|Shares Out. (in M):||10||P/E||16.0x||N/A|
|Market Cap (in $M):||1,795||P/FCF||N/A||N/A|
|Net Debt (in $M):||0||EBIT||0||0|
First Citizens Bancshares, Inc. - $172/share - Groundhog Day 2010 - Ticker FCNCA/B
Over the past few months, I've been migrating my portfolio from some of the juicier stuff offered twelve to eighteen months ago into conservative, stable investments that offer long-term double digit return potential albeit perhaps at the expense of not being short-term multi-baggers.
First Citizens BancShares is a family-controlled Raleigh, North Carolina based bank holding company (BHC) that just acquired its third loss-share via an FDIC assisted transaction. It trades for 1.1x stated book, though I suspect book will accrete faster than normal over the next few years implying today's purchase price is really at or slightly below book. I've been an owner since this fall and it seems more clear to me now than ever that this conservatively managed bank is exceedingly well positioned to expand its deposit franchise both organically and via acquisition (the latter method being self evident, at this point). I believe downside is fairly limited.
The BHC refers to itself as "BancShares" and is a holding company that operates two primary brands: First-Citizens Bank & Trust (FCB) as well as IronStone Bank a federally chartered thrift ("IronStone" - a solid sounding name if there ever was one - not simply iron, not simply stone, but ironstone). In any case, BancShares is a well positioned BHC with a long track record - it's more important brand, FCB, was founded in 1898.
The FCB arm has been the acquirer in each of the three FDIC assisted transactions - Temecula Valley Bank, CA on 7/17/09, Venture Bank, WA on 9/11/09 and First Regional Bank, CA on 1/19/10. I believe most FDIC assisted transaction are low risk and most offer attractive upside.
By my tally, only two banks have participated in more than three FDIC assisted transactions since the onset of the crisis, both of which are privately controlled banks based in Minnesota (each getting four deals). Not only has FCB acquired three failed banks, it has done so in an incredibly risk-averse manner. It has acquired $4.54 billion of assets (before acquisition accounting fair value markdowns) but did so with loss-sharing agreements with the FDIC covering 88% of those acquired assets, significantly limiting BancShare's downside. This compares to less than 80% coverage for the typical loss-share backed transaction.
Along with those $4.54 billion in assets, BancShares acquired $4.1 billion of deposits and deepened its footprint in two geographies that it was already involved with: Southern California and Washington's Puget Sound counties. Both face obvious cyclical headwinds but have long-term secular tailwinds. Also, each is likely to be an ongoing epicenter of bank failure, leading to incremental opportunity for FCB. [As a brief aside, I think the Pacific Northwest is going to be the Georgia of 2010]
The acquisition of $4.5 billion of assets and $4.1 billion of deposits compares to its June 30, 2009 balance sheet of $17.3 billion of assets and $14.4 billion of deposits. This means BancShares added around 25% to each assets and deposits during the past seven months (June 30th is used because it's the last balance sheet that predates any of the three transactions). This was all accomplished without raising any new capital or taking on much in the way of incremental risk.
To show you the limit of what can happen, the FDIC assisted East West Bank of Pasadena, CA in acquiring United Commercial Bank in November 2009. East West had $12.5 billion of assets and about $10.5 billion of deposits beforehand and added $10.2 billion of assets and $7.5 billion of deposits (with $7.7 billion of loss-sharing). East West Bank thus added 82% and 71% to assets and deposits, respectively. That implies that from the perspective of the FDIC, BancShares has ample flexibility to continue acquiring.
BancShares has not paid a deposit premium for any of the acquired failed institutions.
Capitalization: 8.76 million A shares and 1.68 million B shares with identical economic value, but class B holding 16 votes per share. The overwhelming majority of equity is tangible equity.
Family Controlled: The Holding family has controls and runs FCB. They and their family and trusts continue to own well in excess of BancShares votes. They pay themselves fairly - only three employees received over $1 million in total compensation during 2008, none received over $2 million. Two of those employees, Lewis Holding (Chairman & CEO $1.9 million) and James Hyler Jr. (Vice Chairman and COO $1.1 million) retired in early 2009 after 40 and 25 years of service, respectively. Frank Holding Jr. ($0.6 million and 25 years of service) was named Chairman and CEO. His father is the Executive Vice Chairman and has 40 years of service ($1.9 million). Neither retiring executive received any sort of unusual retirement benefits. With a $1.8 billion market cap and insiders owning more than half of it, it's fairly clear that they make money when we do: through building and distributing value to shareholders. No golden parachutes exist.
The company has not issued new shares in many years.
Well Capitalized: If the fact that in the past seven months the FDIC has allowed BancShares to make three acquisitions - including one last week - doesn't provide a hint that BancShares is well capitalized, then nothing will.
BancShares refused TARP money and is considered well capitalized by virtually every objective standard. The company ended 2008 with 13.2% Tier 1 capital ratio, 15.5% total risk based capital ratio and 9.9% leverage capital ratio. According to a recent 8-K, 2009 ended with 13.3%, 15.6% and 9.5%. The FCB subsidiary was 12.7%, 15.1% and 8.7%. Each of these is well in excess of the minimum requirement to be considered well-capitalized (6%, 10% and 5%). While all details haven't yet been released by BancShares about the most recent acquisition (First Regional Bank on 1/29/10), I expect it will not meaningfully negatively impact either BancShares' overall or FCB's specific capitalization scores. BancShares also has a TCE ratio in excess of 8% (calculated before the most recent transaction).
BancShares continues to increase allowances for losses faster than chargeoffs are coming through, despite the fact that both metrics have begun moderating on a quarter over quarter basis, creating some hope that the worst is behind BancShares. The conservative reserving and modest trend improvement both provide hope for reserve release at some point in the future.
Statements like the following one, from a recent 10-Q, provide a qualitative reflection on management's conservative approach [emphasis added]:
"Financial institutions frequently focus their strategic and operating emphasis on maximizing profitability and measure their relative success by reference to profitability measures such as return on average assets or return on average shareholders' equity. Historically, we have placed primary emphasis upon asset quality, balance sheet liquidity and capital conservation, even when those priorities may be detrimental to short-term profitability."
Operating Performance: Over the past fifteen years, BancShares has earned approximately a 10% ROE. That average has been pulled down somewhat by the last several years as the bank was around an 11.5% ROE business for many years through 2000. BancShares earned $11.08 in CY09 vs. $8.73 in CY08, however $6.12 of 2009 was due to acquisition gains from the two 2009 transactions. Book value is approximately $150 per share (tangible book value is about $140 per share).
Chargeoffs in 2009 on non-loss-share assets was 0.56% vs. 0.40% in 2008, though it declined modestly in Q409 at 0.50% vs. 0.62% in Q408. Provisions grew to $77 million vs. $66 million in 2008.
PPOP in 2009 was around $154 million off from $205 million in 2008. Much of this decline is to be expected because as BancShares takes over more troubled assets and new banks, it increases its cost structure. Initially, however, it does not proportionally increase its interest income due to the fact that a large portion of those acquired assets are non-earning assets until they are worked out and the cash is redeployed into earning assets. PPOP in the fourth quarter was somewhat higher than the 2009 annualized rate at $48 million (or $192 million annualized). I expect that over time, it will continue to grow.
Recurring profitability, over time, will expand substantially as the $4.5 billion of assets acquired is redeployed without new equity needing to be raised. More accurately, the fair value of the assets acquired is closer to $3.5-4.0 billion (we don't yet have that detail for First Regional Bank), so the redeployment opportunity is probably more like $3.5-4 billion. In essence, BancShares has increased its loan portfolio and deposit franchise without needing to raise fresh equity. In fact, the growth has come without any meaningful change in BancShares risk metrics because each transaction has been immediately and substantially accretive to book value (described below in the Assisted Transaction section). As such, incremental spread will accrete undiluted to owners. This will improve returns to equity substantially. Over time, I'd expect an incremental $40-50 million of annual income (and growing) from these acquisitions.
I estimate that BancShares could earn in excess of $25 per share on a normalized basis in three years as assets grow from $21 billion and equity grows to over $2 billion. Upside optionality exists in the form of further assisted transactions, faster deployment of excess capital, sustainably higher NIM, and better than expected charge-off experience which releases reserves back to owners.
Steep Yield Curve: While my assumptions do not project substantial NIM spread percentage expansion, it is clearly quite possible. The yield curve is approximately as steep as it has ever been and the lending environment remains very favorable to lenders. This combination amounts to a great environment for banks to operate in, at least for new business. Legacy business obviously remains a challenge for most players as they deal with the less attractive book of assets from the several years leading up to the crisis. As older loans mature and cash is redeployed in a more attractive lending environment, it seems reasonable to expect that this roll leads to average profitability growing over time, even if the size of the asset base stayed the same. I expect that assets will continue to grow as BancShares continues to deploy some of its excess capital and it hunts for additional transactions.
Integration Risk: As with most banks that are larger than $10 billion of assets, BancShares has some history with acquisition and integration. Since 1990, but excluding the three recent assisted transaction, BancShares has acquired nine companies. It wisely stopped buying banks in 2003 as quality and price both degraded. Rather than acquire overpriced banks during the past ten years, BancShares has focused on organic growth of FCB and it has developed its IronStone brand growing it from its founding in 1997 to a $2.1 billion thrift today. I believe the BancShares management team will comfortably manage the integration of the FDIC assisted transactions.
How Does An Assisted Transaction Work: When an FDIC-insured depository institution fails, the FDIC typically conducts an auction to find the highest bidder (actually, to find the "least costly" solution). Some people bid for the entire bank including all assets and liabilities while others bid on pieces. The FDIC seeks to minimize losses to its Deposit Insurance Fund (DIF). This is the normal process.
Historically bidders want to avoid buying bad assets for two reasons: 1) they require management, so there's a personnel cost; and 2) they tie up capital that could otherwise be productively deployed in performing assets. However, during the ongoing crisis, the FDIC was finding that so many bidders were excluding such a large number of assets from their bids that the FDIC began promoting an option to buyers called a "loss-share" agreement where the FDIC covers 80% of losses up to a pre-determined threshold and then 95% of losses beyond that threshold.
In practice this works approximately as follows: the FDIC sets a loss threshold for bidding purposes. This is the threshold where the loss-share triggers from 80% to 95%. For example, on a $100 million asset bank, if we expect losses to the portfolio of $35 million and the loss threshold for bidding purposes is set at $20 million, then the FDIC will absorb losses calculated as 80% x $20 + 95% x ($35 - $20) = $30.25 million (we'll round to $30). Further, because capital is tied up in these troubled loans and resources must be focused on working out those loans, the bidder needs to charge their bid for those carrying costs. Those might be $11 million. As such, the bidder will take the banks pre-existing reserves as its "equity" (call it $10 million for this example), subtract $35 million of losses and $11 million of carrying costs, then add back the FDIC's loss share of $30 million for a total bid of negative $6 million (+$10 of beginning equity - $35 - $11 + $30 = -$6 million). If we were to win the bid, the FDIC would cut us a check for $6 million to take over the bank (and loss-sharing reimbursements would come over time as realized).
All of this is fine, but it still leaves the newly acquired bank with zero equity. A financial buyer (which these days comes in the form of a blind pool) might have to overcapitalize the bank with fresh capital, but a strategic buyer can take its existing capital and apply it to the newly acquired bank. In any case, for BancShares we could assume they'd need to contribute/tie-up $10 million of capital to acquire this hypothetical bank which would make it well capitalized. In essence, they buy this bank at 1.0x book with book being the newly contributed or assigned capital.
This is attractive for a variety of reasons. Importantly, the carry cost asset (the $11 million in our example) is based on a number of assumptions about how long it will take to liquidate bad assets, what the opportunity cost of that capital is, the resources needed to manage those assets, discount rate, etc. The bidder will generally be conservative in this assessment. In the extreme case, if the buyer were to liquidate all of the bad assets for zero, the book value would immediately increase by $6 million (60%) as we'd lose $5 million on the bad assets after the loss-share but we'd accrue the $11 million carry cost asset in its entirety.
This extreme scenario will never happen because the FDIC is both your partner in the assets, your source of future attractive deals, and a key regulator. As such, you would never completely screw them. However, you can see that even if the loss-share portfolio performs extremely poorly, it may actually be a positive event for BancShares. The result is that it is highly likely that the acquisition will end up being at less than book value and that the bank will be overcapitalized and capable of making attractive loans in a lending-friendly environment.
Actual Assisted Transactions:
Temecula Valley Bank - On July 17, 2009, FCB acquired Temecula Valley Bank (TVB) from the FDIC in an assisted transaction. TVB operated eleven branches in Southern California (San Diego and Temecula Valley east of SD). Prior to being shut down, TVB had $1.38 billion of assets and $0.97 billion of deposits (this excludes $304 million of brokered deposits the FCB refused). FCB purchased the $1.4 billion of assets at a $135 million discount. Day one, FCB wrote down the carrying value of the loan portion of TVB's asset portfolio from $1.21 billion to $0.86 billion and its REO from $66 million to $58 million. The bulk of the remaining assets were cash or cash-like, readily marked investment securities and a small amount of "other" assets.
No cash was paid by either FCB or the FDIC at closing. In essence, this was a zero bid. Losses are 100% FCB's on the first $193 million, then split 80/20 FDIC/FCB until losses meet $464 million and then are 95% absorbed by the FDIC thereafter (i.e., the loss threshold set by the FDIC was at $464 million). The term of the loss-share on residential assets is ten years, whereas non-resi real estate is five years with respect to loss-sharing and eight years with respect to loss recoveries. FCB recorded a $103 million loss-share receivable at the time of acquisition and in the first two and a half months identified $32 million in net losses to submit to the FDIC.
In summary, when the net assets of TVB were adjusted upward for the $103 million loss-share offset by the marking of the existing assets an liabilities, FCB records a $58 million "gain" which is effectively $58 million of equity that FCB can use to support the $856 million of loan assets that FCB acquired. FCB likely needed to use another $30 million of its capital to support those assets.
Venture Bank - On September 11, 2009, FCB acquired Venture Bank (VB) from the FDIC in an assisted transaction. VB was located in Seattle/Olympia Washington and operated eighteen branches. Prior to being shut down, Venture Bank had $0.85 billion of assets and $0.71 billion of deposits that FCB inherited (as well as $57 million of other liabilities). FCB purchased the $0.85 billion of assets at a $110 million discount. Day one, FCB wrote down the carrying value of VB's loan book from $650 million to $456 million and its REO from $52 million to $43 million. The bulk of the remaining assets were cash and cash-like investments, investment securities that are easily marked, a small amount of "other" assets.
The FDIC paid to FCB $19.4 million of cash at closing (a "negative" bid). The loss-share was tighter on VB as well. All losses are shared 80/20 FDIC/FCB until losses meet $235 million and then are 95% absorbed by the FDIC thereafter (i.e., the loss threshold set by the FDIC was at $464 million). The term of the loss-share on residential assets is ten years, whereas non-resi real estate is five years with respect to loss-sharing and eight years with respect to loss recoveries. FCB recorded a $139 million loss-share receivable at the time of acquisition and in the first nineteen days identified $8 million in net losses to submit to the FDIC.
In summary, when the net assets of VB were adjusted upward for the $139 million loss-share and the $19 million in cash from the FDIC, FCB records a $46 million "gain" which is effectively $46 million of equity that FCB can use to support the $456 million of loan assets that FCB acquired, meaning that very little to no new capital is required for FCB to take on VB. Further, the faster FCB can put losses to the FDIC, the faster it frees up its capital and resources for productive redeployment.
First Regional Bank - On January 29, 2010 FCB acquired First Regional Bank (FRB) from the FDIC in an assisted transaction. BancShares has not released in-depth detail on the FRB acquisition yet. The basics are that FRB was located in Los Angeles, CA and operated thirteen locations. Prior to being shut down, First Regional had $2.17 billion of assets and $1.87 billion of deposits that FCB inherited. $2.0 billion of the assets were acquired with loss-sharing. Regionally, this fits in with FCB's Temecula Valley Bank acquisition, giving them ample opportunity to continue an in-fill branch strategy or to acquire other contiguous bank footprints.
Return Opportunity: I'm sure nobody follows my VIC write-ups particularly closely, but those that re-read them will notice I am not a fan of valuation targets. However, recognizing most VIC members want something tangible, I'll note the following.
Over the past 15 years, FCNCA has generally traded at a price to book of between 1.1x and 1.9x , creating a valuation arbitrage between the capital BancShares is deploying in FDIC assisted transactions and the market valuation multiple it receives on its book.
Given our acquisition price is at or near book value, it is difficult for me to imagine that over an extended period of time, our shareholder returns lag the returns on equity that BancShares generates. If the bank did no further transactions and were simply to generate 10% ROEs for the next five years, our return would be about 60%. Further, if the P/B were to expand from 1.1x to 1.5x, that would add an incremental 36% to the return. Combined, it seems reasonable that our total return over the next five years will be around 100% or a 15% annualized return. I believe this is achievable with low downside risk and substantial upside optionality from things like the attractive lending environment that prevails, improved scale leading to cost efficiency, "winner" banks receiving premium valuations and potential future attractive acquisitions.