The End Of The (Mortgage) World As We Know it, Again

Institutional Investor News - October, 2006

News Flash: Interest rates are climbing from historic lows. Mortgage rates have moved up by a dramatic 100 point basis points over the last 12 months. The realestate market will never be the same. The shock of higher interest rates will wreak havoc on everything that we have come to understand about the mortgage market...

If you suspect sarcasm, you are right. Recently there have been a slew of articles about the home price appreciation rate (or the decline of it) and its effect on every aspect of a borrower's behavior. Clearly, there is some effect from the slow down in HPA, but its magnitude is secondary to the same old questions. What is the long-term connection between a projection of HPA and the probability that a borrower will default? What is the long-term connection between a projection of mortgage rates and resulting prepayments?

The emphasis here is on long-term and connection. An MBS generates cash-flows for about 30 years. It is the long-term connection that determines the value. If one could predict the direction of interest rates in the future, one should not be in the MBS market, one should play in the futures market. If one could predict the direction of HPA in a particular region, one should be investing (or going short) in REITs. Valuation of MBS is entirely determined by the two long-term connections. These connections are generally referred to as default models and prepayment models.

How does one make money in the MBS market? One of the most reliable ways for sophisticated players has been to profit from relative mispricings of CMO tranches. Over the years, MBS market players have become increasingly sophisticated, especially proprietary trading and hedge fund operations, as weaker participants have been wiped out. This has led to the virtual disappearance of clear arbitrage opportunities and the deceased returns in most MBS hedge funds.

Option: Default, Refinance

The MBS investment performance can now come only from one's superior understanding of the two functional connections that I have described. A borrower at the point of loan origination is granted two options—to default and to refinance. There already exist institutional structures to control, value and "fairly" price the option to default. A mortgage with a greater propensity to default (givin borrower and loan characteristics) will be priced lower. "Reasonable" credit quality collateral may have the economic value of the default option between two and ten basis points per year or approximately 0.1%-0.6% of the value of the loan.

Also, MBS prices reflect the differences in prepayment propensities of loans with different ages and coupons relatively efficiently. An example would be to compare FNMA 5.5 versus FNMA 6. A "pool-level" prepayment model is used in these valuations. On the other hand, the pricing of the differential value due to other characteristics like loan size, geography, purpose, FICO, LTV, property type, etc. has been haphazard at best. This, almost always mispriced, and variability in the intrinsic value of the mortgages due to the differential propensity to refinance represents an enormous arbitrage opportunity. The variability in the economic values of otherwise similar (and similarly priced) IOs can be as high as 30%.

For fixed rate, good credit quality, current coupon mortgages, one can calculate the option to refinance to be worth between 20 and 90 basis points per year, or approximately 1-5% of the value of the loan. The value of the option depends on the borrower and the loan characteristics, yield curve shape, and the values of interest rates derivatives. To emphasize the point: otherwise similar loans, priced similarly, may have as much as 4% variability in their intrinsic economic value, depending on the deferential propensity to refinance. To obtain a measurement of prepayment propensity with any accuracy at the loan level, such as compare two FNMA 5.5 loans to each other, a loan level score that summarizes these loans' attributes (other than type, age and WAC) is needed.

An Accurate Prepayment Scoring Model

Applied Financial Technology has developed prepayment scores that summarize all loan characteristics, other than age and WAC, in their effect on loan's deferential propensity to refinance or to move. The scores are used as modifiers to a "generic" prepayment model. This "generic" prepayment model is designed to accurately project aggregate behavior of these loans, without taking into account the borrower and loan characteristics other than loan type, age, and WAC.

The prepayment scores simplify for investors the process of loan valuation vis á vis prepayment propensity in the same way as FICO credit scores simplify loan valuations for their default propensities. The prepayment scores do not change with movements of interest rates and implied volatilities, whereas the values of the MBS prepayment options do not change with movements of interest rates and implied volatilities, whereas the values of the MBS prepayment options do.

The chart below demonstrates the ability of the scoring process to discriminate for refinancing propensity. We scored all conforming loans originated in 1997 with WAC of 7.5% (+- 0.25%). These loans were then bucketed into deciles by score. In the absence of refinancing incentive (until mid 1998 and between mid 1999 and January 2001) these buckets prepaid at similar speeds, but when the interest rates dropped, the higher scored buckets always prepaid faster than the lower scored ones. And the difference was dramatic. It is interesting to note that these loans were priced in 1997 identically to each other, even though their economic values were clearly different. Given the interest rates and volatile environment of 1999, the calculated economic values of these loans differed by as much as 3%.

AFT has scored all agency pools in similar fashion based on additional pool information like geographic distribution, loan size, FICO, LTV, purpose. The score's ability to discriminate pool's prepayment behavior is equally dramatic. Utilizing the score-modified AFT prepayment model, we calculated differential economic values of a variety of pools and compared them to market pay-ups for low loan balance pools.

Extensive CMO Pools Database

AFT has further extended the analysis to agency and non-agency CMOs. One can calculate the scores of the agency CMOs based on the pool scores that back them, and of non-agency CMOs based on the scores of the loans that back them. AFT maintains databases of pools backing all agency CMOs and of loans backing about 80% of non-agency CMOs. Most of the analytical systems do not provide pool-level or loan-level information on collateral backing the deals. The combination of scoring and the databases allows customers to both access the information and efficiently utilize it.

There are a number of ways in which market participants have started utilizing the technology to take advantage of the arbitrage. Several proprietary trading desks and hedge funds have reorganized their trading activities around the scoring algorithm. They trade a long position in a low scored IO and a high scored PO of the same collateral type, age, and coupon, and a short position in TBAs. The trade removes most of the market risk and allows one to benefit from the desirable score-based prepayment behavior. This trade can take a variety of other forms involving more complex combinations of MBS derivatives. Some trading desks have been using the algorithm to deliver lower valued (on the basis of score) collateral against TBAs or cherry picking specified pools when the market does not price the economic value of the score properly. Mortgage bankers have started cherry picking their portfolios by selling off higher scored loans and looking to acquire lower scored ones.

The score allows one to immediately identify desirable MBS without having to perform a loan-by-loan analysis of the collateral backing it and to exploit the last remaining large arbitrage opportunities.

Michael Bykhovsky is president and CEO of Applied Financial Technology, an analytic software firm providing mortgage banks and MBS traders with mortgage loan prepayment propensity scoring.