Since our last look at the curve in February, the long end of the curve has dropped, with the 3 to 5 year segment dropping the most. This has led to a kink at the 6 month to 1 year point, with a small parabola forming in the middle of the curve, while the 5s/30s slope has actually increased. Various articles about the predictive power of different types of inversions in forecasting recessions have come out. Not as many articles have considered whether, in an environment where the distribution of the maturity of Fed holdings is to the right of the distribution of Treasury issuance (and hence the Fed is overweight the long end of the curve relative to the market), those forecasting models should still be considered reliable. Meanwhile, mortgage rates rallied a full 100 bps from their November peaks before bouncing back a bit, potentially stimulating both the Spring home purchase market and reviving some refinance activity.

In this “CRT Edition” of the Pipeline, we present two articles. The first, by Alex Levin, uses the CRT market to look at implied g-fees. After reviewing our methodology (from 2016), Alex looks at the history of implied g-fees and explains some changes that we see, including the impact of the recent deceleration in HPA. He also looks at how much of a shift to our 20-scenarios occurs in going from our physical approach to the market-implied risk neutral approach behind the implied g-fee method.

Our second article is joint work by our Richard Cooperstein, Ken Fears of NAR, and Susan Wachter of the Wharton School entitled, “Economic Principles and Housing Finance Reform.” The authors begin by examining overarching goals and how the pre-crisis structure of the GSEs failed going into the financial crisis. Then they lay out an alternate approach which rebalances the tension between market incentives and public mission by establishing a regulated utility.

We look forward to questions and comments on both of these interesting articles.

Many of us from AD&Co will be at the IMN CRT conference at the end of April and look forward to seeing
you there.

In this article, we summarize our method of inferring implied g-fees from agency CRT analysis and provide an update on recent market dynamics.

• Implied G-fees gauged from the CRT prices remain fairly consistent across different CRT types: STACR vs CAS,
low-LTV vs high-LTV. They currently stand at around 25-27 bps (credit portion only).

• The cost of selling credit risk has been trending up over the last year, most likely due to decelerating
home-price appreciation.

• The technical (crOAS) component reflecting illiquidity remains steady.

The National Association of Realtors (NAR) released its vision for reform of the GSEs in February, co-authored by Andrew Davidson & Co’s Richard Cooperstein, Ken Fears of NAR, and Susan Wachter of the Wharton School. Our analysis is guided by the economic theory of imperfect markets and concludes that a well-established solution would work best: a pair of regulated utilities. This structure achieves the core goals of housing finance more reliably than other options. The core goals are to:

• Maximize competitive investment in mortgage interest rate risk and credit risk
• Assure stability in the mortgage market in times of extreme financial stress
• Put the mission before profits

Clients who license one of AD&Co’s valuation/risk systems (RiskProfiler or LoanKinetics) can replicate these projections. Those who license the LoanDynamics Model (LDM) can use our extended scenario files through their vendor systems’ computations.