Hot Topics in CMBS Loan Servicing (and Some Thoughts on Document Drafting for CMBS 2.0)

With this post I've decided to take a discussion from the MBA Servicing Conference last week ("Hot Topics in CMBS," which included panelists from CWCapital, KeyBank, Principal, Moody's and Situs and focused on current servicing problems associated with CMBS loans) and examine loan documentation, particularly what CMBS 2.0 has to correct.

One story ricocheting around the conference involved the servicer who performed an audit of its CMBS loans only to discover that some 6,000 trigger events had been missed when it boarded the loans for servicing. This speaks to the mismatch created by high loan volumes and structural complexity on one hand and the constraints of servicing systems and processes on the other.

To minimize servicing headaches, loan documents should be (i) clearly organized and consistently executed, (ii) closed with deliberate speed and (iii) conservatively structured. Simple enough, but takeaways from servicers reveal how actual practices can be off the mark.

About those loan triggers, for example. Multiple conditions precedent to borrower obligations require special scrutiny from servicers. If you cannot reduce the narrative language to a straightforward formula of the "if this, then that" variety, there will be problems. If the event is not objectively determined or inherently involved (even those DSCR and LTV tests), be careful. Where there is an objective measure involved (be it LIBOR or CPI), what if that measure goes away? If there's an "on switch," is there an "off switch"? You have to spell these out. One panelist described a cash flow sweep provision that went live with a DSCR floor and, for lack of a cure provision, became a "life of loan" feature. In another case, lease approval conditions were so strictly drafted (no "or as otherwise acceptable to lender" language), that a new lease could not be approved in a timely manner and a market lease benefiting the project was lost.

One can hope that springing escrows will be a less frequent loan feature anyway. As the Moody's panelist reminded the audience, the lender only gets "credit" from that rating agency for in-place escrows. (See my earlier post on this subject as well.)

And for those aspiring draftsmen out there, here's another suggestion: the language that showed up in the special conditions to the terms sheet might not suffice for the loan documents to be correctly interpreted and serviced. It's rarely a good idea to cut and paste that terms sheet language into documents, because the expression of deal terms in that context is largely conceptual, clearly truncated, and often intentionally finessed. But it has happened.

Consider inaccurate cross-references and mis-referenced exhibits. It's certainly true that the trustee exception process in CMBS picks up many document defects (missing exhibits and signatures, for example). But mistaken cross-references are unlikely to get picked up by the trustee, and those can prove mischievous. It may not be particularly elegant, but "Intentionally Deleted" section and exhibit headings serve the practical purpose of limiting that kind of tie-out.

Key provisions can also be misplaced and be missed altogether by servicers. If the same shop originates and services the loan, there's an opportunity for all material loan information to transition seamlessly to servicing. Where those functions are split, as happens with third party servicing arrangements, the servicer may have to be part forensic expert to do its job. Some provisions can show up in unexpected or unusual places. Permitted mezzanine debt, for example, can, depending on program architecture or the draftsman's whim, appear in the due-on-sale clause (a transfer of ownership notion), SPE covenants (as an exception to restrictions on debt—even though it's not debt of the borrower), or as a special provision altogether. Elegance is aspirational, to be sure, but this is the situation to aim for that hobgoblin of small minds—consistency. It's ultimately fine if things are spotted in the usual places. Where they are not, as in the anecdote one panelist shared involving a guarantor financial covenant trigger for a lockbox that appeared as an add-on to a borrower property covenant, things do not end well. I happen to be a fan of hit-them-between-the-eyes section headings for key loan features for this reason.

Considering the breadth of loan document problems encountered, one servicer put it this way: anytime manual intervention is necessary to interpret loan documents, the servicer has heightened risk and the CMBS process has, in a real sense, failed.

Loan originators should appreciate the consequences of complexity. Or loan servicers should adjust their pricing of servicing rights for high-touch loan structures. One panelist shared the story of a contingent performance escrow structure that was so involved (including interest rate changes, among other things) the loan had to be serviced manually using an excel spreadsheet. A bit less rocket science in this business might not be a bad thing.

Speed-to-closing and cost constraints invariably lead to less quality control as well. A variety of functions will be re-priced in post-recession CMBS. With the attendant liability from Regulation AB and other financial regulatory reform initiatives, not to mention skin in the game requirements, that now confront market participants, the scope and depth of what's required to close and securitize loans has been transformed.

Comments (0)