalized/underloaned: this provides cushion to cover losses, but this is at the expense of relatively weak performance on the asset side of the balance sheet
  • Accounting concerns
    • Possibly be under reserved 
    • Aggressive pension assumptions
  • Somewhat high overhead costs relative to deposits
From a quantitative standpoint, FCBN appears relatively cheap even after adjusting for some of the items described previously:

Current Shares 821,400
Deposits 6,963,127
Interest spread 1.40%
Pre-tax earning power/ Share 119
Deposit growth assumed 2.7%
Post-tax yield assumed 10.0%
Pre-tax yield assumed 15.4%
Per share value 940
Per share pension issue -115
Per share reserve issue -70
Net estimate 755
2/3 of estimate 503
Price 625

However, given the qualitative negatives discussed (which makes the valuation estimate dubious), I am currently not comfortable enough to initiate a position. I plan to follow this company and work to better understand the credit culture and asset quality.

Disclosure: None