Deal Agent: Prerequisite or Pipe Dream for PLS?
Stephen Kudenholdt of Dentons chaired an interesting panel on the new Deal Agent Proposal Sunday afternoon at ABS Vegas. Participating were Yehuda Forster of Moody’s, Dmitri Rabin of Loomis Sayles, Eve Kaplan of US Bank, and Jeff Berg of Clayton.
Significant progress has been made in defining the nature of the Deal Agent role. Triple A investors’ demand for fiduciary responsibility (the “F word,” as it was termed) has been distilled down to “duty-of-care” and “duty-of-loyalty”: standards specific enough that potential Deal Agents feel comfortable assuming them. Three primary areas of responsibility have also been defined, clarifying responsibilities:
- Reps & Warranties Review
- Reporting Reconciliation
- Servicer Oversight
This would seem to suggest a role that is well enough defined, and limited enough in liability, that firms would be willing to provide the service. And yet…we don’t see deals printing with Deal Agents. Why not?
One explanation given was that the quality of the underlying loans is so high today that that there is no real need for a Deal Agent on super-prime deals—why pay for insurance that will never be used? There was general agreement that Deal Agents become more important as credit quality deteriorates. And yet…it was noted that there are sub-prime deals that are printing that not only lack a Deal Agent, but also contain weak investor protections for breach review.
In particular, automatic reviews are not a feature in some recent deals. Instead an investor vote is required to direct a review. Ratings agencies to whom we have spoken at DealVector consider these to be weak reps and warranties frameworks given the current difficulty in getting investor votes (something that, ahem, DealVector enables), and are penalizing these deals by around 200 bps in credit enhancement. Still, the deals are selling. Why?
Panelists offered one explanation and one caution. The explanation is that we are in a collateral constrained environment, these were small deals, and you can always find a way to place a $100M offering since not many investors are needed.
The caution was that these weak investor protections on reps & warranties could become the template for future deals as the PLS markets start to expand. This would leave us in the same state as pre-crisis…or just perpetually stalled. The question is really who blinks first: investors or issuers.
Another reason that the Deal Agent is having difficulty getting traction is the age-old question of who pays. One way to solve this would be to recognize that a Deal Agent makes the deal safer for investors, and this should be reflected in lower credit enhancement or other favorable treatment for deals, partially or fully offsetting the cost. However, while Moody’s and the other credit ratings agencies believe that a Deal Agent is “directionally positive” for credit, Yehuda made it clear that it is very difficult to quantify this (and it can vary by deal). Without a way to quantify a benefit to credit enhancement, it becomes harder for issuers and investors to arrive at an appropriate price they’d be willing to pay—let alone to know whether that price would be sufficient for potential providers to assume the risk of the role.
Ultimately the Deal Agent is an attempt to establish a new “best practice,” and new best practices are incredibly hard to create. They require buy-in from a large percentage of the market (with players whose interests don’t align perfectly), and—because they are new—the data doesn’t exist to determine how much value adoption would create. This of course makes it difficult to get the required large percentage of the market on board.
Dmitri referred to analysis suggesting that losses in the GSE pools (which provided a model for the Deal Agent framework) were less than half the losses in PLS pools, even after controlling for loan quality and type. If the Deal Agent could truly prevent half of all loses during the next period of stress, the structure would clearly pay for itself. But it was also noted that the GSE’s have other benefits helping them control losses, including incredibly dominant market power. How much of the better performance can be replicated by Deal Agents, and how much is inherent to the GSE’s themselves? Until we go through another stress period—with Deal Agents in place on some deals and not others—we won’t have the data to know. Yet unless we implement Deal Agents in advance of the stress period, we won’t get the data either.
Someone stopped by the DealVector booth yesterday and said he felt that our industry moved in 9 year cycles. If you date those 9 years from 2007, things are going to start going south again soon. If you date it later, we’ve got a few more years. We do know, however, that the credit cycle will turn at some point. Will we have put in place the best practices needed to better weather the next storm in time?
