It seems that companies like Thornburg Mortgage
Thornburg specializes in purchasing, originating, and securitizing adjustable rate (ARM) jumbo mortgages with an average loan size of more than $800,000. These borrowers tend to be affluent, and so far have been paying off their loans (Redwood Trust
For the quarter, Thornburg decreased expenses (measured by non-interest expenses as a percent of average assets) to 20 basis points, down three basis points from the year-ago period (if this seems trivial, one basis point of Thornburg's roughly $50 billion in assets is $5 million). This, coupled with an $8 million gain on the sale of some assets, as well as a decrease in premium amortization (explained below), boosted net income 10% for the quarter to $80.2 million over the year-ago period.
For the year, Thornburg's mortgage originations increased 13% to $5.6 billion versus Mortgage Bankers Association projections of a 17% decline in originations industry-wide. Thornburg achieved this growth by increasing the number of correspondent lender relationships 33% to 282 lenders, making it the 18th-largest correspondent lender. Thornburg's other channels -- retail and wholesale -- should also have room to run from a low starting point. Thornburg thinks it can continue its hot streak into this year and boldly forecasted a 21% increase in originations to $6.8 billion, despite the association's forecast of a 5% decline in originations for the industry this year.
Thornburg's management also noted that it had $7.2 billion in hybrid ARMs originated in 2001 through 2003 that were coming off their lower fixed-rate periods (at 4.63%) and should reset in the next two years to higher yields (at today's rates, they would be up 150 basis points), which would boost interest income.
One of the most notable points during the earnings call was the confusion caused by the decrease in amortization to $4.8 million -- it's usually a $20 million quarterly expense to amortize the premiums Thornburg had to acquire the loans in its portfolio. The confusion was this: When Thornburg pays a premium to acquire a mortgage, it then has to amortize (expense) that premium over the life of the loan. However, if a borrower prepays, then that premium has to be amortized right away. Thus, prepayments are bad. Thornburg's management, however, noted that prepayments were actually much slower than anticipated. It had believed that almost all borrowers in the last year of the fixed-rate portion of their hybrid ARM loans would prepay -- they now believe that those borrowers will have a much higher rate of survivorship -- and actually stay on at the higher floating rate for a longer time than expected.
The math is kind of confusing: Thornburg on average paid about 72.5 basis points of the premium for its mortgages, which results in roughly $360 million in premiums that have be expensed at some point or another ($50 billion times 72.5 basis points is roughly $360 million). At $5 million of premium amortization a quarter, or $20 million a year, this would imply an 18-year life per loan ($360 million divided by $20 million in premium amortization per year). This is way too high -- however, some of the decrease in premium amortization was a "catch-up" adjustment to bring the rest of the portfolio in line with the less conservative assumptions (still with me?).
In other words, management believes it was way too conservative earlier, and is now overcompensating (or actually undercompensating) their amortization expenses to make up for it. However, amortization expenses will eventually have to be adjusted sharply upward as loans reset to the higher rates, but because borrowers are staying on for a while at the higher floating rate, Thornburg will earn higher interest income along with the increase in amortization.
All in all, Thornburg -- a mortgage lender set up as a real estate investment trust -- had a strong quarter and year in terms of origination, asset growth, credit quality, and ability to fund its portfolio through securitization. If it can continue to dodge loan losses and higher prepayment rates, it should do well; however, this one's tough to understand, so I think I'll watch it from the sidelines.
For those of you who can't get enough of REITs, check out:
Want to know what others think of Thornburg? Check out Motley Fool CAPS, the new investor intelligence community.
Fool contributor Emil Lee is an analyst and a disciple of value investing. He doesn't own shares in any of the companies mentioned above. The Motley Fool has a disclosure policy. Emil appreciates comments, concerns, and complaints.