Description

Capitalsource (CSE) shares offer compelling risk/reward with easily 100% upside over time and very low probability of loss.  CSE trades at just below 50% of tangible book value because investors fear it may be the next CIT.  However, the structure of CSE and recent actions to address existing liabilities make this outcome very unlikely.  Over time shares will reflect the value imbedded in the company.

CSE was pitched by gigi404 in 2005.  For an overview and history, please see gigi404's writeup.  That pitch was made when CSE was in the process of converting to a REIT.  A lot has changed since then as CSE took over a bank through an FDIC deal, de-REITed, announced the intention to sell its net lease business, and put its commercial lending business into runoff.  The share price today is only $3.50 (vs $22.30 then) so I believe risk/reward is more favorable now.

In its current form, CSE is generally thought of as 3 separate pieces:  a bank, a healthcare net leasing business, and a runoff portfolio of legacy loans.  A more accurate description of CSE is a holding company with a mix of recourse and non-recourse liabilities.  This distinction is critically important, because recourse liabilities define risk (i.e., downside) while the non-recourse liabilities define embedded call options (i.e., upside).  So this suggests a more proper way to view the company is to split it into 5 pieces:  a bank, a healthcare REIT, a runoff portfolio with non-recourse liabilities (embedded call options), runoff portfolio with recourse liabilities, other assets, and other recourse liabilities.

Breaking these pieces down shows the significant value embedded in CSE shares.

1. Capitalsource Bank - $2.50/share of value @ ~1x BV. As of 6/30 the bank had $865m of ending equity and $173m of intangibles. The bank was created in July 2008 when CSE took over the deposit base from Fremont bank in California through an FDIC assisted transaction. CSE placed higher quality loans into the bank. While this bank will have its share of problem loans, it is relatively "clean". I assign it a 1x BV value even though comparable banks trade at much higher multiples (1.5x-2.5x). Capitalsource Bank had a TCE/tangible asset ratio of 12.5%, a tier 1 leverage ratio of 12.5%, risk based capital ratio of 16.8% as of 6/30/09. New loans are being written today at LIBOR + 500 (or more) with LIBOR floors, giving the bank substantial earnings power in the future.
2. Healthcare net lease - ~$2.00/share of value @ a discount to peer valuation. OHI and LTC are the closest public peers today and trade at cap rates of 9.5% and 8% respectively. AVI, which is due to IPO the week of November 2, is expected to price between 7.5%-8% cap rate. Applying a 10% cap rate to the ~$110m annualized NOI for the healthcare business yields $1,100m of value. The healthcare business had $470m of debt, leaving equity of $630m. At 8% yields, the net lease business would be worth closer to $2.75/share. The net lease business operates 181 skilled nursing facilities in the US.
3. Legacy loans with non-recourse obligations - ~$2.00/share of value @ high end of management cumulative loss scenario. The non-recourse obligations include the six securitizations, the combined credit facility, the CSVII credit facility, and the CS Europe credit facility. The securitizations have a combined $1 billion of CSE residual interest and the credit facilities have a combined $370m residual interest. CSE provides the mix of collateral type for each facility and securitization and also provides a consolidated cumulative loss estimate for its entire loan portfolio. I apply the high end of the loss range (30% cumulative loss on CRE, 15% on cash flow loans, 3% on asset based loans) to each securitization/credit facility to get ~$700m of residual value.
4. Legacy loans with recourse obligations - ~$1.50/share of value@ high end of management cumulative loss scenario. This is the syndicated credit facility which was reduced in July 2009 to a $600m commitment. The facility has $1,276m of collateral at current marks (originally value of $1,650m) and $531m of outstanding principal after the CSE capital raise in July 2009. At the high end loss scenario provided by management, the facility would sustain $260m of losses, leaving $480m of value left for CSE. Though management has not provided full credit quality details of the collateral pool, management has indicated that the embedded CRE in the pool is much higher quality than the overall book. As this facility is 2.4x overcollateralized, CSE would need to sustain 60% cumulative losses in the asset pool to default on the structure.
5. Other assets - ~$0.35/share of value. The company has $93m of equity investments and $95m of REO. At a 40% discount, these would be worth $110m.
6. Other liabilities - ~$4/share of liability. The company has $580m of convertible debt ($330m puttable in July 2011 and $250m puttable in July 2012), $300m of senior secured notes (July 2014), and $439m of trust preferred debt (mostly matures after 2035).

Combined, these pieces give $4.35/share of value if all losses were realized today.  This is 20% above where the shares trade today.

However, to this total we would add pre-provision earnings over the next two to three years (assuming the legacy book losses occur over a two to three year period instead).  Pre-provision earnings were $60m in Q2 2009 and were $70m in Q1 2009.  The $300m notes raised in July 2009 will add an incremental $10m of interest expense per quarter.  Offsetting this partly is the ~600bp interest yield improvement between new loans written by Capitalsource Bank and low yielding $888m "A" Participation loans that will mature within the next year.  If we assume that the company generates $55m of pre-provision income per quarter as the legacy portfolio winds down - then the CSE will earn another $1.50-$2.00/share.  This puts the value of CSE at $4.85 - $6.35/share or 40-80% above the current share price.

The above assumptions are very conservative.  They assume high end losses in the portfolio, losses are equally distributed between the different loan pools, and very low multiples for CSE businesses.  Using 2x BV for the bank (which is in line for a bank with similar profiles to Capitalsource bank) adds another $2.50/share of value.  Using 8% cap rates for healthcare net lease adds $0.75/share.  Reducing loss estimates to management's expected case of $400m of additional provisions yields $0.60/share of upside.  And if, as management has indicated, some of the worst quality legacy loans are in the securitizations and non-recourse credit facilities, this could add $0.50 to $1.00/share of upside (maximum loss for CSE in these structures is CSE's equity, additional losses beyond the CSE equity accrue to debt holders of these structures, not to CSE).  These assumptions add another 100%+ upside to the current share price. 

I am not suggesting that CSE benefits from all of these factors or benefits fully from any specific one, but I think the above analysis demonstrates 100%+ upside from the current share price is easily achievable.  And at 100% upside, CSE would still trade at a discount to tangible book value today.

A quick alternative exercise further demonstrates the potential value in CSE.  At the end of Q2 2009, CSE had a tangible book value of $2.4b or $7.50/share adjusted for the subsequent capital raise.  Adjusted for a deferred tax valuation allowance of $137m in Q2, the company had a net capital loss of $110m and had $104m of loss in Q1.  At this pace, it would take 12 more quarters of provision at the same levels as seen in the last six month in order for CSE to trade at 1x tangible book value.   That would imply an additional $2.1b of provision or almost 3.5x management's high case estimate of $620m provisions.  Total portfolio losses would reach nearly $3b or 34% cumulative loss versus management estimate of 17% cumulative loss.  Even if you believed losses at this level were possible, CSE would still then be trading at 1x BV, would have a clean loan portfolio, would be writing new loans at very high rates (L+500 with LIBOR floors), and would have reduced competition - factors that would warrant a premium to book value.

So why does CSE trade so cheap?

Investors dislike CSE for several reasons:

• the next CIT?
• a toxic legacy portfolio that has significant exposure to top of the market commercial real estate loans and mid-market mezzanine loans
• liquidity risk from undrawn commitments to customers
• liquidity risk form maturing credit facilities and converts
• limited trust in management

Let me address these issues.

The next CIT?
No.  CSE would have been the next CIT (or possibly CIT would have been the next CSE) if CSE had not taken over the deposit base of Fremont in 2008.  However, this completely changed the funding sources for the company as it significantly reduced the reliance on wholesale funding.  CSE then successfully amended and extended maturing credit facilities this spring and summer to further address its funding issues.  Also, while CSE and CIT compete for mid-market lending, CSE is not exposed to factoring receivables, which is what caused the run on the bank at CIT.

CSE's toxic portfolio
There are two points on the toxic portfolio - that it limits the upside in the shares and that it creates downside risk.  To the first point, the analysis above suggests that despite this toxic portfolio, CSE shares are incredibly cheap.  The second point is that the toxic legacy portfolio creates risk to CSE shares only to the extent that these assets are attached to recourse liabilities.  As of today, the only recourse liability attached to specific assets is the syndicated loan facility.  As described above, this facility is 2.4x overcollateralized and would need 60% cumulative losses to default.  Under virtually any scenario, it is difficult to see how this is even possible.  Using CSE's high case loss scenarios only generates 20% cumulative loss for this portfolio, but this pool of loans is likely better than the average loan at CSE.  Approximately 70% of the loans in the pool are senior notes.  The largest single industry exposure is nursing care facilities at 23% of the collateral pool.  Approximately 17% of the pool is asset based loans, for which CSE has seen and expects minimal losses. 

Liquidity risk from undrawn commitments
Conceptually, I would understand this argument if someone made it in October 2008 or March 2009.  However, given that CSE has now withstood the financial meltdown of the fall of 2008 and the winter of 2009 without a run on the bank, it is hard to see what would trigger a run now.  Specific information provided by the company supports this notion.  First, the company has total unfunded commitments of $3.0b in June 2009, a 20% reduction from $3.6b in December 2008.  Of this total, $782m is commitments for Capitalsource Bank, which is very well capitalized and is trying to grow its loan base.  This leaves $2.2b of unfunded commitments at the holding company that can actually create liquidity risk.  Of this total, the company expects to fund $200-300m based on a review of all outstanding commitments in the next 12 months.  The reason only 10-15% of the commitments are likely to be drawn is because there are significant limits to funding.  Collateral posting is required for asset-based revolvers, real estate loans have project milestone requirements to draw additional funding, drawing on commitments is not permissible if a borrower is in default, and CSE has some discretion in providing commitments.  Also, to the extent that CSE has fund draws higher than expected, it has the ability to write loans from Capitalsource Bank, thereby reducing the liquidity drain at the parent company. 

There is good reason to believe that the $200-300m estimate from management is very conservative.  In Q2, the company expected 50 loans would draw on commitments and forecasted $122m of draws, but only $25m of net draws came form those loans.  In Q1 and Q2, the company had net negative draws on its parent company revolver commitments.  This is completely consistent with commentary from other banks and federal reserve data that show lower demand for loans and deleveraging.  As revenue decreases (and with it working capital and investment needs), companies have lower need for borrowing.  Revolvers represented $1.9b of the $2.2b of unfunded commitment at CSE parent.  Thus if revolvers actually have been generating liquidity for CSE instead of drawing down liquidity, it is hard to see a liquidity crisis triggered by unfunded commitments.

Liquidity risk from debt maturities
CSE has taken a lot of action in 2009 to resolve any maturity risk.  The credit facilities, and most importantly, the syndicated facility were amended and extended in H1 2009.  New equity and debt were raised in July.  The remaining maturity risk for CSE lies primarily in the convertible bonds that can be put to the company in July 2011 and July 2012 ($330m and $250m respectively).  Very large holders of the converts include Baupost Group and Farallon.  Both own common shares and Farallon has a representative on the CSE board of directors.  The company has exchanged shares and new converts with existing convert holders several times in the last year.  If the shares continue to trade poorly and the company does not generate sufficient capital, then another exchange is the likely outcome in the next 12-18 months. 

The syndicated facility also can create maturity risk.  The syndicated credit facility matures in March 2010, but will pay down $20m monthly beginning in April 2010.  CSE has indicated that it will use 75% of the proceeds of the HUD loan it will receive for the healthcare net lease business to pay $90m of the outstanding balance.  The sale or IPO of the healthcare net lease business will generate further proceeds.  At a 10% cap rate, this business will generate an additional $500m after the HUD loan.  These combined sources should eliminate the maturity risk from the syndicated facility and reduce potential dilution from a convert loan exchange.

Limited trust in management
CSE has gone from a lender to a REIT and now to a bank in the last few years.  This has clearly frustrated investors, and for good reason.  However, given the state of capital markets, the condition of the legacy portfolio, and CSE's capital structure, there is little choice for the company today other than to pursue its current strategy - run off the legacy portfolio, sell the healthcare business, and grow Capitalsource Bank.  Following this strategy should yield tremendous value as indicated above.

Management has also been criticized for selling shares opportunistically, specifically John Delaney.  Delaney has said that he was looking to diversify his own wealth.  While this may be true, he very likely saw the disaster coming in credit and chose to take some profits.  Also, Delaney's compensation structure probably influenced his selling.  Delaney is paid in stock only as his base salary and bonus are both paid in stock.  He made the decision to get paid in stock only beginning in mid 2006 when his shares traded above $20/share.  Today CSE trades at $3.50/share.  Delaney has 7m options at $23.72 that will probably expire worthless.  While I don't feel sorry for the guy, I don't blame him for selling some share.  Today, he still owns over 6m shares of CSE stock directly and indirectly and I am sure would love to see CSE trade above BV.

Finally, management lost investor's confidence after the July 2009 equity raise.  They gave some guidance on chargeoffs and provisions for Q2 results which ended up being too low when the company reported results a few weeks later.  While this is indeed a problem, quarter to quarter results will be volatile as the legacy portfolio unwinds.  Management did not change its longer-term loss projection as a result of Q2 results.

Catalysts

So when will CSE trade higher?  There are a few clear catalysts that play out over the next 12 months that should materially increase share price.

• Improvement in credit trends. Chargeoffs and provisions for the legacy portfolio will eventually peak and begin to decline. Even at management's high case loss scenario, this should happen within the next 12 months.
• Monetization of the healthcare net lease business. The company has said it will consider an IPO (watch AVI the week of November 2) or a sale to a strategic buyer. A sale would remove liquidity risk and eliminate the remaining balance on the syndicated facility.
• Capitalsource Bank earnings grow materially. The bank will get ~600bp improvement in interest yield as the $888m "A" participation loans are repaid within the next 12 months. The earnings power in the bank will dramatically improve as a result of the swap to higher yield and growth in new loans.
• Conversion to a bank holding company. CSE has put on hold its bid to convert to a BHC as it repaired its balance sheet and worked on the disposition of the healthcare net lease business. BHC status will improve CSE's funding sources and provide a better and lower cost deposit base.

Catalyst

Improvement in credit trends

Monetization of the healthcare net lease business

Bank earnings begin to grow

Conversion to a bank holding company