Description

Thesis:

Roma Financial Corp (ROMA) is a misunderstood and underfollowed thrift conversion that combines a number of features that we look for in an investment including: an accounting quirk that distorts economic reality, an attractive valuation, insider ownership and buying, and the chance for significant upside potential.

Business Description:

ROMA is a central New Jersey-based thrift that was founded in 1920. It operates through its wholly-owned subsidiary, Roma Bank, a federally chartered stock savings bank, and its 89.55% owned RomAsia Bank, which is also a federally-chartered stock savings bank. Together (Roma Bank makes up the vast majority) they operate 15 branches (6 of which were opened in 2008) in three counties in NJ. The company has historically been run as a mutual resulting in a very conservative approach to underwriting because there was no upside from the profit motive. It converted to a mutual holding company (MHC) structure and went public in 2006.

What is an MHC?

In 1987, Congress created the mutual holding company structure. In a MHC conversion, there are two steps to full conversion. First, the company sells a minority interest in the bank to the public (first offered to depositors), creating two classes of shareholders: the public (us) and the depositors whose interest is held in trust (and controlled by the board). What makes MHCs interesting is that the majority shareholders' (depositors') shares effectively don't exist because (1) they waive their rights to dividends and (2) when they eventually sell their shares in the second step conversion, the company receives all the proceeds. Therefore, when calculating the value per share, one should only use minority shares outstanding (publicly traded).  This is the quirk mentioned above that distorts economic reality. In the case of ROMA, it changes the valuation from ~2.0x tangible book value to ~0.5x tangible book value because nearly 75% of the outstanding shares are held by the MHC.

The mutual holding company structure causes confusion, and management does little to clarify it by reporting adjusted numbers, holding conference calls, attending conferences, or encouraging analyst coverage. 

TFSL was previously written up and has the same structure. There is lots of discussion of thrift conversions and MHCs in the comments to that write-up. ROMA is basically smaller ($100m mkt cap v. $1.1B) and safer (asset type and geography).

Valuation:

Okay, so the MHC structure is a little confusing making it is worth investigating, but what is it worth?

Price / Tangible Book Per Minority Share as currently configured

I think this is the most appropriate valuation metric because it is a good estimate of liquidation value. The company is basically a portfolio of conservative loans and safe, liquid securities financed with cheap deposits.  Adjusted for the MHC shares, the price / tangible book value is approximately 50%. If the company chose, it could sell the securities, let loans pay off, and take fewer deposits (by offering lower deposit rates) thereby shrinking the balance sheet and paying out the proceeds as dividends. More on why I think the shareholders equity value is very secure below.

For perspective, the median NJ-based thrift trades for 110% of tangible book value (according to SNL). Also, comparable transactions are still being done for significant premiums to book value. Most recently, Danvers Bancorp (DNBK) offered to purchase Beverly National Corp (BNV) for ~144% of tangible book value. Mutuals of comparable size that have already done second step conversions trade at an average of 120% of tangible book (HCBK, PBCT, FNFG, BKMU, BRKL, PBNY, UBNK, ABBC, WFD, and HOME). Recent standard conversions of comparable size trade for an average of 95% of tangible book value (DNBK, ESSA, CBNJ, and LEGC). Between the logic of liquidation value and comparable valuations, we feel comfortable that 100% of tangible value is a realistic and conservative estimate of value.

Potential Reasons that Tangible Book Value is too conservative

• (1) Deposit Franchise / Regulatory Approval / Local Goodwill

Deposits are a sticky and cheap source of funding. Applying a 15% premium to ROMA's DDAs and a 5% premium to its non-brokered CDs results in an increase in the equity value of 24%.  Also, regulatory approval is not easy because regulators want bankers to have experience and a track record of managing a bank, creating a barrier to entry (just ask the PE guys). The combined effects of a deposit base and regulatory approval add at least 25% to the tangible book value.

• (2) Understated Book Value

The company owns several branches that have been on the books for a very long time resulting in very low basis land and fully or mostly depreciated book value. Assuming the 5 branches that are on the books for under $1m would cost $2.5m to replicate (as their more recent openings indicate), tangible book value would increase by ~$.90/share, another 4%.

• (3) Fresh capital at an opportune time

The lack of capital at many banks should make for more attractive capital deployment opportunities than in more typical times. In this way, ROMA is very well positioned because (1) without increasing leverage, they can pick up more spread by selling securities and originating loans and (2) they could nearly double their asset base, picking up the spread between incremental assets and liabilities, and still be very well-capitalized (though we would prefer buybacks).

• (4) Buybacks

The company has executed three buyback programs since coming public, buying 20% of minority shares outstanding (but only achieving ~10% reduction because of ESOP, restricted shares, and stock options). I expect them to continue buying shares which adds value for shareholders by being accretive to tangible book value and which increases the upside in a second-step conversion by further concentrating the tangible book value accretion (if issued above 1.0x) in the hands of fewer minority shareholders. 

Overcapitalization

Another way to look at the company is as two entities: a "well-capitalized" thrift (according to regulatory guidelines) and a safe, liquid securities portfolio. The ratio of tangible common equity to tangible assets was 19.6% as of year-end compared to the thrift industry average of 9%. So, the company could pay out half of the tangible equity (9.8% of tangible assets) to shareholders and still exceed the industry average-and the stock is trading at half of tangible common equity per minority share! Also, for ROMA the two most restrictive capital requirements are Total Capital to risk-weighted assets and Core (Tier 1) capital to adjusted tangible assets, which it has cushions of $130m and $134.5m respectively. This compares to an effective market capitalization of $108m. So, for the price of $12.50 / share, you are receiving a well-capitalized bank with ~$10 of tangible equity/share and $15.66/share of government and agency securities.

Earnings View

Applying an earnings valuation, even adjusting for the share count, does not yield a low PE (25x 2008, 16x 2007, 19x 2006). The company is clearly under-earnings on its capital with ROEs in the low single digits, but it is under-earning for explainable and correctable reasons. First, the company is experiencing a drag from excess capital. Should the company choose to ramp up loan origination and fully deploy its excess capital, its earnings would improve dramatically. The table below assumes that the company was to go to a ratio of 10% tangible common equity (from 19.6%), realizing the same net interest spread (2.67%), the same efficiency ratio (77%), and a 35% tax rate, resulting in an 88.6% improvement in ROE.

Second, ROMA is "under-loaned" at loans equivalent to 49% of assets. Thrifts are more typically at ~80+%. This depresses earnings because securities have lower yields than loans. Assuming 80% loans to assets, the same efficiency ratio of 77%, and a 35% tax rate, ROE improves ~7.6% (changes in the table are cumulative to make the improvement in ROE comparable).

Third, ROMA has a high efficiency ratio (ratio of operating expenses to net interest income). The efficiency ratio has deteriorated to the level of 77% (from a 2003-2005 average of 53%). These extra costs are primarily from the six branches opened in 2008 (resulting in 15). These new branches will take some time to ramp up, but should not be a permanent impediment to the historical efficiency ratio (which management believes is achievable). If the company were to have an efficiency ratio of 55% (which is slightly higher than the 2003-2005 average when they weren't opening branches rapidly or experiencing IPO costs, ROE would further increase by 97.9% to a more reasonable 8.8%. I don't expect this level of profitability soon, but it's achievable and understandable why they are under-earning (unable to load associated table). 

What does the valuation look like while we wait for the value discrepancy to close (naturally or via second-step)? First, minority shareholders are receiving an approximately 2.5% dividend yield. This represents ~50% payout ratio. Second, the rest of the earnings will accumulate on the balance sheet (increasing book value by ~2.5%/minority share as well. Third, management has and continues to buy back shares at a moderate pace of ~5% of minority shares per year. At 50% of tangible book, this adds value of ~2.5% per minority share per year. Assuming the valuation gap doesn't close, that results in a ~7.5% return. However, we obviously expect that the gap should close over time. Using a 5 year time horizon results in a total of ~22.5% (15% from accretion of tangible book towards tangible book value + 7.5% - though if the price is increasing the buybacks will add less value).

Balance Sheet

Since we are using tangible book value to value the company, it's important to understand what comprises book value and how secure it is.

Assets

As of year-end 2008, it had $1,077m in assets comprised of:

• 48% loans concentrated in residential mortgages, commercial loans, and home equity loans
  • 44% residential (1-4 family); no subprime or IOs; conforming standards
  • 30% Multi-family and Commercial; Maximum LTVs of 75% with personal guarantees
  • 26% Home-equity; combined LTV maximum of 75%; 40% debt-to-income maximum
  • Purchased loans of $8m (stopped in 2008)
  • Loan growth was moderate in recent years at 13%, 11%, 10%, and 12% for 2005-2008
  • NPLs: 1.98% of total loans; ALLR a bit light at 21.5% of NPLs (Co says it's because they are so conservatively underwritten)
• 44% Cash and Securities
  • 64% Agency MBS; bank policy not to buy MBS with more risk than underlying (ie: higher tranches)
  • 19% government securities (90% federal / 10% Muni)
  • 17% cash
• 8% Other assets -PP&E is the largest bucket here

Liabilities

At year-end, ROMA's liability mix was as follows:

• 88.3% deposits
  • Loans / Deposits: 68%
  • Deposit growth:
    • 2008: 17% (thrift industry -18% à may be skewed by failures)
    • 5 yr CAGR: 7%
  • No brokered CDs
  • Deposit composition:
    • 56.6% CDs (27.6% of which are over $100k)
    • 26.7% Savings
    • 16.7% Interest bearing demand deposits
• 10.0% borrowings
• 1.7% other

It is often said that a bank's assets are its liabilities and its liabilities are its assets because assets create problems in the form of loan losses and the real value of a bank is its cheap funding from deposits. This liability base should be attractive to an acquirer because of the material core deposits and problem assets are unlikely.

Second-Step Conversion Upside

The upside scenario is a well-priced second-step conversion. A second-step conversion is when the company converts from a MHC structure to a traditional public company structure by selling the shares held by the MHC (slightly simplified). Those proceeds will go to the company, essentially being split between the new buyers and current minority holders (not the trust). Depending on pro forma valuation, this can be a catalyst to unlock value for the minority shareholders. This is typically the case because deals are done at a premium to tangible book value, so it is materially accretive to book value, cleans up the structure, and obviously clarifies the timeline. Alternatively, the company could be re-mutualized either by itself, as a buyout from a private mutual, or as a buyout by another MHC (ISBC would be a good fit, but has some balance sheet issues). However, until three years after it has undergone a second-step conversion, the OTS prohibits a sale to a non-mutual (a non-mutual can buy it before a second step conversion though).

It is difficult to say when a thrift might undertake a second step conversion. Also uncertain is the valuation at which a second-step could be done at. However, looking at the history of MHCs that did a second step conversion, it appears that most of them were done 4-5 years after the IPO which makes sense because the OTS does not allow them to do a second step conversion for 3 years after the IPO (ROMA went public in July 2006). We also think ROMA is more likely to do a conversion because of the actions of the company and insiders (discussed more below). With regard to pricing, previous MHC conversions of comparable size (defined as assets between $700m and $3B) have been done at ~110% of pro forma tangible book value (same list as above: HCBK, PBCT, FNFG, BKMU, BRKL, PBNY, UBNK, ABBC, WFD, and HOME). This includes a couple done at a discount to tangible book value (UBNK and HOME) that appeared to need the capital at the time, making them less relevant. Excluding those two, the average increases to 114%.

The table below demonstrates the upside potential. MHCs that have done second step conversions below 100% of pro forma tangible book value were generally in a position of weakness and needed the capital. This valuation (110% of tangible book value) generally makes sense because, as discussed above, a premium to book value is warranted because of the cheap deposit base. There we computed a 24% premium was justified, and a 10% premium leaves something on the table for buyers of the second-step offering. Also, the downside for a second step buyer is minimal because the excess capital (which would be even greater than now) could be used to purchase shares if the price fell below tangible book.

The table below does not capture dilution from options, additional ESOP contributions, future grants, or the offsetting benefits of stock buybacks, but is representative of what shareholders could expect in a second-step conversion.

A possible reason to undergo a second-step conversion is because management thinks it can sell itself for a subsequent premium. Of the thirty-nine second step transactions announced since 2000, eleven are not yet eligible to be purchased because they did second step conversions too recently. Of the other twenty-eight eligible, nine have subsequently been acquired for an average (and median) price of 2.2x tangible book value. This high percentage being sold makes sense because insiders position themselves throughout the conversion process to benefit from the structure, resulting in significant equity interests that they will want to maximize before they retire.

We feel confident that most MHCs will eventually undertake a second step conversion. If they wanted to preserve absolute safety of deposits and avoid my calls, the board would have remained a private mutually-owned thrift. They choose to go this route because they see an opportunity to do well for themselves economically. We are riding their coattails and simply relying on them to act in their own best interests.

Insider Actions:

Insiders have a 10.7% (% of minority shares) stake in the company through direct holdings and the unallocated shares in the ESOP plan. Management and directors have also received meaningful option grants (~820k shares struck at $13.67 - timing seemed favorable, similar to your typical spinoff) and restricted stock grants (222,000). Assuming a valuation of approximately book value makes the options equivalent to ~410k shares (net of proceeds to the company at ~50% of tangible book), increasing the insiders stake to effectively 15% (rights to buy 10% at $13.67 but TBK/share of $25.56, so ~50% of the proceeds go to pay for the exercise price). The buyback mentioned above becomes more important when they are material owners, signaling that they think it is a good use of capital. Finally, I would note that the chairman and CEO are not young at 90 and 70 years old, respectively. This gives them an incentive to take action while they can realize the benefits before potentially losing control (and so that they have time to enjoy it!). This makes me think they may be interested in value realizing events earlier than if they were younger and wanted to build a bigger stake or bigger company.

It is important to note that management has effective control of the company via the MHC. This provides management and the board the opportunity to enrich themselves at the expense of the minority shareholders. Examples include generous stock options, restricted stock, and ESOPs. These are common in MHC conversions and present in this case, but in tolerable amounts. While we may not like the way that management is splitting the pie, we think the pie is so big that there will be lots left over for minority shareholders and that management is well aligned to increase the size of the pie as well. Further, the signal that they are sending by trying to gain more equity exposure is a positive.

Why is it cheap?

• Misleading share count
• $100m market cap and no analyst coverage
• Lack of hard catalysts

Risks:

• Loan deterioration
  • There is a significant buffer before losses impair the opportunity (19% TCE/TA and 50% P/TBK à9% realized losses before our investment would be impaired v. NPL/Assets under 1%), but we still must take this risk seriously.
• Agency Securities/RMBS exposure
  • If MBS credit deteriorates further and FRE/FNM lose the government backstop, the safe, liquid portfolio is probably not safe or liquid.
• Interest Rate Risk
  • Borrowing short and lending long exposes us to inflation risk
  • ROMA does not retain 30 year mortgages and actively manages interest rate exposure
• Geographic Concentration
  • Loans and deposits are derived from Central New Jersey. A significant downturn in this area could adversely affect the company.
• Regulatory/MHC
  • Changes to the MHC structure and deals that shift the benefits from minority shareholders to MHC owners.
  • Prevention of buybacks, dividend waivers, or second-step conversion (the DC types all seem to think these would be grandrathered in)
  • Interest rate caps / deposit floors
  • Forced lending
• Competitive
  • Competition for deposits, especially if other funding sources are gone (securitization).
  • Banks that can't get capital elsewhere might mis-price deposits for liquidity.
  • Higher FDIC limits benefit weaker banks. Limits have just been further extended.
• Management
  • Cost structure - management may not be motivated to improve profitability
  • Agency Costs - management may enrich itself to the extent that minority shareholders don't realize the opportunity

Catalyst

• Accretive (to book) buybacks
• Insider purchases
• Increasing profitability from maturation of branches
• Second-step conversion - There is some circularity in the timing here. The further the price is away from tangible book value per share, the less likely a second step transaction is in the near term. I believe this is part of why it is cheap. I'm willing to wait.