(This article first appeared in the April 16 issue of International Financing Review, a Thomson Reuters publication)
by Adam Tempkin
NEW YORK, April 18 (IFR) - An aggressive underwriting tactic that was the hallmark of risky commercial real estate lending at the height of the market’s froth from 2005 to 2007 is beginning to rear its head again in loans backing recent CMBS transactions.
According to analysts at Barclays Capital, elements of so-called “pro forma” underwriting -- or basing future property cashflows on (often wildly optimistic) projections rather than in-place or realized income -- have emerged in the latest batch of CMBS transactions marketed by firms such as Goldman Sachs (GS.N), Deutsche Bank (DBKGn.DE), and Cantor Fitzgerald.
The unrealistically optimistic pro forma projections of net operating income made on various commercial properties at the peak of the bubble spelled doom for billions of dollars of CMBS packaged during those years.
One risky pro forma-related characteristic cited by Barclays -- early lease-termination clauses for commercial real estate tenants -- is particularly noticeable on Cantor Fitzgerald’s inaugural CMBS deal, the $635 million CFCRE 2011-C1, which was being marketed last week and is scheduled to price today or tomorrow, according to investors that have looked at the deal.
The offering is the bond broker’s first foray into the CMBS market after building a new CMBS conduit-loan shop last year. Nearly 30 percent of the loans backing the CFCRE transaction are secured by properties leased to the US government on behalf of the General Services Administration, an independent federal agency that supplies products, communications, and office space for the federal government.
Many of these properties have early termination clauses on long-term leases, meaning that the government has the option to terminate a lease after the loan’s maturity but within a three-year period following that date.
While the likelihood of the government taking that option is extremely low, the potential severity of loss would be quite high if it did happen, which dissuaded some investors from taking part in the deal.
“We could not get comfortable with the heavy dominance of the GSA portfolio because of the risk of early termination,” said a senior CMBS portfolio manager.
“Cantor did not exist in this market beforehand, so they’re out there competing for loans against the big boys. There are so many conduit shops with their shingles out in the market now, so the pendulum has swung from levels that were not sustainable in 2006 and 2007, to highly conservative levels, and now it’s swinging back again, with a loosening in underwriting.”
The investor acknowledged that a certain level of risk -- including the possibility of loans with cashflows that are not yet in place -- is baked into every transaction, balanced out by a layer of protection, called credit enhancement, which is set by the rating agencies.
It’s the delicate balance of these two factors that makes a deal attractive to investors and helps them determine the risk premium they should demand for the level of risk they’re buying into.
But it’s difficult for investors to assess the loans in a deal if they don’t know the story behind them.
“Buysiders mainly want transparency and consistency from arrangers and rating agencies when assessing a deal, and have the right to know when a certain loan or set of loans is not being underwritten to the ‘black-line’ standard,” the investor said.
“This way, we can demand extra risk premium for the shoddiest loans. Given how competitive the market is now, and the lower quality of most loans out there, we realize there isn’t as much of a market for loans with in-place cashflow. So as investors, we are always struggling with that fine line” between accepting riskier collateral for higher returns and investing in a low-quality deal, he added.
A ‘DEAD’ MALL?
Other investors doing their own due diligence last week on the loans underlying the Cantor Fitzgerald CMBS expressed additional concerns about the collateral.
One well-placed investor at a large, prominent money manager was particularly frustrated, noting several troubling characteristics, including big-box mall tenants that had lease rollovers in the next couple of years, and so-called ‘balloon-risk’, or the risk that loans coming due will not be able to be refinanced because the value of the property has sunk below the amount of the loan.
The investor noted that one loan in the deal worth $52.4 million was backed by the Hudson Valley Mall in Kingston, NY, which appears on a website titled “deadmalls.com”, dedicated to malls that had or have a bad reputation, or are run-down.
According to the website, there were two violent events at the mall in the last six years, including a 2005 incident involving a man walking into a Best Buy and shooting an assault rifle at shoppers (two were injured), and the 2006 murder of the night manager of the mall’s Ground Round by an ex-employee, an incident which occurred on premises.
The website does acknowledge that despite the incidents, “the mall has managed to escape the bad reputation and continue on.”
This still did not satisfy the investor, who said that the proverbial ‘race to the bottom’ was happening all over again in the CMBS world.
“Originators are at their own peril if they assume there’s a deep bid for ever-more risky collateral,” the CMBS investor said. “While some [buyside] accounts may need exposure to this, we’re going to vote with our feet. We’re not going to send the message that this is okay.”
The investor said that the issuer would have to price the deal wider, meaning that investors would demand a higher risk premium for questionable collateral. That extra cost would then be passed on to commercial real estate borrowers.
“There are borrowers that are willing to pay for that increased leverage,” he said. “It gives them increased ability to get a more flexible loan.”
While the rating agencies (Fitch, Moody‘s, and Realpoint rated the Cantor transaction) supposedly accounted for the higher risks in their determination of credit enhancement for the offering, they admit that some properties underlying the deal had been going through a “destabilized period”, and just recently stabilized.
Therefore, there was an assumption that net operating income (NOI) would increase.
For example, some new leases backing the transaction were signed in late 2010, said Robert Vrchota, a CMBS analyst at Fitch who rated the transaction.
Moreover, some of the newer government and military properties in the deal have no historical financials or operating history, so the rating agencies gave credit to these new leases, which are largely untested.
The Cantor deal is not the only one to contain risky characteristics. As competition between commercial real estate lenders heats up, looser underwriting with increasingly speculative assumptions has produced several loans that made their way into recent CMBS offerings, said Julia Tcherkassova, head of CMBS research at Barclays, on a teleconference last week.
“While the level of pro-forma underwriting may not be back to 2007 levels, we are seeing semi pro-forma underwriting characteristics [in recent deals] where cashflows are not based on the 12-month trailing average financials” of the properties, Tcherkassova said.
This means that potential cashflow projections do not rely on the recent actual financial profile of a particular property, but on assumptions that either a tenant will eventually be in place or will not leave a property earlier than expected.
For example, a $16 million loan from an Albuquerque-based mall portfolio within the $856 million COMM 2010-C1 transaction led by Deutsche Bank, which was priced in October 2010, contains an underwritten tenant (Borders) that is now in lease payment default or bankruptcy.
In other instances, various letters of credit or reserves are subtracted from the loan balance in order to decrease loan-to-value (LTV) ratios. This so called “hold-back” of upfront reserves for the benefit of lowering LTVs existed in a $122 million loan for the Fashion Outlets of Niagara Falls, also within the COMM 2010-C1 deal, Barclays said.
This situation also existed for a $49 million loan on an office building at 123 South Broad Street in Philadelphia which was securitized in Goldman Sachs’ $876.45 million GSMS 2010-C2 offering, which priced in mid-December.
“Not every single LTV has been underwritten using actual occupancy or actual at-risk appraisal,” Tcherkassova said.
Other examples of speculative underwriting include cases where the actual occupancy of a property is lower than the underwritten property, or where there was a substantial step-up in NOI.
The most notorious example of failed pro-forma projections was the historical 2010 default on $3 billion of loans backing the 2006 buyout of the Stuyvesant Town/Peter Cooper Village apartment complex in Manhattan by Tishman Speyer and BlackRock.
The two borrowers acquired the complex from MetLife at the height of the market for a record $5.4 billion. The grossly overleveraged deal was underwritten with the assumption that a number of rent-stabilized apartments would convert to market-level rents, which did not turn out to be the case.
The losses spread through several CMBS transactions that had exposure to the Stuyvesant Town loans.
The CMBS market has returned more quickly and more aggressively than most would have thought possible: more than $7 billion was issued in the first quarter of 2011 - almost as much as full-year volume for 2010.
Adam Tempkin is a senior IFR analyst