Farewell to ARMs

by Brian Pretti CFA, Contrary Investor. September 14, 2007

It was a bit funny last month watching Fannie Mae and Freddie Mac stock prices squirt higher on the assumption/rumor that F&F would act to basically bail out or monetize at least some portion of troubled US mortgage paper. Let's be frank with each other; there is absolutely no way on this green Earth F&F could make this happen, despite the fact that they had in the past been key credit cycle provocateurs when needed most to provide systemic liquidity in the late 1990's. The regulators have been all over Fannie and Freddie in recent years. These two need to petition the regulators at this point to expand their balance sheets. Importantly, one day after such a request from Fannie to expand conventional lending limits, the OFHEO turned down the request without even blinking. Any questions? The numbers simply don't add up for these two to even make a dent in the hundreds of billions in mortgage paper that either is now or will become troubled ahead. These GSE's simply are not going to be of significant credit market help this go around, despite the fact that I'm sure a number of politicians will try to override the OFHEO. But what is very helpful is the fact that our friends at Freddie so kindly give us a glimpse into the world of mortgage refinancings now and again. I'm sure you are more than aware that from literally this month all the way through the summer of next year, we are going to see the largest number of adjustable rate mortgages resetting in this country that we will probably ever see in our lifetimes. The peak comes next March. As I've mentioned many a time, household mortgage issues of substance and the ultimate consequences of this circumstance lie ahead of us, not behind us.

So if you don't mind, in relatively rapid-fire fashion, let's quickly review the most recent mortgage refi data as of 2Q 2007. Important why? It's an insight into consumer behavior, and I believe an insight into levels of financial stress at the household level. The charts tell the story. There are plenty of them, so let's roll through them. Please remember when looking at this data that the real crunch in mortgage credit availability has occurred in the third quarter of this year, not the second. This data only runs through 2Q, so in many senses it may indeed be the final hurrah. A farewell to ARMs of sorts. As examples, both WAMU and Wells, two of the largest players in the sub prime game, shut their sub prime lending operations down last month, not during 2Q. American Home Mortgage imploded in August, not in 2Q. You get the picture. Since 2Q period end, residential mortgage terms and credit availability have tightened up significantly. Let's get right to it and have an initial look at level of cash out refi's through the second quarter. As is clear, 83% of all refi transactions involved a new loan at least 5% higher than the prior loan in 2Q. Implication? At least as of the second quarter, households were implicitly continuing to lever up the largest household asset that is clearly declining in value. Do you really think these were choices households wanted to make? Or were forced to make? In case you have not heard, NEVER in US history have we experienced year over year declines in home prices without being in a recession or depression. Does this help frame the character of these actions?

As stated in the chart, it’s my belief that it will not be long until these historically high levels give way to the downside in a meaningful manner in terms of cash out activity. But what is also very clear is that a substantial magnitude of this refi activity was the act of refinancing ARM loans into fixed loans during the period. Now that ARM rates and conventional fully amortizing loan rates are quite close to each other, it makes all the sense in the world, assuming one can qualify, of course. What the chart below importantly shows us, and reinforces my belief that those who are refinancing are getting out of ARM's, is the history of ARM loans as a percentage of mortgage applications of the moment.

It just seems a shame Greenspan himself was suggesting households take on ARM debt at quarter century lows in interest rates a number of years back, when ARM's as a percentage of total mortgage apps were near 35%. He must have gotten one hell of a deal on a sub prime negative option ARM loan at the time for he and Andrea, right? Sure he did.

Quite important is the fact that the average age of refied loans in 2Q came in at 3.5 years. As you'll see below, it was one of the highest numbers in close to seven years. This is saying something quite loudly about the ability of recent year homebuyers to refinance their loans. It's saying that it's getting tougher and tougher for those who bought in the 2005 and 2006 period to refi. Of course these were the exact years that the percentage of sub prime and Alt-A loans as a percentage of total loans made literally skyrocketed. As you know, one of the conventions in ARM loan production past was to offer 3-5 year lockups on rates before the initial adjustment period commenced. The 3.5 years fits the ARM refi characterization into a fixed like a glove. No surprises here at all.

Very quickly, in 2Q, a whopping 5% of loan refis resulted in a lower loan amount relative to the prior refied mortgage. Again, this is happening while the collateral values are declining. As a quick mea culpa, in the chart below I stated that 95% of loans refied during 2Q resulted in a higher loan balance. My math is incorrect. Here are the correct numbers. Five percent of refis resulted in a lower loan balance, 83% resulted in a higher loan balance and 12% resulted in no change in the loan balance. Regardless of my hasty math error while creating the chart, the current cash out percentage is simply unsustainable over time. I expect a dramatic drop off ahead. But in like manner, I do not expect what you see below to spike up any time soon. Not a chance.

I'll wrap this up with a last two charts and call it a day. The next two charts are important because they contain both historical as well as projected numbers. Below we're looking at Freddie's numbers for historical nominal dollar cash out refi experience as well as their estimates for what's to come both this year and in 2008. Certainly the chart speaks for itself. Say bye-bye to cash out dollars that have supported consumer spending, especially over the last three years.

The final chart helps clarify a bit the magnitude of cash out dollars relative to prior loan balances in the refi process. As you can see, for 2007, the estimate is an average 25% increase in loan balances (cash taken out increasing the loan balance) for cash out refis undertaken. Relative to historical context, this is still one very big number.

So there you have it, quick and dirty. What are the important take aways from this exercise? ARM holders are opting to refi into fixed loans. No big surprise. Households doing refis are still taking out very significant amounts of both nominal dollar cash and increasing their loan balances significantly at the exact time they are fully aware that the collateral against which they are borrowing is shrinking. They know their equity in the property is diminishing; yet they are still taking out cash and increasing loan balances meaningfully. I can't see how this is still happening unless households are under a certain amount of financial stress. At this point in the housing cycle, there seems little other explanation. Finally, given the average age of refied loans in the period, purchasers of the last three years seem to be in a bind when it comes to refi possibilities. We're going to find out starting this month through the entirety of next summer just how much of a bind. Again, please remember these are the numbers and character of refi activity prior to the mortgage credit terms and availability carnage of July and August of this year. As Merle Haggard asks us in one of his hit tunes of yesteryear, "Is the best of the free life behind us now and are the good times really over for good?" Merle, don't you worry pardner, we're going to find out real soon. I'll be sure to update you on the character of 3Q refi activity when I have the data, of course assuming you really want to see it.

Brian Pretti

Contact Information

Brian Pretti CFA | Editor and Publisher, Contrary Investor
P. O. Box 4402 | Walnut Creek, CA 94596 | Observations | E-mail