Friday 25 November 2011

Icelandic Indexation, Part II

In my first entry about Icelandic indexation of debt and credit, predominantly mortgages, I touched on how the cost of inflation is added onto the principal of the mortgage instead of being in the form of higher nominal rates as is the case in most other countries. The feedback I got from one of my readers was “I still don’t understand how this thing works, it looks nuts the way you describe it and it can’t be right.”

I’m sorry, but it is. But admittedly, I didn’t explain in enough detail how the principal first grows and then contracts only much later in the loan period. Second, I didn’t explain how or why the monthly payments are growing exponentially in nominal terms as the loan closes in on its terminal date. So I hope this post clarifies some of the grey zones that I left in the first part about Icelandic indexation.

An example
The best way to explain the functionality of anything is, in my humble opinion, always to take an example. To make things more coherent between this part of the story and the first I’m going to use the same loan and the same given figures. They were: an indexed mortgage of 20.000.000 ISK for 40 years at 5.0% real (NOT nominal) interest rates. Payments are monthly and calculated on an annuity basis, i.e. every single monthly payment (monthly payment = repayment of principal + interests) should be equal given a certain principal that is repaid over the loan period and fixed interest rates. Average inflation is 2.5% (which is the inflation target of Central Bank of Iceland).

The following screen shot from Excel shows how this mortgage would develop in the beginning, i.e. the first year of 40. Notice that the borrower does not owe “remainder of original nominal principal” but “remainder of inflation-compensated principal after payment” once he has dutifully paid the monthly payment. And that amount grows spectacularly in the beginning of the loan period.

Table 1: The development of an indexed mortgage in Iceland. Notice that the "remainder of the inflation-compensated principal after payment" grows even though payment has been carried out (click to enlarge).

In fact, it continues to grow, in the case of only 2.5% inflation, until month 212 (for almost 18 years!) when it finally tops in 24,055,720 ISK. Notice that at that time, the original nominal principal has shrunk down to 15,487,314 ISK. But again, that’s not what the borrower owes. He owes 24,055,720 ISK after having paid dutifully of the mortgage for 17 years and 8 months. And the original amount borrowed was 20,000,000 ISK.

Table 2: The principal owed does not begin to shrink in nominal terms until after almost 18 years in case of the mortgage listed in table 1 (click to enlarge).

But 2.5% inflation is low in Iceland. If we put more historically reasonable inflation rate of 5% (average annual inflation from 1990 to 2010 was 5.2%) the inflation-compensated principal doesn’t begin to shrink until after – hold your breath – 25 years and 11 months. By that time, the nominal amount owed has in fact more than doubled!

Table 3: 5% inflation is historically lot more realistic in Iceland than 2.5%. In that case, the amount owed more than doubles before it begins to finally shrink after almost 26 years (only 14 years left of the loan period, click to enlarge).

For a more detailed comparison, the graph below shows the development of the principal owed in the case of 2.5% annual inflation on one hand but 5% inflation on the other.

Graph 1: The development of the amount owed (ISK) on the mortgage in tables 1-3, given 5% annual inflation or 2.5%.

An enlightening graph is also the development of the monthly payments (payment = monthly repayment of [inflation compensated] principal + interests) in the cases of 2.5% and 5% inflation. The graph below shows the monthly total payments. In the case of 5% annual inflation over the whole loan period – remember the average annual inflation for the last 20 years is 5.2% - the monthly repayment grows by a factor 7. This is not a joke! From this graph one can also estimate that a person who borrowed money with an indexed 40 year mortgage 20 years ago is paying today about 2.5 times more than in the beginning of the loan period. Anyone who wants to bet this person to be bankrupt in 10 years? And does anyone want to think about the feedback between inflation and necessary nominal wages-after-tax? We’ll come to that in later posts.

Graph 2: The monthly total repayments grow exponentially. The higher the inflation is, the higher is the exponential growth. Notice that in the case of 5% inflation, the borrower will in the end pay 7 times higher nominal amount than he did in the beginning of the loan period.

But why does the principal not shrink?
Because the cost of inflation – the indexation – is added onto the principal instead of being included in higher nominal interest rates as is the case of almost every single mortgage system in the world (apparently Chile and Israel, of all countries, do something similar to this but I’ve never found the necessary data to do a complete comparison between the systems. The Central Bank of Iceland loves to make reference to those economies whenever it is challenged on whether the Icelandic system is clever or not. If somebody can tell me anything about the indexation of mortgages in Israel and Chile, please contact me, it would be most appreciated!)

I hope this explains somewhat better the madness of the Icelandic indexation. If you’ve got a feeling that this system cannot work out you’re most profoundly right! But the rabbit hole goes deep and we are getting closer on being able to discover it properly now that, I hope, the functionality of the indexation is somewhat clearer.

If there are any questions that are still left unanswered or you’ve got any other comments, leave them in the comment box and I’ll do my best to answer them. If there are a lot of troubling matters still left to be dealt with – beside the sheer shock of thinking men that some nation thought it might be a good idea to organise its mortgages in such a spectacularly mad way – I’ll try to clarify them further in later post(s).

But surely, I haven’t even begun introducing you to the madness of Icelandic economics. Wait until I explain the pension system!


  1. Aside from the indexation part (which is bad in and of itself), for some reason they structured these as negative amortisation loans.

    I have to think that they must have had some completely bizarre notion about how the banking system behaved and thought this would somehow counter the 1970-80's hyperinflation - but it doesn't in fact work like that at all. I'm fairly sure for most people these things end up being mathematically unrepayable.

  2. This is just INSANE. I thought I was misunderstanding, but apparently and horrifyingly not. Next, the pension system -- which as far as I an tell people are supposed to rob in order to pay part of their mortgages????

  3. Thank you very much for posting this. I've heard about this before but never had a coherent explanation until now. But (without running the numbers) I have to assume a "traditional" mortgage of, say, 9% would be more appealing than what is currently offered in Iceland. Why have banks, or other private investors, not stepped into the market with alternatives?

    And I've been to Iceland several times. I can't turn around without seeing a tricked out Land Cruiser, which cost a fortune. Is there some kind of tax benefit to maxing out one's expenditures? Maybe you guys are just super rich. I can't wrap my head around it.

  4. I'm not sure I'm seeing the same problem.

    5% inflation over 40 years means that everything -- presumably including housing and wages -- would cost 7 times as much.

    I would expect that indexing would increase the damage if inflation isn't measured properly (and therefore the incentive to measure it badly), or if wages don't keep up with inflation -- but that doesn't seem to be what worries you.

    I would expect indexed mortgages to be riskier (for both borrowers and lenders). Inflation means that a non-indexed mortgage requires a decreasing portion of expected income over time, and that decrease provides a larger margin of safety in case a particular borrower has troubles a few years out. But you don't seem to be worried primarily about risk.

    That same extra safety margin gets treated (on average, when the safety isn't actually needed) as greater disposable income, which provides a bias towards economic growth. But even that should be growth on a smaller base, since the money otherwise spent on "early" amortization is presumably spent on something else at the time.

    Am I missing where you concerns lie, or missing something entirely?