Spent most of yesterday sorting through my database full of info... updating, upgrading and quite a bit of deleting. Long over due as it was a colossal mess, and I needed a refresher over what I had. It was good and got a ton of ideas about things I could cover... and which in all likelihood will be forgotten by time I can get around to it.
And in that spirit, I'm doing a post on something I just randomly stumbled upon while surfing archive.org today. Found a old page that was once from the foreclosurescanada.com site and thought it would make an interesting topic.
It's just basic information on 88 foreclosures in Edmonton and Calgary from a couple week of July, 2003. Amount owing, original mortgage amount, month issued and interest rate. Figured it made a nice little random sample to examine, and even better, it's from a period when the market was far more balanced then what it's been like the last four or five years.
We can test some axioms, and see where it leads. In articles and interviews I've heard it stated that most foreclosures occur in the few years of the loan. So, lets start there:
Seems that claim holds up, at least as far as our block sample goes. Over 60% of those foreclosures had occurred within three years... over 75% within five... and 95% by year ten.
We can also see that most occur in the second year in particular, over 1/4 in fact. Years one and three are the next highest at around 16-17%... from there it seems the general trend it that the longer the mortgage is held, the less likely it is to default.
Those that looked at the linked page may have noticed that a few of those had some rather extreme interest rates, as high as 20%... obviously not made to prime borrowers. So, I also ran the numbers again for rates less-than-or-equal-to 10% just for shits and giggles and so we can see what it looks like for move conventional loans.
Pretty much looks the same, except years one and two are slightly lower, and the rest slightly higher. Not really a surprise, as obviously people taking those elevated rates are higher risk and the rates themselves make the vicious circle complete.
Of the 12 that had the >10% rates, four were foreclosed upon within a year... six in the second year... and the remaining two in the third year. Typically these were smaller loans, often under $25,000, so presumably second mortgages, but a couple were over $50,000.
Now we'll look at a slightly different angle, this is how much is owing relative to the original loan at the point of foreclosure... or more accurately, how much has been paid off. By far the highest range here are those that actually owe more then their original loan... almost 40% (also did this graph for loans with interest rates at or below 10%)
The incident rate quickly drops to about 24% of foreclosures for those that have paid back between 0-5%. In this sample almost 2/3 foreclosures involved borrowers with less then 5% of their loan paid off. Oddly the rate plummets in the 5-10% range, then bounces back up, but I'd chalk that up to it being a small sample. If we had a larger sample the curve would probably be much smoother.
In any case, we can see that the less principle one has repaid, the more likely they are to fall into foreclosure... or conversely, the more principle one has repaid, the less likely they are to fall into foreclosure.
This would obviously jive with our findings earlier in this post... and as one should expect, those defaulting early in their mortgages life would in all likelihood have repaid less principle.
While I wouldn't take any of these stats as gospel, as this was just a random block sample, at least it was a fair sized one and the results appear as one would expect intuitively. Anyway, I thought it was interesting to analyse, hopefully you feel found it interesting to some degree.
Thursday, October 15, 2009
Comments for this post
All comments