Four years ago, iStar Financial's (SFI) primary business was providing shorter-term senior debt for real estate development projects. With the onset of the financial crisis, the collapse of collateral values and a liquidity vacuum, the majority of their loans were unable to obtain replacement financing as they came due. As a result, these loans became non-performing and SFI has spent the past four years taking back, sometimes through lengthy legal battles, the collateral underlying the loans. At the same time, the company has been reducing the size of its balance sheet by diligently paying down its own debt maturities through refinancing and asset sales.
That process is nearing its final days, with the result that SFI is now a completely different animal than it was four years ago. Loans will soon represent less than a quarter of its assets, with the balance comprised of a mixed portfolio of real estate -- including a huge land portfolio, multiple high profile condo projects, its legacy net lease assets, and some key other investments.
The graphic below shows how the company's net asset profile has transitioned since 2008.
Click to enlarge images.
Here is a brief description of the assets in each of the company's five business segments:
As of June 30, 2012, the gross carrying value of the loan portfolio was $2.85 billion, representing approximately 100 loans. Of this, $1.73 billion are performing loans and $1.12 billion are non-performing loans (NPLs). There are specific reserves against the non-performing loans of $507 million (45%) and a general reserve of $57 million, resulting in a net carrying value of approximately $2.3 billion. The collateral is spread out geographically as well as among real estate type, including apartment/residential, land, retail, hotel, mixed use, industrial, entertainment, and office.
2. Net Lease Assets
These are facilities that have been purchased by SFI and leased back to corporations on a long-term net lease basis. Property types include industrial/R&D, entertainment, office, retail and hotels. As of June 30, 2012, net lease assets were 91% leased with a remaining average lease term of 12.2 years with a weighted average effective yield of 8.4%. An example of one of the larger transactions was a sale/leaseback to AMF Bowling in 2004 of 185 bowling centers for $250 million on a long-term lease.
3. Other Investments
These include equity investments in other real estate ventures or funds, and a 24% interest in LNR Property, which is the nation's largest special loan servicer.
4. Other Real Estate Owned ((OREO))
These are properties that have been foreclosed on and are held for sale; approximately 65% residential/for sale condo projects, with land, retail, hotel, and mixed use making up the balance. Of the 18 condo projects (approximately 1,500 units) some are high profile projects, such as 10 Rittenhouse in Philadelphia, The Residences (and Hotel) at Mandarin Oriental in Atlanta, and Trump Plaza in New Jersey. (Check out the others here.)
5. Real Estate Held for Investment (REHI)
These are also projects that have been foreclosed on and which SFI plans to hold for an extended period of time, about 66% being land assets, mostly for residential development. Some of the projects are currently selling lots (Magnolia Green, Richmond, Va. ), some are in various stages of rezoning or permitting (Ponte Vista, San Pedro, Calif., Asbury Park, N.J.) and some are simply being land banked for when the housing market is more robust (Atrisco, N.M., Surprise, Ariz., Los Valles, Calif., Indiantown, Fla.). As of June 30, 2012, the land is held on its books at $815 million, making it one of the largest owners of residential development land in the nation. As the Lending, OREO and Net Lease divisions reduce their assets, the weight of the land component as a percent of the total portfolio is expected to increase.
In what follows, I will first touch on liquidity, and then assess the value of each asset component to arrive at a "Level 2 or 3 type" valuation of the company. Bear in mind that the company has many assets, which are not easy to ascertain from the SEC filings, and require a fairly intensive time commitment to uncover and analyze. I will highlight the major ones and hopefully give you enough to make a bigger picture judgment with a sufficient margin of safety.
With non-performing assets not generating any cash flow, together with downgraded credit ratings making access to capital more difficult, the key hurdle for SFI through the transition has been to ensure that there has been sufficient liquidity to meet its operating losses and debt maturity obligations.
This was mostly accomplished through two large secured credit agreements in 2011 ($2.95 billion) and 2012 ($880 million), the sale of a $1.6 billion portfolio of net lease assets in 2010, purchases of their unsecured debt at a discount, some smaller asset sales and refinancing, and, recently, a new issue of unsecured debt ($275 million). The secured facilities are self-amortizing via sales or repayment of the collateral supporting them. The newer debt has come with a higher coupon than the debt it replaced and so the cost of debt is negatively impacting its margins. It will continue to trend higher with the replacement of an October 2012 convertible issue with a .5% coupon, with debt in the 8% to 9% range.
The following charts show the trend in its weighted average cost of debt from 2008 through 2013, and its debt maturity schedule from 2012 onward.
There are some significant maturities in 2013 and 2014, which will be addressed with a combination of new secured or unsecured debt to replace the existing secured debt, proceeds from loan repayments, proceeds from OREO sales (mainly condo sales), and to a lesser extent net lease asset and loan sales, a possible sale of LNR, and cash on hand. The most recent rating by Fitch for SFI's latest $275 million unsecured Senior Note issuance provides an excellent assessment of SFI's credit risks.
The company has always responsibly addressed its debt maturities, never missed a dividend payment on preferred stock, and announced two separate stock buyback initiatives in the past 15 months, which would not be prudent if liquidity was a concern. Absent a significant global macro event, I think it is reasonable to infer that liquidity is no longer a concern.
Absent liquidity concerns, the key issue becomes trying to establish the real market value of the assets in each division, and estimating a timeline for realizing the value. Given the variety of assets, some of the values will be proven out in the near term, such as loan repayments and condo sales, while others, such as master planned communities, will play out over 10-20 years before the full value is realized.
SFI currently does not generate positive GAAP earnings, partly because the remaining NPLs do not pay interest, due to both the large land holdings it has have accumulated and to higher interest rates on new debt capital -- which has replaced maturing debt obligations. After adjusting for depreciation and other non-cash items, adjusted net income is at breakeven or slightly negative, but not materially so.
Segment Outlook and Valuation
Current net interest margin is very small -- for the three months ended June 30, 2012, the weighted average cost of debt (6.51%) is only slightly below the weighted average effective yield on its performing loans (7.6%). And this margin will narrow after the October 2012 convertible issue is repaid. After allocating G&A, it's close to breakeven. I do not see much growth in this segment unless demand picks up, and then it would still require a lower cost of funds to be profitable. I expect to see this segment continue to get smaller as loans are repaid. With the expertise it has in problem loan resolution, perhaps we will see SFI opportunistically buying distressed loan portfolios, but that's just a guess. The key issue here is whether they will recover the full book value of its remaining loans outstanding.
The performing loans have a weighted average loan-to-value ratio of 76%. If the non-performing loans were made an original loan-to-value of 75%, this would imply an original value for the $1.1 billion of NPLs of $1.5 billion. The net carrying value of these loans is $605 million, or 40% of original value, so that seems like a reasonable margin of safety to expect a full recovery of the current carrying value. In addition, there is a general reserve of $57 million, or 2% of the total loan portfolio. In light of management's conservative accounting and the improvement in the commercial real estate sector, I would expect a full recovery and possibly a recapture of some of the provisions.
Net Lease Business
This segment currently has an annual NOI of $134.5 million based on Q2 2012 NOI annualized. Capitalizing this at 8.5% -9.0% gives a value range of $1.5 billion to $1.6 billion. There are some weak assets in this portfolio with large vacancies, particularly in the office component. Leasing these up would generate an additional $30 million per year for an additional value of $350 million. For example, 500 Woodward Ave. in downtown Detroit, a one million square feet Class A office building, has approximately 500,000 square feet vacant. There has been some leasing activity this year, and with the revitalization momentum in downtown Detroit, this space should get absorbed over the next few years.
The most significant asset here is LNR Property. It is public knowledge that it is shopping this asset, with reports estimating the price to be at least $1 billion. With a 24% stake, SFI would receive at least $240 million from a sale or $65 million above the current book value. Reported bidders include Lennar (LEN), Starwood Capital Group, and Island Capital.
Other Real Estate Owned
This segment consists mostly of 18 for sale condo projects (1,500 units). Condo sales activity has been brisk, consistent with new home sales in general. A number of larger trophy projects have only recently been readied for sale, after some additional investment, upgrades and repositioning from SFI to maximize its values. Sales in Q2 were significantly higher than Q1. Based on year-to-date sales of $149.7 million and income of $20 million, the gross markup on these assets is running at 15% above book value. However, this is being eroded somewhat by the holding costs of the unsold inventory. Based on the methodology used to account for condo sales, this margin should increase. From Fitch rating review:
Reported EBITDA understates the company's cash generation power, given that the accounting for non-performing loans and real estate owned allows the company to recognize income only when certain thresholds are met. Fitch expects that reported earnings will improve going forward as certain of these thresholds are achieved related to unit sales at iStar's owned condominium properties.
CEO Jay Sugarman has indicated that the company has "healthy unlevered IRRs" on all 18 projects. Taking these factors into consideration, I project that it will recover somewhere from 110%-120% of current book value after holding costs.
Real Estate Held for Investment
Land value is the biggest unknown in the portfolio, and difficult to value. The best we can do is reflect on what management has stated, infer by comparing to other similar assets, and reviewing any actual recent transactions. Here are some facts:
- During the Q2 2011 conference call a little over 12 months ago, Sugarman indicated that at that time the company was carrying the land on its books at about 50% of original capital stack, or original purchase price (equity plus debt).
- Demand for land from homebuilders has become quite brisk. In the second quarter of 2012, Lennar spent $287 million on new land purchases, Toll Brothers (TOL) spent $189 million, and KB Homes (KBH) $83 million. The S&P Homebuilders ETF (XHB) is up 60% from a year ago.
- Other publicly traded entities with significant land exposure such as Howard Hughes (HHC) and St. Joe (JOE) are up 25% and 10%, respectively, over the past year, measured in terms of enterprise value.
- It was reported last month that SFI sold a commercial parcel in Hollywood, Calif., to Kilroy Realty for $71 million, which is recorded on its books at $65 million, so almost a 10% premium to book value on a fairly large size asset.
- The company has added an in-house land team of 18 people to maximize the value of the land platform. It is investing additional capital and working with local cities to rezone or re-permit as necessary to maximize value. A good example is Asbury Park, N.J., where it has navigated the interests of all the constituents to ensure a successful development.
Given management's historical prudence in taking reserves as necessary, I assume that as of Q2 2011 the book values of the land were a reasonable reflection of market value at that time, and use that as a starting point. Based on some pretty strong evidence of increased values for land related entities over the past 12 months, the sale referenced above and SFI's success in repositioning assets, I would argue that it is reasonable to estimate the current value at 110%-120% of book value.
The land has become quite capital-intensive for SFI. It is investing additional capital into the land portfolio, and more importantly, carrying the land at an 8% cost of debt will weigh down on the eventual returns. Sugarman indicated on the conference call that it would look for ways to "supplement the capital resources for that portfolio." My guess is that this would be in the form of a joint venture, or replacing some of the debt with equity capital in the future.
Estimating the Value of all the Pieces
To ascertain the combined value, we take the current balance sheet, and first allocate the NPLs equally to OREO and REHI, which is where they will eventually end up (the equal allocation is an estimate but not material to the overall picture). I then estimate a low and high value for each of the assets, based on the outlook provided above, and with further footnotes in the table below.
I come up with an estimated value as of today, for the common stock, somewhere between $18 and $23 per share, which is significantly above the both the current book value of $12.29 and the stock price of $7.18. If the company completes the remainder of the share buyback, these estimates will increase slightly.
Some investors may like the preferred stock, which trades at $20.50 and offers a 9.5% yield. It has never missed a dividend payment.
I am interested in the common, which trades at a 40% discount to book value and a 65% discount to the average of the high and low values I calculated in the previous table. While a 65% discount may seem like a stretch, this company is highly leveraged and so the equity is very sensitive to changes in asset values.
In July 2012, the company announced a $20 million stock buyback program. There is still $16 million available under the program for repurchases. There are also 16.8 million shares short, representing 20% of the float. Volume has been declining recently and the short ratio, or days to cover, is approximately 40, which is the highest it has ever been. The stock has seen stubborn resistance at the $7.50 level a few times in the past year, so if you are considering going long you might wait for it to break through that level. But I think above that level we will start to see some short covering.
Going forward, we will continue to see the balance sheet pared down, with the result being less uncertainty about the value of the assets and the adequacy of the provisions, and more visibility for the remaining assets.
The underlying value appears to be well in excess of the book value. Where it ends up will depend mainly on the strength of the housing recovery, the sale price of LNR, the pace at which it resolves the remaining NPLs and monetize assets, and any initiatives it takes to modify its capital structure to better accommodate the long life of the land holdings. There are, of course, risks, especially from the global macro arena, but hopefully this article will help you to see SFI in a different light. Being perceived as a mortgage lender as opposed to a housing recovery play is negatively affecting sentiment toward the stock. I believe that is partly the reason why the stock has not appreciated along with other companies exposed to the housing recovery. This will change as its evolving asset composition becomes more visible.
Disclosure: I am long SFI.