January 17, 2013 - 8:24am EST by
2013 2014
Price: 169.00 EPS $17.00 $17.00
Shares Out. (in M): 10 P/E 10.0x 10.0x
Market Cap (in $M): 1,585 P/FCF 0.0x 0.0x
Net Debt (in $M): 0 EBIT 0 0
TEV (in $M): 0 TEV/EBIT 0.0x 0.0x

Sign up for free guest access to view investment idea with a 45 days delay.

  • Regional Bank


Thank you to Chuck307 for previously posting this idea, which strikes me as equally compelling today as when posted earlier, especially after disclosure Thursday (12/20/12) that the bank repurchased 5% of shares outstanding at a 6% discount to the market price.  The repurchase adds nearly $1 per share to earnings power, about $10 to intrinsic value, and over $2 to book value per share.  But I do not think it is widely appreciated.  The brief investment thesis to buy FCNCA is the following: (a) plain vanilla banking is a comprehensible business, (b) management is honest and competent and because the "founding" family owns over 35% there is a strong owner-operator mentality, (3) traditional borrowing and lending supplied by an FDIC insured bank is unlikely to suffer from material technology disruption or another form of obsolescence during our investment horizon, and (4) the stock price trades at less than 85% of book value and only 10x current earnings and earnings power based on a sub-standard 70bps ROA.  


First Citizens operates without many of the complexities that characterize modern banks.  This bank is a traditional borrower and lender.  Evaluating it for signs of complexity yields only a few yellow flags.  The bank holding company owns just one banking subsidiary; there is seemingly no "trading" activity (e.g. investment banking or capital markets activity) and the balance sheet does not include a worrying amount of Level 3 or speculative derivatives; the bank does not finance highly leveraged transactions; and, all noncovered loans are to customers domiciled within the bank’s principal geographic areas.

But there are a few yellow flags to highlight.  First, the bank’s operations are spread across an expansive geographically, even though most of the assets are in North Carolina and Virginia.  The founder's son (Lewis Holding, CEO until 2009) decided to diversify the bank beyond North Carolina in 1995, and the bank did so organically until the recent financial crisis when it acquired six failed banks from the FDIC.  Now First Citizens operates in 17 states and DC, which is more than all but three banks (excluding the four money centers banks which are more financial conglomerates than banks).  The sprawl is further illustrated by a few ratios.  First Citizens has the lowest ratio of assets-to-states of all banks over $10 billion (assets) and the ratio of deposits per branch is also low, even in the bank’s domicile of North Carolina, home to two-thirds of its branches (see charts below). 

Several larger banks, most notably KeyBank, attribute low returns to lack of branch density.  So it is reasonable to conclude that First Citizens has a dual return profile: the operations in North Carolina and some neighboring states like Virginia likely yield satisfactory returns on assets, but the operations in newer geographies with less branch density and size likely yield very unsatisfactory returns.  This provides the bank with a clear path to improving the overall ROAs. And it is probably why the bank has aggressively bid on six FDIC acquisitions.  Florida is one market that the bank might likely retreat from, especially because the press reports so many buyers for Florida banking operations. 

Despite the bank's expansiveness, most of the credit risk is taken in the original markets.  All loans are underwritten centrally, and if operations outside North Carolina (50% of loans and 16 states+DC) are excluded, then the bank’s assets per state is among the highest at $14 billion (see chart below left).  And west of the Mississippi, the bank has no retail franchise and operates as a niche player, focusing on its medical professional lending business.

Second, the bank takes different interest rate risk than most peers. Management prefers to take interest rate risk on loans to customers because there is a business relationship that can grow.  The bank started this practice about 20 years ago and mostly uses competitive loan prices on commercial mortgages to attract new customers.  Once the lending relationship has been established, the bank provides terrific service in order to generate additional revenue from selling other products to the same customer.  Many other banks start taking interest rate risk by purchasing longer duration securities.  But First Citizens does not because the cross selling opportunity is not possible with customers of securitized credit because securitization severs the relationship between the bank and the borrower.     

Given management's strategy, fixed rate loans with a maturity longer than five years comprise 32% of total loans.  This is higher than most other banks.  But to offset the longer maturities on loans, the securities portfolio is almost entirely short duration government guaranteed securities.  One way of illustrating the relatively short duration of the bank's securities portfolio is by looking at the yield.  The yield on securities is the lowest of all banks larger than $1.5 billion.  Specifically, the yield in 2Q12 was 0.73%, while the median of all banks was around 2.6%.  The yield at First Citizens is even lower than the custody banks. The interest rate risk taken on loans is further mitigated by the fact that deposits fund 100% of both loans and securities.  The bank utilizes very little term funding.  Less than one-quarter of deposits are time deposits.

A third source of complexity stems from distorted financial statements owing to six purchases the bank made from the FDIC.  The acquired loans amounted to one-third of the bank’s legacy branch originated loans.  GAAP necessitates the gains on these deals be amortized so currently book value is understated and earnings are benefitting from some non-recurring gains.  In fact, nearly every major line on the income statement requires adjustment to form a proper view of earnings power, as discussed more below.

In addition to having fewer complexities, there are several reasons to believe First Citizens is safer than peers.  The 10x financial leverage that banks employ is a source of potential catastrophic loss for equity investors.  But several factors suggest the risk at First Citizens is acceptable. 

  • The bank is family owned.  The family’s economic interest is 36% while it controls over 80% of votes.  The family gained control in the 1920s and the third generation is now in executive leadership (three executive officers and one director). The stake in First Citizens represents the bulk of the family’s wealth.
  • Management’s strategy explicitly favors safety over returns:  “Bancshares ROA and ROE have historically compared unfavorably to the returns of similar sized financial holding companies...we have consistently placed primary strategic emphasis on balance sheet liquidity, asset quality and capital conservation, even when those priorities may have been detrimental to short term profitability.” The Market does not discount FCNCA's earnings at a lower rate because of the safety these lower returns reflect. Moreover, the bank has a clear path to improving returns through greater branch density.   
  • The financial record is consistent with a culture of risk-aversion.  The bank did not lose money in the 42 years of data that is readily available.  The poorest returns since 1982 occurred in 2008 when the ROA was 0.6% and the ROE was 6%.  The ratio of tangible common equity to assets is only slightly above the median for peers, but adding the accretable yield to equity boosts the ratio by 1.0%.  Furthermore, liquidity is strong. The ratio of loans to deposits is 76% compared with peers of 85%.   

A permanent loss of principal would most likely arise from the bank's second lien mortgages.  The bank holds $2.3 billion of residential revolving mortgages, an amount equal to 20% of noncovered loans and 130% of tangible common equity.  Only five banks have a higher concentration: FHN (214%), TCB (186%), WBS (171%), FNB (152%), and HBAN (138%).  Revolving mortgages are risky because of the structure: typically, the loan is interest only for ten years and then principal and interest amortize over ten years.  While interest rates are low, debt service is quite manageable.  But if rates increase or the term hits ten years and the loans begin to amortize, credit costs are likely to rise.

Several factors make me less concerned about this bank's heloc portfolio.  First, one-third of the loans are secured by first liens.  And for the second liens, First Citizens is gathering updated information on first lien balances and home prices to determine whether their second liens are effectively unsecured (up-side down).  Management intends to use the information to boost monitoring on revolving mortgages with increasing credit risk.  Second, the bank originated all the loans based on standard criteria and no loans were acquired in the secondary market.  Over 90% of the loans are held by the bank’s customers in North Carolina and Virginia. Finally, the variable rates on revolving mortgages provide a sort of hedge to the credit risk. If rates increase, then debt service becomes more onerous and defaults may rise.  But as rates increase, the bank’s capacity to absorb credit losses will increase also because income from the performing book will increase. 

Thinking through a stress case, one can formulate credit costs by looking at peak losses this cycle for first lien mortgages and for credit cards.  For the one-third of revolving mortgages that are first lien, one may estimate stress credit costs equal to 150% the recent cycle peak for first lien mortgages, or 3.0% annually.  For the remaining two-thirds, one may estimate stress losses equal to 100% of the recent peak for credit cards, or 10% annually.  Given these loss assumptions and no change in rates, the bank would lose $35m after tax, or 2% of tangible common equity.  For this loan category to remain profitable for the bank, however, interest rates only need to increase by 2.0%.

A new scion at the helm is another risk.  First Citizens was run for 51 years by Lewis Royall Holding, who became CEO in 1957 when his father died.  No longer.  Lewis himself died in 2009.  Just before his death, his younger brother’s son, Frank Jr., took the helm.  It is reported that Lewis was an authoritative figure, and his nephew is unproven and certainly less experienced.  In a recent speech, the new CEO discussed banking and noted the following: (a) the bank has taken the road less travelled and it was not always easy, (b) risk aversion is part of our corporate DNA – one cannot turn on and off character, and (c) success in banking is about building long-term relationships with clients because the most important asset of a bank is the customer’s trust.   

The key profit drivers of profits are operating leverage and interest rates.  For most banks, the key driver of profits is credit costs.  But First Citizens has a remarkable record of low and stable loan loss provisions (chart below left).  The bank’s loan loss provision has averaged just 35 bps (% of loans) over the last 22 years.  That is nearly the lowest on record among banks and is on par with outstanding banks like MTB, CBSH, UMBF, and NTRS which trade at P/B ratios of around 1.5x.  In fact, the bank’s highest LLP of 66 bps is the lowest of the bunch.  The bank’s stable credit expense has contributed to stable returns on equity, but the level of returns is near the low end of peers (chart above on right).  Interestingly, returns on equity took a small step down in 1995 and continued at the lower level for the past 15-20 years (14% for 1983-95 and 10% for 1995-2011).  The two most likely factors are geographic expansion and lower interest rates, but it is not simple to establish these as causes.  


The family’s long record is fairly unblemished, as far as I can tell.  From appearances, this bank operates with care, the way banks should. The result is growth in earnings and book value per share at high single digit rates (the CAGR from 1982-2011 in EPS and BVPS is 7% and 10%).  The bank has not lost money in the past 40 years, but earnings have declined in seven of the past 30 years.  Out of 140 banks for which 20 years of profit data was readily available, about two-thirds lost money in at least one year.  

A strong credit culture steered the bank clear of the latest real estate (and construction) bubble.  Credit decisions are made centrally, although the bank only lends to branch customers.  While the largest concentration is commercial real estate, the underwriting focuses on the cash flow of the business operating in the property and not the value of the real estate collateral.  In my experience, some banks (perhaps including First Citizens) view commercial mortgages as business loans with recourse to real estate collateral.  Management avoids loans classified as commercial and industrial “C&I” because those loans are typically secured by just inventory or receivables, which are less reliable sources of recourse than real estate. 

The emphasis is on smaller businesses with sales of under $25 million.  While the legal lending limit is $275 million, there are only about 25 loans over $25 million and 3-4 relationships that exceed $50 million.  Loan officers are judged on credit quality not loan growth.  The bank utilizes a proactive monitoring process to catch and to address credit deterioration early.  Credit costs have averaged just 42bps for the past 30 years, and peaked at only 97bps in 1985. 

Real estate construction loans caused the bulk of credit losses for regional banks during the past decade.  First Citizens did not avoid this business. In fact, construction credits accounted for 7% of total loans in 2007 (management does eschew construction loans for large subdivisions).  But the median for banks in North Carolina at the time was 20%.  This suggests that management not only manages its exposures well (not too much construction exposure), but that the underwriters select risks better than peers also.

Management is fairly paid, but there is no variable compensation.  The bank compensates executives with salary and no pay at risk (variable compensation).  So salaries are high – the comp committee targets 150% of the peer median -- but total compensation for the top five executives is near the bottom of peers – though the committee targets 100% of the peer median (compensation for the top five executives during the past five years is #10 of a group of 11).  Compensation for directors is also modest (#9 out of the same group).   The rationale for not paying bonuses is unexplained, but I suspect it emanates from the family’s strategy of managing for the long run, which means not paying incentives because they are all overemphasize the near term.

Communication with Wall Street is a low priority, but management is accessible.  Management does not travel to meet investors or to participate in broker sponsored investment conferences.  But the annual reports are informative and the CFO welcomes visitors.  We were unable to dig up any salacious stories about the Holding family members.  There are not many press stories because employees (except for the top 2 executives) were historically forbidden from speaking to the press. 


Superior management is the primary margin of safety.  Given the financial leverage and opacity in large banks, there is rarely sufficient margin of safety in the ostensible price to value equation alone.  My subjective conclusion from a brief discussion with the CFO is supportive of the aforementioned objective factors that indicate a margin of safety including: (a) high family ownership, (b) priority of liquidity and safety above returns, and (c) more than 40 years without a loss.

Neglect is a secondary source of margin for error.  No Wall Street brokers rate the stock (FCNCA) according to Bloomberg and Thompson Reuters.  The $1.7 billion market capitalization is far larger than the float given that the family owns so much of the stock.  So the rush to buy bank stocks that will benefit from a recovery in the housing market has whooshed right past FCNCA.  The chart below illustrates that the correlation of FCNCA to other bank stocks is among the lowest in the bunch.

Earnings power of ~$17 suggests value in the range of $165 (0%) to $200 (+20%).  There are two methods that we use to derive earning power equal to $17: (1) historical returns and (2) adjusted 2011 earnings.  At 10x and 12x, the value range for FCNCA is $165 to $200.  With recent prices in the range of $160-170, we can purchase a conservative bank below fair value and expect fairly reliable long-term growth of intrinsic value in the high single digits. The factors that could contribute to higher growth are more likely, in our view, than those that would contribute to lower growth, at this point in the cycle and owing to the company's opportunity to improve its scale and hence returns. (The table below shows that the average ROA from 1996-2011 was 0.78% and +/- 1 standard deviation is 0.65% and 0.91%, for EPS of $14-$20.   

Returns based approach:  The average ROA since 1982 is 0.81%.  The average 10-year ROA has been as low as 0.72% and as high as 0.86%.  Based on the lowest 10-year average of 0.72% (2001-10) and assets per share of $2,068, earning power is $16 per share.  The average ROE since 1982 is 12%. The average 10-year ROE has been as low as 9% (2001-10) and as high as 14% (1983-92).  Based on the lowest 10-year average of 9% and an estimated book value per share of $198, earning power is $18 per share.

Adjusted current earnings approach: Unlike most bank’s, First Citizens is not under-earning today because of the accounting for banks acquired from the FDIC.  There are four main adjustments.  First, net interest income is overstated because the discount rate used to determine the present value of the future cash flows on acquired loans (15%) is higher than the interest rate on the loans (7%).  As a result, First Citizens paid the FDIC less than unpaid principle balance and the difference is amortized over the life of the loan.  Second, provision expense is overstated because most of it relates to further deterioration in the credit quality of acquired loans for which First Citizens will be reimbursed (roughly 80%) by the FDIC pursuant to the loss share agreement. Third, non interest expense is overstated, primarily because of gains on acquisitions from the FDIC, but also from other factors like the loss share reimbursement from the FDIC and regulatory changes.  Finally, non interest expense is overstated because of costs to workout troubled loans acquired from the FDIC. (The table below shows adjustments to earnings to derive EPS of 2009 = $5, 2010 = $15, and 2011 = $14. ***The adjustments below do not include reduction to the share count that has transpired which adds nearly $1 to EPS.

Relative valuation is convoluted.  Not surprisingly, the bank’s valuation looks attractive versus peers on a “Graham+Dodd” type P/E.  The G+D P/E is around 10x, which is comparable to BOH, TCB, JPM and PNC but far superior to STI, FITB, ASBC and MTB.  The chart below to the right illustrates that the trailing 7-year earnings yield of 8% is higher than 80% of observations since 1993.  

The bank’s ratio of price-to-book value is lower than 95% of the recordings since 1986, and today’s book value per share is understated by around 10%.  But the P/B ratio of peers is similarly low today and the peaks have been much higher for peers than for FCNCA.  So there is a major risk that the Market could bid up peers that take much greater risks to much higher valuations. 


I do not hold a position with the issuer such as employment, directorship, or consultancy.
Neither I nor others I advise hold a material investment in the issuer's securities.


    show   sort by    
      Back to top