Tuesday, January 29, 2013

Housing Affordability Metrics

Despite the fact that I’m interested in the debate about whether we’re in a housing bubble and whether we’re headed for a housing crash that takes down our economy, I have no opinion myself. I care what happens, but I don’t know what will happen. Two of my favourite writers on this topic are Larry MacDonald, who likes to shoot down housing bear arguments, and Potato, who likes to shoot down MacDonald’s arguments. I won’t enter their debate except to make some observations about housing affordability metrics.

Housing bears tend to focus on debt-to-income ratios. They look at how many years of income your mortgage (and other debts) represent. Of course, you can’t spend all your income on debt repayment; there’s interest to pay, and you probably need to eat. So, the actual number of years needed to pay off a debt is much higher than the debt-to-income ratio.

If we focus on just the debt-to-income ratio, the situation in Canada seems dire. The average ratio in Canada keeps hitting new records. What is saving us is the current ultra-low interest rates.

This brings us to the other housing affordability metric: payment-to-income ratio. Because interest rates are so low, payments are low, and this metric doesn’t look so bad. In fact, we’re only slightly above the long-term average payment-to-income ratio. If we focus only on this metric, all seems right with the world.

So, which metric is right? In an interview with Preet Banerjee, Ben Rabidoux says that in most centers “we’re still above the long-term norm in terms of the percentage of income it takes to carry a mortgage and that’s with record low rates.” To those who say that we’re only marginally above long-term trends of affordability, Rabidoux says “I don’t see that as a good sign because we really don’t have a lot of downside in rates right now and so when they inevitably normalize that ratio or that reading is going to blow out quickly.”

Rabidoux is right that payment-to-income ratio doesn’t tell the whole story because interest rates are so low. If rates were to rise even modestly, this metric would rise very fast. On the other hand, debt-to-income doesn’t tell the whole story either because it ignores interest rates.

The truth is that the real picture is somewhere between these two metrics. If you read any argument that focuses on just one of these metrics without observing that today’s low interest rates make the metric somewhat misleading on its own, then you should be doubtful of the writer’s conclusions.

10 comments:

  1. Michael,

    Interesting points and I like your take on it. I struggle with this issue myself.

    I agree that looking at just debt-to-income ratios paints a negative picture but it makes sense that ratio would be so high what with the price of homes these days. I suspect mortgages represent a very large chunk of that ratio.

    And payment-to-income ratios obviously look OK in today's low interest requirement.

    The thing is - it feels like I've been hearing talk about the coming housing crash my entire adult life. As a 42-year-old today - I was in my twenties in the 90's and missed that decade's boom and bust. Then came the new millenium and housing prices took off. Here we are 10 years later and I'm still waiting for the crash.

    I'm lucky in that I bought my home in 2000 and have greatly benefited from the rise in home prices and have enough equity that any eventual housing crash will not affect me but I still wonder when and if a housing crash will occur and how bad might it be.

    ReplyDelete
    Replies
    1. @Steve: Thanks. I've had my house a little longer than you have and I've seen its value double (taking into account inflation). But I have no plans to sell it soon. I'm not too worried about the possibility of its value dropping, but I am interested in the effect on the general economy if housing prices don't land softly.

      Delete
  2. Michael,
    you are absolutely right in that most writers avoid looking at the whole picture by whatever reason. Any single number is just a point, not a demonstration of a trend. I think, that payment-to-income ratio would reflect a picture way better, would it be given some depth. I would like to see a graph showing a projection of the all-debt-payments-to-income ratio in relation to interest rate. In other words, how many households will go under, should interest rate increase by 1%?

    ReplyDelete
    Replies
    1. @AnatoliN: You ask an important question, but I don't have the data to answer it. I sure hope that a 1% increase wouldn't take down too many households.

      Delete
  3. Hi Mike
    A timely post for me - I just finished writing a piece for the Globe on this topic (should be published in a few days). As for Ben R.’s comments on the affordability index, I would say that interest rates don’t go up in isolation, as he appears to assume. Looking at business cycle dynamics over history, interest rates and household income tend to rise together.

    ReplyDelete
    Replies
    1. @Larry: You're right that interest rates are likely to be held down until the economy improves and incomes rise. This may save the average person, but outliers whose incomes fail to rise with the averages may be in trouble.

      Delete
  4. Payment-to-income ratios at long term average interest rates would be interesting to look at. Although maybe not some much if incomes do indeed rise with interest rates as Larry MacDonald suggests. I would think that it is probably more wages and interest rates both rise with inflation.

    Back in the '90s I had a 7% mortgage for 5 years (17 year amortization). That was about as cheap as it got in the '90s. I would guess that is pretty close to the long term average mortgage rate.

    ReplyDelete
    Replies
    1. @Greg: Moving away from macroeconomics for a second, I think individuals should consider the possibility that they will have to carry their debts, including mortgages, at much higher interest rates and without a corresponding rise in their incomes. I say this not because I think it's the most likely outcome for most people, but because it's a possibility that they should protect themselves against.

      Delete
  5. Well, the measure I focus on most is price-to-rent (or a full buy-vs-rent analysis) as that's what determines the individual micro decision of where I should live.

    ReplyDelete
    Replies
    1. @Potato: That's certainly a good metric for deciding whether to buy or rent. Most people put too much emphasis on comparing mortgage payment to rent, but I think both metrics should be considered.

      For those who already own a home with no plans to sell and want to know if they're OK, metrics such as payment-to-income and debt-to-income, along with the security of their jobs are important. Curiously, though, I find most people have unrealistically low estimates of the likelihood of losing their jobs.

      Delete