CMBS 1.0 vs 2.0 | FT Alphaville

CMBS 1.0 vs 2.0

We’ve been monitoring trends in CMBS 2.0 — the next generation of Commercial Mortgage-Backed Securities — since early this year, when it became clear that issuance was set to begin climbing again.

And, of course, they have — with deal sizes, the number of loans collateralising each deal, and the number of tranches all continuing to increase on the initial, more conservative deals of CMBS 2.0.

The last time we checked in, we discussed a Standard & Poor’s presentation showing that underwriting standards had fallen since early last year, though they remained strong compared against the immediate pre-bubble years. Okay, that’s not saying much.

The presentation also showed that the prospect of an outburst of further issuance was likely to be dampened by uncertainty over proposed risk retention regulations and the dearth of critical B-piece buyers.

Well, Moody’s has just come along with some proprietary calculations of its own, having reviewed many of the same deals, and has drawn similar conclusions.

Below are a few highlights from the report, which you can find in full at the usual place.

First, debt service coverage ratios remain well above pre-crisis levels:

Moody’s conduit loan DSCR, an indicator of term default risk, was 1.46 during Q1 2011 (Moody’s DSCR = Moody’s net cash flow/actual debt service). This is consistent with DSCR in the period from early 2004 thru early 2005 (as shown in figure 1) before the bulk of the CMBS 1.0 credit degradation took place. The protective cushion provided by current DSCR is roughly twice what it was in 2007 when it fell into the 1.20’s.

Second, the loans in these conduits are less leveraged than in those bubble years, but leverage is creeping up:

As shown in figure 2, issuer LTV based on then current appraised values was typically in the mid 70’s until the tail end of CMBS 1.0. In contrast, Moody’s LTV highlighted the growing credit risk late in CMBS 1.0 and is much more consistent with the now evident performance patterns of recent vintages.

In Q1 2007, the quarter immediately prior to Moody’s announcement that it was raising subordination levels due to credit deterioration, more than 30% of loans had a Moody’s LTV between 110% and 120%, and an additional 20% had Moody’s LTV’s greater than 120%. In contrast, Q1 2011 saw only about 5% of loans above 110% and none above 120%. Figure 3 also shows that Q1 2004 and Q1 2011 are roughly comparable, but with an edge to 2004 as it had more loans in the low leverage grouping and fewer in the high leverage grouping.

There has long been a focus on the average LTV of pools, but Moody’s believes a focus on the distribution can be equally if not more important. Since the majority of loans in conduits are not crossed (where the strong loans offset the weak) the distribution makes it easier to spot where potential losses and extensions may come from.

And finally, debt yields are still higher:

Debt yield in effect represents the cap rate at which a property could be sold to pay off its debt. Debt yields can be measured against historic cap rate patterns to assess the relative degree of protection offered. They are especially relevant to below investment grade CMBS investors and mezzanine lenders as they have exposure to owning a property at the basis implied by its debt.

In 2007 the average debt yield based on Moody’s cash flow was 8.3%. For CMBS issued in Q1 2011 debt yield averaged 10.5%. This implies that borrowers are now taking out about 20% less proceeds against a given dollar of cash flow (and perhaps even less than that given that are current cash flows are more likely to be beaten up).

Figure 4 illustrates the distribution of debt yields across Q1 2011, Q1 2007 and Q1 2004. As with the distribution of LTV, the 2007 vintage clearly stands out as the weakest, with 2011 and 2004 being roughly comparable. In 2007 about half of newly made loans had debt yields of 8% or less, which is the case in only a fraction of those from Q1 2011. The greater leverage on 2007 loans will manifest itself as greater losses for loans that default during their loan term, and greater balloon risk for the ones that don’t.

The report looks at a few other categories, but the takeaway is the same for each. Underwriting standards remain decent, and that shouldn’t be surprising: the bigger deals of CMBS 2.0 have mostly come this year, so it remains early in the asset class’s return to something approaching normalcy.

But this is something to be watched closely, especially given the nascent signs of deterioration.

Nothing to really worry about yet, but it doesn’t hurt to remember that the relatively benign conditions of 2004 eventually gave way to those of 2007.

Related links:
CMBS issuance and the B-piece buyer problem – FT Alphaville
Premium capture is the new 436(g) – FT Alphaville
Back to the future with CMBS – FT Alphaville