The “why are they doing them” question is easy. Their regulator is telling them to. And Wall Street and the investment community support them because they make a lot of money off them.
I’ve just seen a good example of why few others call Fannie and Freddie’s CRT program into question. After the paper “Credit Risk Transfer and De Facto GSE Reform” was published by the New York Fed as a staff report, I wrote a note to the authors, sending them two of my blog posts and the blog comment I made last week about their paper on the program. They (collectively) sent back a note that I did not feel adequately addressed my criticisms, so I wrote the following (repeated in full despite its length, which I apologize for).
“I am coming at this from the perspective of a former CFO of Fannie, who for fifteen years was responsible for pricing and managing the company’s credit guaranty business in a way that maximized its risk-adjusted return to shareholders.
As you point out, CAS insures against unexpected risk, not catastrophic risk. Were I still Fannie’s CFO, I would not have issued any CAS since the inception of the program, because of my assessment that the interest payments on them (which, if not paid, would be retained earnings for my shareholders) are far greater than the dollars of credit loss I could hope to transfer to the buyers of the CAS tranches, given the quality of loans the CAS are insuring and the amount of the first-loss risk I still retain. As CFO of Fannie, all CAS would be doing for me is reducing my company’s earnings volatility (that’s what “unexpected risk” is); they won’t ensure its solvency. Paying huge premiums for insurance against a remote risk we (the company) are able to bear ourselves (earnings volatility) would be a very bad business decision, not a “success.” It’s that simple.
The fact that Fannie and Freddie are in conservatorship doesn’t change this analysis. Today, the “taxpayer” gets all of the companies’ earnings. By issuing costly CAS and STACRs against high-quality books of business, the companies are reducing the payments to taxpayers now, in exchange for the possibility of being able to transfer modest amounts of credit loss (but far less than what they’re paying in CAS and STACR interest) to risk-sharing investors in the future. That, too, is a bad decision for taxpayers.
In theory, this could work out differently over an entire business cycle. But the problem is that when the delinquency rates on the older books of business start to rise, CRT buyers will withdraw from the market, and not buy future issues. True “success” for an issuer of CRT securities is if it can transfer more dollars of credit risk to CRT buyers than it pays them in interest. But that same outcome is failure for the CRT buyer, and if they think there’s any chance of that happening they’ll run. The CRT issuer thus will end up paying more in interest than it receives in credit loss transfers on its “good” books of business, and bearing all the credit risk itself on its “bad” books. That’s also not “success;” that’s a badly flawed program.
Briefly, in response to your four points:
1. In your profit margin table, you’ve left out the expected dollar amount of risk transfers from the CRTs. Fannie and Freddie surely model that. When I was Fannie’s CFO we modeled our credit guaranty profitability over 500 different combinations of random home price and interest rate environments (with the most extreme being the threshold beyond which we had a “catastrophic” loss). Fannie and Freddie won’t tell us what these scenarios show, but I would be willing to bet that CAS and STACRs increase projected profits in fewer (and probably far fewer) than 10 percent of those scenarios. Again, that’s not “success;” it’s a waste of money.
2. Overpaying for the insurance on your good books then not being able to get any insurance at all on your bad books makes the program worse, not better. And as I noted in the analysis I did for my post “Risk Transfers in the Real World,” even during a downturn CRTs issued against older books of business don’t help a company that much, because of their first-loss thresholds and the fact that the tranches can pay off before the credit losses hit.
3. I didn’t say Fannie and Freddie are diversified across business lines (they aren’t; because of their shrinking portfolio business they’ve effectively become mono-line credit guarantors)—their credit risk is diversified across product types, regions of the country and origination years. This means that, at the entity level, they can use excess earnings from their good books to pay losses on their bad books, and remain solvent. That doesn’t happen if you try to manage risk just with CRTs. The fact that some investors made money on your 2014 CRTs doesn’t mean that others will be willing to lose money on the ones you want to issue in 2020. They won’t; they’ll walk away, and you can’t claw back the excess payments you made to the 2014 CRT buyers to help you cover the losses you’ll be stuck with on your 2020 book.
4. I agree that CRTs are priced in a competitive market. My sole argument is that Fannie and Freddie (or other future credit guarantors) should be able to compare their potential guaranty fee pricing (or profitability) with CRT insurance and without it, and make the choice that is best for their shareholders (or homebuyers). That’s not happening now. With the companies in conservatorship, FHFA and Treasury have mandated that they issue CRTs. Once we’ve decided on what a reformed mortgage finance system should look like—and the credit guarantors in that system have been given updated and binding capital requirements—CRT issuance should not be mandatory; it should be based on economics, in which CRTs are assessed on the basis of their equity capital equivalency, and the guarantors are free to use them when they believe they will result in more efficient guaranty fee pricing, but in the alternative are not required to.
I’d be happy to continue this discussion, perhaps through a conference call.”
One of the authors quickly wrote back with this brief response: “Thanks again for your thoughts on the topic. At this point we would choose not to further discuss this given our different points of view, which seem somewhat ideological, and hence we likely won’t bridge the gap.”
I found this answer curious. I had never heard anyone refer to a risk management technique as being “ideological”; it either is sensible and economic or it isn’t, for financial and analytical reasons that have nothing to do with ideology. So I Googled the authors and found out that two of them are executives at Annaly Capital Management. That made sense. Their answer sounded to me like, “we love these securities as investors because we make a ton of money off them and take very little risk, and choose not to discuss your criticisms of them.” The big surprise is that the NY Fed would publish a blatant marketing piece under its Staff Reports series. But that’s just where the reform dialogue is these days.