Young people borrow, to finance school, house, car, and kids. Their debt reaches a maximum, somewhere around age 35. Then they start to pay off their debt, and save for retirement, and so reach a maximum stock of savings somewhere around age 65.
So, if there were a big bulge in the population at around age 35, and very few people around age 65, the average level of debt would be highest, right?
If age were the only thing that mattered, and if people were otherwise identical, and if everyone were age 35, debt would be zero. That's because there would be nobody for the 35 year-olds to borrow from. It takes two to tango. A borrower and a lender. Dollars borrowed equals dollars lent.
You would get maximum average debt if the age distribution were bi-modal. Half the population aged 35, and the other half aged 65. The 35 year-olds would want to borrow the maximum amount, and the 65 year-olds would want to lend the maximum amount. Every borrower can find a lender, and every lender can find a borrower. The whole population is on the dance floor, tangoing, because everyone can find a partner. And both partners want to do a lot of dancing.
Now, what sort of country has demographics like that?
Canada does. OK, it's not precisely like that, but the Canadian age distribution is bi-modal. There's the boomers, in their late 40's and early 50's; and the echo-boomers, in their 20's. The boomers aren't 65 yet, and the echo-boomers aren't 35 yet. But wait another 10 or 15 years, and they will be.
Have a look at Stephen's graphs here, especially his animated gif.
Right now, at DMax-day minus 10-15 years, with the countdown on, Canada is approaching its destiny with maximum debt, if this simple little model is correct. So it's no surprise that average debt is increasing over time.
But perhaps Canada, a single country, is not the right unit of analysis. We borrow and lend within Canada, but we also borrow and lend across national boundaries. Perhaps we also need to look at the world age distribution.
Then it gets a bit trickier. Poor countries can neither borrow nor lend as much as rich countries. A rich country, with twice the per capita income across the whole age distribution as an otherwise identical poor country, would be like a country with twice the population. So we would need to get some sort of GNP-weighted global age distribution.
What is the exact statistic of that GNP-weighted global age distribution that we are looking for? It's not precisely the variance of that distribution. It's some sort of descriptive statistic that would take on the value 0 if everyone were exactly the same age; take on the value 1 if half the population were 35 and the other half were 65; and be somewhere between 0 and 1 the "closer" it were to one extreme or the other. Call it the "35-65 bi-modal statistic", because it needs a name, whatever it is.
I can't figure out exactly how to calculate it properly. Never mind, because you should have gotten the gist of it.
But my hunch is that, globally, that statistic has been rising over the last few years. 35 year olds didn't used to be able to find enough 65 year olds to dance with. Now they can. And if my hunch is right, that means more and more people are finding partners who want to do the debt tango. And Global Finance has been having to work harder and harder to do the intermediation for more and more couples of borrowers and lenders, who want to dance more and more dances.
Maybe Global Finance just buckled under the strain of overwork.