An example of a bad ARM

To understand why FHA & VA ARMs are relatively safe, we need to first look at a less-safe ARM, of the sort that someone might have obtained in, say, 2002. ARMs are usually described by six numbers; for example, an ARM that someone may have taken out in 2002 might have been a “3/1 ARM with 5/2/10 caps and a margin of PRIME+3″. The first two numbers describe how long the initial rate is fixed for, and then how often the rate resets after that initial fixed period, so a 3/1 ARM is one that has a fixed rate for 3 years, and then the rate resets every 1 year after that. The next three numbers are caps on how much the rate can increase–the first cap affects the initial reset, the second cap affects each year after that, and the third cap is a lifetime cap. The final number (really more of an equation) determines how each rate reset works.

So for that hypothetical loan from 2002 (again: a 3/1 ARM with 5/2/10 caps and a margin of PRIME+3), imagine that the borrower who got it was taking out a $300k loan, and got an initial rate of 3.500%. Their initial PI payment in 2002 would have been $1347/mo, and that would have been fixed for 3 years (the “3″ in “3/1″). But what happened to that borrower after 3 years? In 2005, their rate would have adjusted, to PRIME+3, or 3 percentage points above the Prime Rate. In 2005, the Prime Rate ranged between 5.25% and 7%; if the fixed-rate period ended in Jun 2005, it would have been 6%, so the new rate would have been 9% (6% + the 3% margin). The loan does have an initial adjustment cap of 5%, though (the “5″ in “5/2/10″), so that would limit the new rate in 2005 to 8.5% (the initial 3.5% rate plus the 5% initial adjustment cap) rather than 9%. So in 2005, the new rate would become 8.5%, and the new payment would be $2224/mo.

And the rate would continue to adjust annually (the “1″ of “3/1″), so in June of 2006, the rate would adjust again. At that point, the Prime Rate was at 8%, giving a new theoretical rate of 11% (8% + 3% margin). This time, the cap in effect would be the annual adjustment cap of 2% (the “2″ in “5/2/10″); adding 2% to the last actual rate of 8.5% gives a cap for that year of 10.5%, so since the new theoretical rate of 11% is higher than the cap, the rate “only” adjusts to 10.5%, with a new payment of $2616/mo, nearly double the original payment just four years after the loan was taken out.

Those adjustments would continue annually, with the rate potentially going as high as 13.5% (the initial 3.5% plus the 10% lifetime cap (the “10″ in “5/2/10″)); thankfully, in late 2007 the Prime Rate did start to go down rather than up, but by then most people in loans like this, confronted with payments that had doubled in just 4 years, had already gone into foreclosure.

A better example: an FHA ARM

FHA and VA 3/1 ARMs are currently much safer, because those six key numbers are much more favorable: government regulations set them as 3/1 ARMs with 1/1/5 caps and a CMT+2 margin.

A borrower taking out an FHA 3/1 ARM today might get an interest rate of 4%; on a $300k loan, that would result in an initial PI payment of $1432, which would be fixed for 3 years (again, the “3″ of “3/1″). After that, the rate would start to adjust, but those adjustments would be capped at 1% per year (the first “1″ of “1/1/5″), so in a worst-case scenario, in 3 years the rate would go to 5%, resulting in a payment of $1628/mo. (And remember that a borrower getting a fixed-rate mortgage today would probably be getting an interest rate of about 5% anyway.)

Over the life of the loan, the rate would also then adjust annually (the “1″ of “3/1″) but could only increase in similar 1% increments each year (the second “1″ of “1/1/5″), up to a maximum lifetime adjustment of 5% (the “5″ of “1/1/5″). That means that, again in the absolute worst-case scenario, in 8 years the rate could be 9%, for a monthly PI payment of $2329.55.

That worst-case scenario would be unlikely to occur, however, because the CMT is a different index than the Prime Rate, tied to the yield on US Treasury Bills. In order for the worst-case scenario above (of the rate eventually climbing to 9%) to occur, the CMT index would have to climb to 7% (since the rate is the CMT + 2). The CMT is currently at 0.39% (as of 11/4/09), so that would mean it would have to go up more than 6% in that 8-year period; in the last 47 years that the Federal Reserve has been tracking the CMT, there have only been three times where the CMT was more than 6% higher than it was 8 years earlier–in 1979, 1980, and 1981.

So the worst-case scenario for a goverrnment ARM is better than the worst-case scenario for many other ARMs, and it’s also less likely to happen to boot. FHA and VA ARMs aren’t for everyone, but they are definitely an option that consumers with some tolerance for the risk that their payment might go up by a relatively small amount each year should consider–especially if they are already planning on using an FHA or VA loan to minimize their down payment.

Interested in learning more about government ARMs? Contact us and we’d be glad to answer your questions.

Possibly related posts (automatically generated):

1. More on ARM sticker-shock
2. Estimating Monthly Payments (the old version)
3. Refinancing to lock in a fixed rate

It’s worth noting, however, that while their numbers are fairly accurate, they make two assumptions that don’t necessarily hold true in this market:

• They use the national average ARM amount of $300,000; in my experience, borrowers in this market have been getting ARMs in closer to the $450k-$600k range. That means that the increase for the “typical” ARM that they calculate would be closer to $500-$750 more per month, rather than the $385 that they calculate.
• They note that most 3/1 ARMs have caps that limit rate increases to 2 percentage points per year. In this area, at least, most lenders actually have three different caps that apply to their ARMs–a lifetime cap that limits how high the rate can ever go (usually around 5-8 percentage points higher than the initial rate), a per-year or annual cap (often the 2 percent cap that the article describes), and most importantly, a “first-adjustment” cap that supersedes the normal annual cap, and which is often the same as the lifetime cap. For an ARM with 8/2/8 lifetime/annual/first-adjustment caps, then, the initial sticker shock could be much worse–and unfortunately, much harder to predict, since the exact amount of the initial adjustment will depend on the specific value of the index that the ARM is tied to.

Especially for homeowners whose ARMs have large first-adjustment caps, it's extremely important that they not base their final decision on just one possible outcome (such as the best-case or worst-case scenario, or even a middle-of-the-road outcome). Instead, they should look at all of the possible outcomes for both refinancing and staying with the current ARM, and do a thorough risk analysis: outcomes should be sorted into categories like “unacceptable and need to be avoided”, “painful but tolerable”, and “desirable”, and then the options that lead to unacceptable outcomes should be eliminated, and contingencies and hedges should be considered to see which “painful but tolerable” outcomes might be worth the risk.

Interested in figuring out if you should refinance or not, but not sure where to start? Let us know and we'll gladly help you analyze your options, and put you in touch with a good lender if you do decide to refinance.

Possibly related posts (automatically generated):

1. FHA/VA Adjustable-Rate Mortgages Can Be A Great, Safe, Deal
2. Refinancing to lock in a fixed rate
3. Real Estate News Roundup

If long-term rates were rising but short-term rates were staying low, sticking with an ARM might make sense for more risk-tolerant homeowners who wanted to try to “ride out” their ARM in hopes that the savings on lower short-term rates now would offset the higher rates in the future; similarly, if short-term rates were rising but long-term rates were staying low, some owners might opt to stick with the lower ARM rates for now, betting that they would still have a chance to refinance before long-term rates rose significantly.

Since all signs point to both rates continuing to rise for the forseeable future, however, I'd advise most holders of ARMs (or hybrid ARMs whose fixed-rate period will expire in the next few years) to at least look into refinancing to lock in a fixed rate. Many homeowners who bought in the DC area in the last few years did so with less than 20% down; for many of those homeowners, refinancing now would also present a good opportunity to take advantage of the accumulated equity that the hot market has created for them, by allowing them to refinance without a second mortgage or PMI if their current equity is greater than 20% of the value of their home.

Interested in finding out about your options for refinancing? Drop us a line and we would be glad to talk to you about your situation, including showing you some scenarios for how your monthly payments could change over the next few years if you do or don't refinance. If you think refinancing might make sense for you, we'd also be glad to put you in touch with a good lender.

Possibly related posts (automatically generated):

1. More on ARM sticker-shock
2. FHA/VA Adjustable-Rate Mortgages Can Be A Great, Safe, Deal
3. When Should You Lock In The Rate On Your Mortgage?