Bank Lending - Another information imbalance

I was reading an article in the Sunday Business Post, Banks blame regulator - for wrong reason, and it got me thinking about how this is another example of information asymmetry. We have a small number of banks in this country, they were all giving out developer loans and they were all giving out mortgages, you have to wonder why they would keep giving out loans to developers when it had to be apparent to them that the consumers were going to run out of money.

Take this for example (taken from the Sunday Business Post article):

When the consumer could no longer afford the mortgage repayments spread over the traditional, prudent 20 years, the bankers devised new products to allow the borrowers spread the payments over first 25 years, then 30 years, then 35 years and, in some cases, 40 years.

The banks had to have some projections, and they had to have known that they couldn't keep giving out more money. That there is a tipping point, once you take too much money out of someone's pocket to pay back their mortgage, you can take no more!

So, we have the banks, who have information on the mortgages they are giving out, and they know those mortgages have been becoming a bigger strain on people for several years (with larger salary multiples and longer terms, not to mention tracker rates). These same banks are giving out developer loans, that are also getting jumbo in size. Which anyone with half a brain had to realise would mean higher house prices again, which means either a bigger salary multiple or a longer term.

Let's take a quick look at this. My informal method of analysis for this information is as follows:

I'm not an economist, I'm not saying these numbers have some grand meaning, but they do shine a light on a small portion of the picture. The traditional 2.5 times multiple over 20 years would versus the 6 times over 40 years gives you a ratio of almost 5:1, or an increase of 380%. Now, lets take a look at the change in house prices....

According to the Permanent TSB/ESRI figures, the change in house prices from 1996 the their peak in 2007 was... drumroll please... 280%, or 360% in Dublin. To be honest, that these numbers sort of line up is probably a coincedence, I just thought 'My Index' might be an interesting take on it.

Some Facts and Figures

Let's delve a little deeper though, with some numbers. Before I go any further though, we need to decide what numbers are important, what do I compare. I think that the total cost of the lifetime of the mortgage is of interest. For this analysis to mean anything I will hold the interest rate constant across all the mortages at 5%. The cost of the mortgage is the amount the consumer paid over and above the original loan amount, with the percentage cost being the cost of the mortgage versus the original mortage amount. I will also hold the salary constant, at €35,000 (€2917 per month). The salary multiple is what is of interest here, not the salary itself. All figures are rounded to the nearest euro for simplicity sake.

When I see these figures, the one that always stands out is the cost of the mortgage. Take a look at the 6 times multiple and compare the repayments for a 20 and a 40 year mortgage, there is a drop in the montly repayments of 13 percentage points, which in real terms is almost a 28% drop in the repayment. Great, except... the increase in the cost of the mortage is 71 percentage points, which translates to an almost 125% increase in the final cost of the mortgage.

What does this all mean... well, all things being equal (which they aren't because banks penalise you if you pay back early, not mention the opportunity cost of money), if you have a 40 year mortgage, and you increase the amount you are paying each month by 37%, your mortgage will cost you 55% less, and it will be paid back in half the time.

What have we learned?

Well, I don't know about you, but I still have to concentrate when using statistics and percentages. Especially when dealing with drops and increases as we have here. More specifically, a 10% drop requires more than a 10% increase to bring us back to our starting value, e.g., assuming a starting value of 1, a 10% drop gives us 0.9, a 10% increase from there gets us back to 0.99. We would actually need an 11% increase to bring us back to our starting value.

The banks had to have done all this number crunching, and much more, with big supercomputers, and you would think it would pop out an answer at the end that said either "all is good" or "all is bad". For the last few years of the boom, it should have been the latter, and the banks should have known it.