Five CMBS Case Studies
Barclays' CMBS team has completed several CMBS narratives of portfolios that are still current but no longer covering debt service. However, instead of a the traditional quantitative black box analysis, BarCap has provided a unique "story" to each case, providing a much needed and comprehensible perspective to the recent sudden rise and even more spectacular fall in commercial real estate. The current state of economic disrepair means a material amount of properties backing up CMBS will likely go from borderline viable on the margin to fully blown collapses very quickly. We recommend this report as it presents the upcoming crash on the "margin" in easily understandable language.
Five CMBS stories, by Aaron Bryson, BarCap U.S. CMBS Strategy
For several months, we have expected an above-consensus increase in the pace of fixed-rate CMBS delinquencies, especially in recent vintages. The key driver, in our view, is the high volume of current, but high-risk loans, specifically those with most recent debt service coverage of 1.1x or less. Our core thesis is that there will be a much higher rate of delinquencies for such loans than in previous CMBS cycles, given the magnitude of property value declines and overall economic stress during this cycle. We have started to see this trend occur in recent months, and we expect it to escalate. In this article, we select five large, recent vintage fixed rate loans in different property types that are still current, but not covering debt service. With the assistance our surveillance team, we profile each loan in greater detail and attempt to explore borrower behaviour.
First, we revisit the current delinquency landscape, focused on recent vintage, fixed-rate loans. For 2005+ vintage loans, the 30+ day delinquency rate is 6.1%, up from 1.1% at the end of 2008. We still believe that we are in the fairly early stages of this rise. This view is supported by the large volume of current, high-risk-of-term-default loans. This includes any loan with a most recent DSCR NCF of <1.1x. Of the $454bn of 2005+ vintage fixed rate loans that are current, $81bn, or 18%, have a most recent DSCR <1.1x, including $55bn <1.0x (Figure 1). Most of these loans have a mark-to-market LTV well in excess of 100%; the latest Moody’s CPPI release reports that CRE property values are down 43% from the peak in October 2007, and back to late 2002 levels.
Exploring borrower behavior
We do not expect every one of the high-risk loans to default; however, we do expect a much higher percentage of them to default than in prior CMBS cycles. There could be many reasons why a sponsor will continue to stay current even if cash flow does not support the mortgage debt service, for example: availability of pre-funded reserves, expectation of pickup in cash flow due to new leases, use of excess cash flow from other properties to support underperforming property, etc. Many large CMBS loans have a story attached, which may be difficult to grasp from purely quantitative data or conventional default models. We focus on five such stories in recent vintages across the major property types to try to learn why each borrower remains current and whether it will stay that way:
- Hotel: $310mn Shutters on the Beach & Casa Del Mar Portfolio A-note in CSMC 07-C4
(largest loan at 15.0% of deal, in CMBX.4); also, associated with the borrower’s interest in this property is $72mn of mezzanine debt.
- Multifamily: $375mn The Belnord A-note in JPMCC 06-LDP9 (7.8% of deal, in CMBX.3).
- Office: $380mn Columbia Center pari passu A-note, $300mn of which is in MSC07-HQ12 (largest loan at 15.4% of deal, non-CMBX); the remaining $80mn was not
securitized. Also, there is a $100mn B-note outside of the trust.
- Industrial: $305mn Schron Industrial Portfolio A-note in GCCFC 07-GG9 (4.7% of deal,
- Retail: $206mn District at Tustin Legacy A-note in LBUBS 07-C7 (6.5% of deal, in
CMBX.5) backed by a transitional property.
Shutters on the Beach & Casa Del Mar Portfolio
Many floating-rate investors are likely familiar with these hotels, as they have a lengthy history as CMBS collateral. The 198 room Shutters on the Beach hotel was first securitized in a floating rate deal, JPMCC 2004-FL1A. This hotel, along with the 129-room Casa del Mar next door, is owned by Edward Thomas Companies.
Shutters on the Beach began development as a Park Hyatt brand property in 1989. In 1992, the developer entered bankruptcy and construction was halted. The Edward Thomas Companies, in partnership with Goldman Sachs, acquired the fee simple interest in the property in 1992 and invested an additional $37.5mn to complete construction. The property opened to the public in June 1993.
In February 2004, it was appraised at $128mn, and a $60mn A-note/$32.5mn B-note loan was originated in April 2004, along with $20mn of mezzanine financing with an April 2006 maturity and extension options. The hotel was reportedly to undergo a $10.5mn renovation to all guest rooms.
The loan was quickly refinanced. In November 2005, Shutters on the Beach was combined with the Casa del Mar hotel as collateral for a $113.9mn A-note, a trust junior component of $6.3mn, a $104.8mn B-note, and $40mn of mezzanine debt for a total debt balance of $265mn. The appraised value of the two hotels was $320mn as of August 2005, of which $204mn was attributed to Shutters on the Beach. The loan was securitized in JPMCC 2006-FL1A, with a November 2007 maturity and extension options.
In June 2007, at the height of aggressive underwriting, the sponsors refinanced again and extracted considerable equity. The hotels qualified as collateral for a seven-year fixed rate loan. A $310mn fixed-rate, interest-only loan with a coupon of 5.5% was originated and securitized in CSMC 07-C4. The loan had a 7y maturity, and there was $72mn of associated mezzanine debt at a rate of 9.8%. The combined A-note and mezzanine debt balance of $382mn equated to $1,168k loan per key. The appraised value of the properties was $450mn, of which $270mn was allocated to Shutters on the Beach. The sponsor planned $11.7mn of renovation for the two properties in 2007. From February 2004 until June 2007, the property value appreciation estimate from appraisals for Shutters on the Beach was 111%, or 26% annually (Figure 2), compared to a 51% cumulative return in the Moody’s CPPI Index over the period.
The pullback in hotel demand, especially at the high end, has had a significant effect on these properties. The heavy reliance on business travel (47% of the demand comes from corporate transient and 20% from meetings and groups, as reported in 2007) adds additional pressure. The most recent DSCR at the A-note level for the TTM period ending August 2009 was 0.88x, reflecting a 31% drop in NCF from at underwriting in 2007 (Figure 3). The A-note NOI debt yield now stands at 5.6%. Monthly occupancy reached a low of 51% in December 2008, but has since rebounded to 85% in August 2009 due to “very aggressive marketing efforts,” according to servicer reports (Figure 4). The website for Shutters on the Beach lists specials for advance purchase rooms for as low as $335 per night.
We expect the Shutters on the Beach & Casa Del Mar Portfolio loan to default in 9-12 months with a significant loss due to the excessive leverage. The borrower may have remained current as it suspects a sharp rebound in hotel demand and rates, as in prior recessionary cycles.
However, we believe luxury hotel demand will remain weak in 2010, especially in the business segment. Our loss-given default is based on a combination of a cap rate on net cash flow and value/key comparables. There has been limited sales activity in the luxury hotel sector; however, some recent transactions provide data points for reference (Figure 5). In particular, we believe the recent sale of the St. Regis, Monarch Beach, is a meaningful comparable. We apply an 8.75% cap rate to trailing NCF, which roughly translates to a value per key of $535k. This value estimate, along with advancing and liquidation expenses, results in a loss of 55% to the A-note and a 8.3% deal loss contribution to CSMC 07-C4. Also, we assume a 24 month liquidation period.
Much attention has been focused on recent developments surrounding the Peter Cooper Village & Stuyvesant Town loan, which was based on pro forma assumptions about converting rent stabilized apartment units to market rent units. However, this was not the only property that followed this strategy. The Belnord, the collateral for a $375mn 10y interest only A-note securitized in JPMCC 06-LDP9, followed a similar strategy.
The property, located on the Upper West Side of Manhattan, is a 215-unit multifamily building erected in 1903 and owned by Extell Development. The borrower has spent over $50mn on renovations since purchasing the property in 1994 for a reported $15mn. This suggests that the sponsor extracted a significant profit at the time of the 2007 origination. The debt amounts to $1.7mn per unit, excluding the retail portion which includes 60k sf of retail space. The underwritten NCF was $29.9mn, based on assumptions of turning rentcontrolled units to market rent, versus in place cash NCF of only $11.2mn.
At securitization, the breakdown of the units was 74 market, 22 rent stabilized and 119 rent controlled. The unit mix at securitization had an average monthly rent of $2.22 psf, with rent controlled, stabilized, and market rent units at $0.53psf, $1.35psf, and $5.19psf respectively. The underwritten rent at stabilization was $6.66psf. As of December 2008, only a net three units had been converted to market rates, well below the pro forma conversion rate. To cover the shortfall, a debt service reserve of $49.4mn was funded; the current balance is only $29mn.
Given the recent court ruling regarding Peter Cooper Village & Stuyvesant Town, The Belnord strategy is in serious jeopardy. The current DSCR as of June 2009 is 0.55x, due to the sponsor’s inability to deregulate units and the property incurring higher-than-expected operating expenses. It is highly unlikely that the property will reach its stabilization in April 2012 as planned.
Pre-funded reserves will likely continue to support his loan for the next 2-3 years. We expect a term default with a significant loss after we estimate that the debt service reserves run out in October 2012. The sponsor, Extell Development, has several development projects underway in New York City, which could limit its ability to support the debt service on the Belnord. Applying a cap rate of 6.5%, assuming minimal improvement in cash flow, and factoring in advances and expenses, we expect a loss of 55% to the A-note. This would result in a deal loss contribution of 4.3% to JPMCC 06-LDP9.
For the office sector, we focus on a purchase loan, instead of refinancing. The 1.5mn sf Columbia Center office building in Seattle was part of the EOP/Blackstone/Beacon flip and reflip activity. EOP originally purchased the property in June 1998 from Seafirst Bank for $404mn, or $280/sf. In June 2000, it raised $195mn in a single asset CMBS deal: Columbia Center Trust Series 2000-CCT. After Blackstone purchased EOP for $39bn, it quickly flipped $6.5bn of properties located in Seattle and Washington, DC to Beacon Capital Partners, including Columbia Center.
Beacon found long-term financing through the CMBS market, in the form of a 5y, interestonly $480mn mortgage loan with a 5.6% coupon, comprising a $380mn A-note and a $100mn B-note. The A-note was securitized in MSC07-HQ12. The appraised value of the property was $648mn, or $426 per sf. This is another example of a pro forma loan, as the underwritten net operating income was $31.8mn, versus trailing NOI of $20.5mn, based on expectations of rolling leases to higher market rents. Based on in-place cash flows, the Anote DSCR NCF was approximately 0.81x at origination. At closing, the sponsor escrowed a $3.5mn cash reserve that would remain in place until the NOI reached a 1.05x DSCR. In addition, if the NOI decreased during the loan term, the sponsor would be required to increase the amount of the cash reserve to result a 1.05x DSCR on the whole loan. A $10mn up front reserve was also funded at securitization for future leasing costs.
Pro forma expectations have not materialized. The June 2009 A-note DSCR was 0.73x, with an occupancy rate of 73% versus 89% at origination. The whole loan DSCR, which should drive default probability, is only 0.57x. Law firm Heller Ehrman (8% of NRA, 13% of total rent, Figure 6) went bankrupt and vacated its space. According to the servicer, Beacon was in the process of leasing some of the Heller Ehrman space and has requested using a portion of the $10mn rollover reserve to fund leasing costs. The current reserve account has a balance of $9mn. In addition, according to REIS, effective rents in the Seattle submarket have declined from $30.2 psf in 2008 to $24.9 psf in Q3 09, with further drops expected through 2011. If these forecasts are realized, there would appear to be little near-term upside from a modification.
We expect pressure from tenant vacancies to escalate. Amazon.com, the largest tenant, is highly likely to vacate or downsize on its lease expiration in 2011, as it is completing an 11- building campus that would lead to a consolidation of its space in the Seattle area. However, due to the existence of the reserve accounts, we do not expect a term default, but a maturity extension of 36 months with a 35% eventual loss. The current NOI A-note debt yield is only 5%, suggesting significant refinancing pressure even if conditions improve before 2012. If the performance of the building deteriorates further, a maturity default is likely, with a higher expected loss due to servicer advancing and liquidation costs.
Schron Industrial Portfolio
The collateral for the Schron Industrial Portfolio is a diversified pool of 36 industrial/flex properties across Long Island, NY. The property was originally purchased by Ruben Schron and other investors from First Industrial Realty Trust in October 2000 for nearly $180mn. The purchase was financed with floating rate debt. In late 2006, the sponsor refinanced and extracted considerable equity. Goldman Sachs provided a 10y $305mn interest-only loan with a 5.5% coupon. The financing represented $125mn over the original purchase price back in 2000. At time of origination, the appraised value of the property was $390mn on trailing NOI, which translated to a cap rate of approximately 5%. We marked the loan as pro forma, as the underwritten net cash flow was $24.7mn, based on the expectation of rolling leases to higher market rents. At origination, over 50% of leases expired prior to 2012, which at the time was viewed as a significant positive.
The pro forma upside has not yet materialized. As shown in Figure 7, occupancy has been on the decline, as the sponsor struggles to find tenants for the expiring leases. Net cash flow was assumed to rise to the underwritten value of $24.7mn; instead, it has slumped to an annualized $15mn, which is below the annual debt service. The sponsor has increased marketing expenses for leasing.
We expect the property to remain under pressure and the loan to hover around a 1.0x DSCR NCF coverage for a prolonged period. Our base case scenario assumes a back-ended default, as we believe the loan is a potential modification candidate. The loan is backed by a portfolio of 36 properties; one potential modification could be a release and sale of some of the better performing properties to partially pay down the principal balance. We apply a cap rate of 8.5% to recent NOI, which results in a loss of 35% to the A-note, with a 24-month liquidation period.
District at Tustin Legacy
The last loan, from the retail sector, is a transitional story. The $206mn District at Tustin Legacy loan is backed by a newly constructed 985k shopping center in Tustin, California. The project was part of the redevelopment of the 1,600 acre former Tustin Marine Corps base. The center is anchored by Costco and Lowe’s, which are independently owned and not part of the collateral, along with Target, which is on a ground lease, and Whole Foods, which is part of the collateral. The collateral consists of 522k sf, resulting in a loan balance psf of $395. The property is owned by Vestar and Kimco, a major shopping mall REIT. The 6.9%, 10y interest-only loan was originated in November 2007 and securitized in LBUBS 07-C7. We view this loan as transitional. At securitization, construction at the property had not yet been completed. As of November 2007, only 82% of the space was occupied, despite leases signed for 99% of space. An upfront reserve was established with a balance of $540k. The transitional nature of the property was likely a factor in the higher loan coupon; we estimate that the relative loan SATO was +55bp. (Loan SATO stands for spread at origination, and is adjusted for comparably issued loans made at the same time. A number > zero suggests a higher-than-average loan spread and higher perceived risk from the issuer standpoint.) The borrower stated that the property would stabilize in 1H08.
Also, the loan terms are unique in that they suggest a partnership with the City of Tustin. The city must approve any lease of space over 20K sf and receives 25% of all percentage rents payable to the borrower. The sponsors invested a significant amount in infrastructure improvement in the area around the site.
The property has yet to stabilize, due to delayed store openings and lower-than-expected rents. The completion of the project coincided with a severe housing bust in the region. MSA level unemployment touched 11.9% in September, well above the national average, as a large portion of employment was tied to housing. High unemployment and falling property values have contributed to weak retail spending and net absorption in Orange County (Figure 8).
The 2008 NOI for District at Tustin Legacy came in at $13.1mn, versus a budgeted amount of $16.8mn. This was due to a decline in average rent to $20.7 psf, versus $26.2 psf at underwriting. Mid-year financials suggest a similar result for 2009, and the most recent DSCR was 0.86x. The loan has been on the servicer watch list since November 2008, and the upfront reserve has been depleted. On the positive side, occupancy has rebounded to 99% as of June 2009, but we suspect that rent concessions have risen as well.
We see continued pressure on retail rents in Orange County; however, we expect the sponsors to give this new development more time to stabilize. Also, the site benefits from strong anchor tenants: Costco, Lowe’s, Target and Whole Foods, with long lease terms. Despite a current DSCR <1.0x, our base case assumption does not include a term default for this loan. We suspect that reported cash flows could improve as free rent periods expire. Also, we see some degree of political downside for the sponsors if they walk away, given the partnership-like structure with the City of Tustin.
However, the current debt yield is only 6.2%. Even under a recovery scenario, we see tremendous refinancing pressure at loan maturity in 2017. For now, we extend the loan 24 months post-maturity and assign a 15% loss, based on applying a 7.75% cap rate to a stabilized NOI projection of $15mn. We believe this loan could also be a modification candidate.
We randomly focused on five, large “high-risk” loans in recent vintages to explore borrower behavior and the potential for default. Across the $81bn of high risk loans, we see the mostaggressive pro forma assumptions and underwriting in the largest loans. As summarized in Figure 9, we expect losses on each of these loans ranging from 15-55% in our base case, over different time horizons. In terms of exploring behavior, we suspect that these loans are still current due to either pre-funded reserves (The Belnord, Columbia Center), expectations of a sharp recovery (Shutters on the Beach & Casa Del Mar, District at Tustin Legacy), hopes for a modification (Schron Industrial Portfolio), or some combination. As reserves dwindle, we are sceptical that a recovery will be sharp enough to narrow the tremendous gap between these loans and the new underwriting paradigm.
Another common theme is that we see little equity at stake for the borrower, especially in the refinance loans. Many recent vintage borrowers extracted considerable equity at the peak of the market; as a result, we are generally sceptical about the efficacy of loan modifications for such borrowers unless they bring significant new equity into the equation.