Which opinion is more reliable: A well-informed one tainted by financial arrangements or one free of such conflicts but based on limited data and possibly colored by spite?
Investors in securitizations are now getting more of the second kind, though not as often nor in the precise form regulators hoped for when they enacted reforms encouraging unsolicited opinions.
For example, Fitch wasn’t hired to rate single-borrower securitizations in the fledgling single-family rental market, but it recently assessed these transactions’ risks.
The company’s analysts recently said in a ratings comment that issuers risk not be able to refinance large balloon loans underlying deals when those loans mature. Fitch mainly cited debt yield, a metric that shows a property’s capacity to support financing. Debt yield is calculated by dividing the property’s cash flow by the debt. Three recent deals in the single-family rental sector had debt yields below 5%, compared to roughly 9% as typically found in Freddie Mac multifamily securitizations, the rating firm noted.
Fitch also has capped its ratings for the single-borrower, single-family rental deals at a midrange investment grade rating of single-A, initially due the sector’s lack of operating history, and more recently due to this refinancing risk.
In contrast, presale reports by other ratings agencies generally indicate refinancing through securitization is just one option in a maturity default where a loan can’t refinance. They base credit and in some cases top triple-A credit ratings more on the potential recovery through foreclosure and property sales, a practice Fitch disputes.
“From our perspective, especially for investment grade debt, you shouldn’t have to foreclose. The property should generate sufficient cash flow that a take-out lender at maturity would feel comfortable refinancing,” said Dan Chambers, a managing director at Fitch.
Unpaid ratings agency opinions became more accepted after 2007 when waves of downgrades hit a huge number of mortgage securities that previously had top triple-A ratings, according to Mark Adelson, a veteran bond market executive and researcher known for his early warnings about mortgage securitization risks before the downturn.
“Back in the ’90s or early 2000s a rating agency would say of unsolicited ratings, ‘This is sour grapes,’ and might get an investor to dismiss those opinions. It’s a lot harder for the sell side to make those arguments today,” said Adelson, who is now chief strategy officer at BondFactor, a company created to help improve insurance claim payment certainty in the municipal market.
Unsolicited ratings or ratings comments may be unbiased by seller payment but still competitively biased, particularly given that competition has grown because the downturn created opportunity for new entrants who could bill themselves as untainted by pre-crisis ratings mistakes.
But the bias in unpaid ratings reports creates balance because it tends to present a more conservative view of deal risks, rather than downplaying them as paid ratings often are more prone to do.
This is important because regardless of whether ratings agencies are new entrants or former players, sellers generally tend to “ratings shop” in choosing which ones they want to work with, said Adelson. This encourages raters to de-emphasize risks in seller-paid ratings over time because sellers generally tend to hire the ratings agencies that offer the most favorable views of their bonds, he said.
Encouraging unpaid ratings has been the aim of the Securities and Exchange Commission’s Rule 17g-5, which creates a mechanism for paid ratings agencies to register information they have asked for and gotten from sellers, making it available to unpaid ratings agencies for unsolicited ratings purposes under certain conditions.
The rule encourages unsolicited ratings in theory, but in practice it has its limitations, said Adelson.
Ratings agencies not hired by the issuers have published some unsolicited commentary based on publicly available data, but full unsolicited structured finance ratings based on 17-g5 information are scarce, said Brian Grow, a managing director in Morningstar Credit Ratings’ residential mortgage-backed securities group.
There is an obligation related to accessing the 17-g5 information, which paid ratings agencies provide. After accessing information from nine deals, that company must agree to rate the 10th deal, monitor and report on its performance throughout that transactions’ life, he said. It’s costly work to do unpaid.
“There’s not a lot of upside in the current market to publishing a rating on an unsolicited basis,” Grow said.
Some of the 17-g5 information can’t be cited in a ratings report, he noted.
“A lot of the 17-g5 information is confidential,” said Grow.
Also, because the 17-g5 data is based solely on sellers’ responses to other paid rating agencies’ questions, it may not have the information needed by an unpaid rating agency to rate a deal, said Becky Cao, who also is a managing director in Morningstar’s RMBS group.
Morningstar does use 17-g5 information primarily for its subscription-based surveillance of transactions over time, something investors and other market stakeholders subscribe to in order to get access to performance information on a broad range of deals in the market.
Ratings agencies have attempted investor-paid models which would remove the issue of seller-paid bias, but so far it’s been tough to survive on alone, as seller-paid ratings revenue tends to outweigh it, said Adelson.
Differences between paid and unpaid ratings commentary aren’t necessarily huge but even if they are a matter of nuance, together they paint a more well-rounded and multifaceted picture of deal risks.