This has been a much requested topic, and one I would have touched on sooner but I made the mistake of over-thinking the question initially. It wasn't until BMO came out with a report last month, that I realized it was far simpler an analysis... in fact, incredibly simple.

Now, obviously all the variables cannot be accounted for as we know banks generally lend variable at rates of prime-plus-, or prime-minus-, thus leaves an infinite number of possible combinations... but with a little intuition we know that whatever the rate is relative to prime would merely result in a shift in data. So, you can use your imagination, we'll be doing all the calculations using the average five year rate straight against the prime rate.

Here is a look at how the two have moved over time. Obviously track very similarly, though prime is generally lower (but not always). These are spot rates though, and to get an idea of how the two stack up in practice over time we much take into account how they perform relative to each other over their 5 year terms, and even over the entire life of the mortgage.

This is a more useful presentation of the data, comparing the 5-year fixed rate at any given moment against the moving-average of prime over the next five years. You'll notice it is remarkably similar to the graph from the BMO report, that's cause it's the exact same data, except their's only goes back to 1975.

As we can see, in general variable has provided the more beneficial choice the vast majority of the time, as BMO noted, 82% of the time since '75, even higher at 89% going back to '51... and it has exclusively been the better choice since 1987. That is not surprising though, as rates have going down generally since 1981.

The times fixed has been the better option has been during periods immediately proceeding large spikes in interest rates. Beyond that, during periods of generally rising rates (the period up to 1981) even when fixed is the less preferable option, it tracks much closer to how variable performs.

So, as rates have for all intents and purposes hit absolute bottom, such info is worth considering going forward as rates are sure to rise above current levels in the years to come. That goes for whether you're buying, or if your mortgage is coming up for renewal. This may be an ideal time to hedge your bets and go fixed, at least while rates are still near record lows.

Here is a bit of a different perspective of the previous data. Any time the line goes into the red area, fixed was the cheaper option... conversely, when it's in the yellow, variable was the cheaper option.

We kind of already discussed this, but this just gives you an idea of the proportional difference in these cases. It went as high as 4.4% into fixeds favour, and as low as 8.8% into variables. Over the presented period, the average was -1.33%, and median of -1.14%, both in favour of variable.

So, on average there is certainly something to be gained by going variable, but it's usually only a 1-1.5% advantage. So in times of uncertainty, such as now, you need to carefully weigh the potential positives and negatives of either route when you make a call like this.

This is another take, assuming if a person went fixed or variable at the start, they stayed with it through the remaining renewals for the full life of the mortgage. Here we can see how payment advantages from prior periods can compound over time. This is because at different rates, you also pay off different amounts of principle during any given term (except the final term of course)... basically, the lower the rate, the more principle you will pay off.

So if you make the right choice in one period, it will pay off for the remaining life of the mortgage... conversely if you got the other way, you'll be paying for it for the rest of the mortgages life. This measure is a measure of how much is saved relative to the more costly option.

We can see that over 25 years (five terms) the scale of the advantage is much larger. Going as far as 28% in favour of variable. But again, remember this is over periods when the majority of the life of the mortgage would be while rates on going down.

We're not in that situation today, in all likelihood rates are heading up at least a point or two in coming years, thus, pay particular attention to the data in the 50's and early 60's when in a similar situation. During that period, the results were much more mixed.

The dotted area is of mortgages and/or terms have not year completed, and the further right you go, the younger the mortgage, and thus the more the data will respond to future data. Those plots/figures have not been included in any figures I've discussed, they're just there in case you were interested in such things.

Finally I just threw this one in for shits and giggles, this is what they'd look like if we had 30-year fixed rates like they do in the US, rather then our preference for the 5-year terms. This is using the Canadian data as I'm too lazy to look up the US numbers, and from experience their 30-year rates are close, if not often lower then our 5-year rates (no wonder our banks are so profitable huh?!) and prime would be very similar.

We can see here the results are more muted then the prior graph. During periods of increasing rates, fixed performs better, and during ones of lowering rates, variable does. No surprise, that's what intuition would suggest.

I know real estate is still ugly in the states, but you gotta figure in a lot markets prices have returned to their long term trendlines, if not dipped below. That combined with this current interest rate environment, there may never be a better time to buy if you're a potential first-time buyer south of the border. Even if prices slide a bit further, you've locked in at a sweetheart rate, and thus probably still better off.

We north of the border on the other hand, are at the opposite end of the curve, still near the top (or in some cities, right at it) of the bubble. We still have a world of hurt to come before we return to market conditions that are ripe for those looking to enter the market, buying in now is more a debt trap then anything.

Anyway, hope this answered some of your questions about the historical performances of fixed vs. variable rates. But, if you have any more, fire away!

## Thursday, November 12, 2009

Comments for this post
All comments