Bank mergers roll on unabated; Odd rate sheet pricing; Demand for non-agency loans picks up!

http://globalhomefinance.com 

The number of banks in the United States grows smaller every week. Not because they are being shut down and folded into others, but because of mergers and acquisitions. I am sure that we can expect the same in residential lending as smaller firms decide the costs are too great to go it alone and join forces with larger companies, and/or owners decide to take the proverbial chips off the table. But returning to recent bank news, in Minnesota KleinBank ($1.6B) will acquire Prior State Bank ($206mm) for an undisclosed sum. The parent company of Platte Valley Bank ($445mm, NE) and two other banks will acquire Mountain Valley Bank ($157mm, CO) for an undisclosed sum. In Virginia Middleburg Bank ($1.2B) will sell its 62% stake in mortgage banking firm Southern Trust Mortgage to Sonabank ($716mm), EVB ($1.0B), and executives of the mortgage firm. Arkansas's Simmons First National Bank ($3.2B) will acquire Delta Trust & Bank ($431mm) for about $66mm. William H.W. Crawford IV, President and CEO of Rockville Bank and Rockville Financial, Inc. and Richard B. Collins, United Financial Bancorp, Inc.'s Chairman, President and Chief Executive Officer, today announced that the Connecticut Department of Banking has approved Rockville Bank's application to merge with United Bank of West Springfield, Massachusetts.

Pricing for protected classes aside, let's take a look at something odd occurring on rate sheets everywhere. I received this note: "Rob, the difference in my rate sheet prices has gotten way out of whack. (Editor's note: why doesn't anyone say, "In whack"?) In the past it was about .5 for every .125%, but now the price differences are all over the map, but much higher. What is going on?" I am happy to answer that question, which basically involves several aspects of the capital markets, and simplify it somewhat - but it is a complex topic. That being said, every LO worth their salt should know a little about this topic in order to converse with the secondary marketing group. 

To start, every agency loan across the nation is priced to the sum of the price of the mortgage-backed security into which it will be placed (think of them as buckets), the value of its servicing in the market, and the sum total of loan level price adjustments. Oh, and we can't forget profit margins. The agency MBS market is liquid, its prices easy to obtain. (Due to an archaic but accepted method, MBS prices are quoted in 32nds; thus 99-16 equals 99 and 16/32, or 99.50, 99-08 is 99.25, etc. For this write-up, we'll switch to guns, uh, I mean decimals.) After converting to decimals and rounding a little, a recent price run showed Fannie 2.5s at 92.50, 3.0s at 96.50, 3.50% at 100.625, 4.00% securities at 104.00, and 4.50% securities at 106.75.  

So far so good. Investors will pay more for pools of higher rate/yielding mortgages, which makes sense. (The math is actually more complex, but we can talk about that, and how mortgage rates are set, some other time, like partially in tomorrow's - Saturday's - commentary.) But something smells fishy, and it is not my cat Myrtle's wet food. I did the math, and the differences between the securities/buckets, spaced .5% apart, are 4.0 points, 4.125 points, 3.375 points, and 2.75 points. In the "old days", lenders could basically count on a ratio of 4:1, so that every .125 in rate equated to .5 in price (4x.125). This was because the security market priced MBSs that way through supply and demand. It was pretty much linear around 100.00 but got a little flighty away from par which we will conveniently ignore. So using the above example, prices might have been more like Fannie 2.5's at 96.625, 3s at 98.625, 3.5% securities still at 100.625, 4% securities at 102.625, and 4.50% securities at 104.625. 

So there are big differences between where security prices are trading now versus where they were in the past, relative to each other. We still have the value of the servicing (usually higher for lower rate loans, given the tendency to stay on the books longer), the loan level price adjustments, and the profit margins. Every lender out there has the same MBS market prices as a base. After that, anything goes! Investors in servicing might value servicing different for different agencies. Investors and/or aggregators may have different loan level price adjustments, including buyup and buydowns (which are much higher and punitive now than in the past) and different profit margins. And note that, for the most part, we're not seeing these price differences as much in jumbo pricing: the difference in .125% in rate is much closer to the historical norm of .5 in price. 

Yes, lenders set their own profit margins and control their own overhead, but unfortunately for scores of capital markets personnel who set rates and prices every day, they are at the mercy of the agencies, who set the LLPAs and buyup/buydown grids, and large investors, who set the value of servicing and their own profit margins. And let's throw in some additional nuances. Let's say a particular bank is behind in satisfying its CRA requirements. It will either cut its profit margin on those loans, or increase the pay-up to sellers in order to increase the flow of that business. And while that is happening, other investors have to compete. There will be more information tomorrow, but if you have more questions, ask your capital markets person about "convexity." 

It's no secret that many equity firms have started to monetize the poor housing market with a basic business model of: buy distressed homes, lease them, securitize them, repeat. So much so, that Congress has started to take notice. As a reminder, in late January a Bloomberg article entitled "Wall Street Bonds Draw Scrutiny Where Subprime Spread Mortgages"; John Gittelsohn and Heather Perlberg wrote about this niche practice which has become known as "Rental Bonds"; a practice which has caught the attention of first-term Congressman Mark Takano. "Deutsche Bank AG led the first rental bond sale in November, raising $479 million for Blackstone's Invitation Homes, which has spent more than $7.8 billion on 41,000 homes. Goldman Sachs, JPMorgan and Wells Fargo & Co. are now preparing to sell as much as $500 million in bonds for American Homes 4 Rent (AMH), the second-largest landlord, with more than 21,000 homes. JPMorgan and Credit Suisse Group AG are working with Colony American Homes Inc., a property company run by Tom Barrack that owns more than 15,000 rental houses." The implications of such practices remind many of similar practices leading up to 2008, and call into question similar fiscal responsibilities of buyer and sellers alike.

In the realm of securities financing, "nearly one-half of dealers reported an increase in demand for funding of non-agency residential mortgage-backed securities (RMBS), and two-fifths of respondents also noted increased demand for term funding against such collateral," the Fed said. Here is the "SCOOS" report released by the Federal Reserve yesterday. 

Supply and demand drives mortgage prices, and the market activity Thursday was forgotten after the unemployment data this morning. As a benchmark, the yield on the U.S. 10-yr T-note closed Thursday at 2.79%, and prior to the numbers it was unchanged at 2.79%. Nonfarm Payroll came in at +192, but January was revised +15k and February +22k, for total revisions of +37k. The Unemployment Rate came in at 6.7%, and Hourly Earnings were unchanged. Right after the data rates improved slightly, and the 10-yr is down to 2.78% and agency MBS prices are a shade better.

http://globalhomefinance.blogspot.com