Scotiabank released a report today saying 40-year amortizations are “changing the longer-term risk dynamics” of Canada’s mortgage market. (Financial Post story)
Three out of four insured applications are now for 30 to 40-year amortizations, says Scotia.
40-year amortizations comprise 1/2 of that total. (Don’t be surprised if this 1/2 grows to 3/4+ over time.) Prior to 2006, 25-year “ams.” were the standard for decades.
According to the Financial Post, Scotia’s Derek Holt (the report’s author) cites some potential risks of this trend:
- If homeowners someday start paying off their 40-year mortgages sooner, and en masse, it could heighten [prepayment] risk to buyers of mortgage-backed securities containing these mortgages. That could in turn cause problems for mortgage originators.
- If people become highly leveraged using 40-year amortizations, and the economy goes in the dumper, the repercussions to our housing market could be material.
We have no stats to back this up but most people seem to be getting 40-year ams. because they need them (to get the home they want). If wages sink and unemployment spikes, you have to wonder how many 40-year mortgagors might be living too close to the edge to keep up their house payments.
On the other hand, given the quality underwriting standards we see day in and day out, it’s difficult to foresee anything akin to a U.S.-style collapse heading Canada’s way. Long-term though, Holt says nobody knows what the effect of 40-year ams. will be.
We’ll try to get our hands on Scotia’s full report and follow up on this story shortly.
Last modified: April 25, 2014
Actually, I think its quite easy to imagine an American style collapse.
Yes, we have different underwriting standards. That just means that people can afford the mortgage in the first place, so we won’t have people defaulting because their rates reset.
However, many people are taking out the absolute biggest mortgage they can afford, having only a tiny downpayment, and amortizing over 40 years. That means they have ZERO wiggle room if something bad happens. And rates are very low right now. If the housing market goes down, or even stagnates, they may not be able to refinance when the mortgage is up, especially if rates increase. And this scenario will affect a LOT of people. Somehow I doubt someone who could afford their mortgage when it’s at 4.6% can afford it at 6.6%. (A 2% increase is not unrealistic in 5 years)
Or, if they find that they have to sell before the mortgage is up, the realtor’s commission will eat up more than the small amount of equity they have.
I’m not certain this is a significant concern; if as you say our clients want to maximize their ability to buy (and therefore impact on their cash flow) at 40 years won’t they do the same at 25?
Yes a 2% increase in rate has somewhat more impact at 40 years than at 25 but the incremental difference in monthly increase is dwarfed by the overall increase in total payment.
At the end of the day brokers still have the responsibility to insure our clients are fully advised now and for the future and not in over their heads as a result of our attempts to maximize commissions; which was very much a part of the U.S. problem.
Hi Diane,
Excellent points and thanks for the comments.
As with every major housing correction (and Canada has had some big ones), a lot of people will get hurt. Even with 25-year ams. there were always a contingent of people who overextended themselves before a downturn. The point above is just that a U.S.-style collapse seems unlikely (to us).
In the U.S. people were way underqualified for their mortgages up front so many were doomed to default. For example, people were being qualified on 3% option ARMs that were destined for 7%+ two years later. There was no possible way people could afford those payments from the outset. The defaults that resulted greatly exacerbated the U.S. housing market plunge and tripped off a vicious cycle.
That type of thing isn’t happening in Canada. With most lenders in Canada, if you want a 4% variable rate, for example, you still have to qualify at posted rates (e.g. 7%).
Regardless, I do share many of your concerns. Debt loads are growing and it’s been economic bliss for so long, that we seem due for a “break.” (but who knows when or for how long…)
Just a comment on pre-payment risk.
Generally MBS’ have payout rates of 10% per year which means 40-50% of the mortgages are paid out prior to maturity. A 25 vs 40 year am doesn’t make much of a difference.
MBS’s almost always sell at a discount. So if a mortgage pays out early the investors get 100% of the balance back instead of the discount they payed.
MBS pools also pay a rate differential indemnity for mortgage payouts which compensates the investor for re-investment risk
Hi Rsimpson,
Thanks for the perspective. You’re right about people’s proclivity to maximize their buying power, regardless of amortization.
You’re also dead on about our duty to advise borrowers properly as to affordability and future risks.
John,
Thanks very much for sharing this info on MBS pre-payments. Your comment that “A 25 vs 40 year am doesn’t make much of a difference” is one of the reasons we’re trying to get our paws on the Scotia study. I’d like to see exactly what Holt had to say on this topic.
Cheers,
Rob
“A 25 vs 40 year am doesn’t make much of a difference” ”
But how long have MBS existed? If less than 10 years, then their entire performance history occurred during a time when house prices were consistently rising and interest rates were falling.
How will models of default and prepayment risk stand up on the other side of the cycle?
My understanding is that pre-payment always works out in the worst possible way for the MBS holder:
If someone is locked into a long term fixed rate and interest rates rise, almost no one would do the MBS holder the favor of prepaying (and freeing up the MBS holders now “low yielding capital”) because they would do better to make their scheduled mortgage payments and use excess cash to buy bonds which yield more than their mortgage interest. If interest rates fall, those same people would just refi which creates a pre-payment event for the MBS holder who loses the value of their now relatively “high yielding” paper.
I bring this up because the Americans got into so much trouble by building these statistical models that it turns out were based on “history” that didn’t include an environment of falling house prices. In fact, their models were annihilated when house prices simply went flat!
Will the black-box models created on this side of the border turn out to be any better? Tune in next week, for “As the MBS turns!!!”
Tom,
I promise it’s a soap opera we will watch avidly. :)
Interesting enough, prepayments are most commonly correlated with falling interest rates and/or rising home prices. These are two major factors built into the pricing models you refer to.
There’s one thing we’re not so sure of, however. A lot of people take out 40-year amortizations largely for payment flexibility. They swear they will make repayments every chance they get to lower the effective amortization. If even half these people are telling the truth, these pre-payments could be a major factor not fully priced into MBS.
Maybe this was the risk Derek Holt was referring to (we’re not sure). Incidentally, we read his report and it didn’t make reference to the pre-payment risk that the Financial Post had noted in its story. So we can’t be certain what he meant. Maybe we’ll give him a call.
Cheers,
Rob
Canadian MBS pays a rate differential indemnity to the bond holder.
The indemnity increases as bond rates decrease.
So if a mortgage pays out when rates have risen the investor gets zero indemnity, 100% of the principal instead of the discount they paid and they get to re-invest at a higher rate
If a mortgage pays out when rates have fallen the investor gets an indemnity equal to their re-investment risk,100% of the principal instead of the discount they paid and they re-invest at a lower rate but are compensated by the indemnity
It won’t work out perfectly but it does backstop the downside on holding Canadian MBS
Hi John, Excellent and helpful post! Thank you!
Ok I have a question here. Putting an end to 40 yr amortizations, does that include 35yr amortizations as well? I live in Fort McMurray, AB and considering the average family home here is $800,000, a 35yr amortization is the norm.
Hi Ken,
How are things up in oil country? Property values still holding up?
35-year amortizations are safe for now. So are uninsured mortgages with 40-year amortizations.
Cheers,
Rob