On Thursday, Amanda Lang—co-host of The Lang and O’Leary Exchange—made a comment worth exploring.
While talking with Kevin O’Leary about the risks of variable mortgage rates, she stated:
“Some people can forecast [rates] and arrange their affairs accordingly. Rates won’t…gap up. They will climb in some orderly fashion.
…It’s very rare to see a multi-move up or down with interest rates…People of means…can actually take their time…and then still lock in with plenty of time to get a decent longer-term rate.”
If you’ve studied rate history, however, you know that interest rate behaviour doesn’t cater to these assumptions.
In the modern era of Canadian monetary policy (1991 to today):
Rates have moved swiftly at times
- In 1994, prime rate shot up 425 basis points in 13 months
- In 1997, prime rose 250 bps in 12 months
- In 2000, prime fell 375 bps in 13 months
- In 2005, prime jumped 175 bps in 9 months.
- In 2007, prime dropped 400 bps in 17 months
Rates sometimes climb in leaps, not steps
- Since ’91, there have been four instances where multiple 50+ bps rate hikes occurred over spans of six months or less
There have been periods where today’s fixed rates would have outperformed a variable
- Coming off a cyclical bottom, the average increase in prime rate has been 3.16% (that’s from trough to peak, over the last three major rate cycles)
- After these rate bottoms were made, prime rate was 1.23% higher, on average, over the next five years. Put another way, if you had picked the worst possible time to get a variable-rate mortgage (i.e., right before rates increased), your rate would have averaged 1.23% more over the following five years.
Locking in on time isn’t easy
- Traders sell bonds at the first hint that future inflation could exceed the BoC’s comfort zone. Historically, that selling has occurred anywhere from 1-6 months before increases in prime rate. When bond prices fall, bond rates rise in lockstep, which lifts fixed mortgage rates in the process.
- History has shown that it’s costly for variable-rate borrowers to wait for the first increase in prime rate before locking in. Fixed rates have often risen 50 bps or more leading up to initial rate hikes. Waiting for the 2nd increase in prime is even more costly.
- Folks also have to remember that conversion rates are almost always higher than regular mortgage rates. Due to breakage penalties, discharge fees and aversion to change/inconvenience, lenders know that variable-rate customers are captive. Lock in today, for example, and you probably wouldn’t get 2.94% on a 5-year fixed. You’d get 3.09-3.19%, if you’re lucky.
Of course, 22 years of rate history doesn’t tell us what will happen next year, or the year after. The message here is more of a reminder not to overestimate our rate timing ability.
A strategy based on locking in after the first BoC rate increase is usually counterproductive. When the difference between fixed and variable rates is as tight as today’s 39 basis points, most rate-lockers would be better off with a low fixed rate from the get-go.
It also doesn’t help to use rising bond yields as a signal to lock in. The problem there is that the bond market creates more fakeouts than Barry Sanders in his prime. A 75-basis point upmove in the 5-year yield could easily be followed by a 75-basis point drop, leading you to lock in for nothing.
For the record, our sense is that rates won’t rise materially for several months—and when they do, it should be a gradual incline. But that is more of an educated guess than a useful conviction.
When inflation threats eventually appear, they could surprise the market and force bond traders to rapidly reverse their positions. When investors rush to dump bonds, fixed rates can climb like an F-35 on takeoff. And if this were to happen in the next year or so, fears of a deflating “bond market bubble” could intensify this selling.
In short, believing we can lock in “at the right time” is overoptimistic, to put it mildly. Variable and short-term mortgages are indeed the best fit for some borrowers, but anyone with visions of saving ½ point in a variable and then converting to a fixed should get acquainted with history.
Rob McLister, CMT