Click here to join our mailing list to receive the latest news and updates as they happen. Unsubscribe any time.

Variable or Fixed — The Research – Part II

Part II of our interview with Professor Moshe Milevsky follows below…


On the chances of someone being able to predict variable mortgage rates next year:

Dr. Milevsky: One of the least understood “secrets” about successful mortgage financing, is that to predict whether to fix or float you have to accurately forecast two things: a) What is the Bank of Canada going to do tomorrow; and, b) what is the bond market going to do tomorrow. But it’s not that simple. The Bank of Canada can lower rates but long-term yields can still go up. Vice versa is true as well. Having to predict two things correctly is therefore exponentially more difficult than predicting one. That said, I actually am willing to entertain the notion that some observers are able to forecast the direction of the next move of the Bank of Canada, especially as we get closer to the meeting dates.

The Bank of Canada tends to change rates at predictable times. Knowing that, a person may plan to float now and then lock in later. For this strategy to be successful, however, you have to predict when variable rates will go up and you also have to predict when fixed rates will go up. The bet will only pay off if both things happen as expected. This is where people stumble. The long-term mortgage rate is driven by the bond market, and that is truly a coin toss when it comes to predicting.

On the correlation between fixed and variable rates:

Dr. Milevsky: There is a positive correlation between fixed and variable rates but it is not perfect. It’s somewhere between 50 to 60%. What this means is that in any given month, there is a 50% chance that they do NOT move in the same direction.

On how much Canadians should rely on the long-term interest rate forecasts of major economists:

Dr. Milevsky: Bond traders may want to listen to what the community of forecasting economists are saying but I wouldn’t give much credence to it in the arena of personal finance. In the 15 years I’ve been tracking economic forecasts, I would say that broadly speaking, 50% of the time forecasting economists predict right and 50% of the time they predict wrong. Of course, these folks make a habit of reminding you–advertising all the calls they got right. The non-financial Canadian homeowner getting a mortgage shouldn’t get involved with predicting interest rates. As a homeowner you need to look at it as a risk management exercise. What does your balance sheet look like? What are your assets, liabilities, equity? How would an adverse rate change affect your financial situation? Think holistically.

On fixed payment variable-rate mortgages being an optimal solution:

Dr. Milevsky: I like them. I use one myself. They enable people to plan. At the point of sale, people have to understand the different scenarios that can unfold. If I were getting a mortgage I’d want to see a stress testing analysis. What are the probabilities something bad can happen? Give me a chart that says if rates go above this level my payments are going to be this. It should be required. People should know where rates would have to go for their payments to go up, and they should sign off on it.

On patterns that appeared during his study:

Dr. Milevsky: When floating didn’t work there was a certain degree of clustering. If floating didn’t work this year it tended not to work the next year. These cases aren’t as random as you’d expect. In an inverted yield curve we might see some of this clustering. In plain English, if you can get a long-term fixed-rate mortgage for 4%, go for it and don’t be greedy.

On the possibility of using systems (like moving average bands) to help time when to lock in or float:

Dr. Milevsky: People and so-called experts who use predictive strategies are trying to outguess the credit and bond markets. If you can do this successfully, you could go out and make a killing in the bond market. Why waste your time on mortgages? Open a hedge fund. Mortgage basis points are peanuts in perspective.

In short, I’d say I’d be sceptical. At least, I would have to see the data and results before I pass judgement.