Tyranny of numbers

Iran’s multiple exchange rate system which has been in effect since last October has led to much confusion about the new dollar parity for the rial.   Most people in Iran consider the “free” or parallel market rate, which has fluctuated around 30,000 rial per USD, as the new equilibrium exchange rate.  Published Western reports have used a similar number.  For example, an article in Foreign Policy last month put the decline in the value of the rial at 300%, which is not easy to interpret but probably means 3 times what it used to be a year ago (around 10600 rials per $), or about  30,000.  A story in Washington Post yesterday noted that the rial had declined by more than 40 percent relative to its value in August — again, around 30,000 rials per $ since the free market rate in August was around 19,000 rials.  A Reuters report today used the exchange rate of 30,000 to convert rials into USD.  But has rial really fallen by this much?

There are two ways to approach this question. One way is to ask what the rate would be if all, not just part of, Iran’s foreign exchange was sold at the free market.  Surely, it would be less than 30,000.  About a quarter to one-third of Iran’s forex is now selling at 12,260, which is the official rate, and a good chunk (no estimates available) at the rate set at the Exchange Center, about 25,000, and only the rest are traded at 30,000.  Following this logic, last October I estimated that the true rial-dollar parity is closer to 20,000 (which follows from my estimate of the devaluation of about 100%).

A second way is to go by the level of economic activity before and after the devaluation, both evaluated at rials and dollars.  Like the previous method, this one involves some guesswork as well.  Iran has not published details of its GDP since 2008.  The latest summary figures from the Central Bank’s “Current News” series is for the second quarter of 1390 (July 21 Sept 20, 2011) puts the value of the non-oil GDP at 1,265,040 billion rials.  (Non-oil GDP is a good indicator of economic activity and fluctuates less than the total GDP which includes oil.)  In dollar terms this would be about $120 billion (after dividing the rial value by an exchange rate slightly over 10,000).

To arrive at the corresponding number for the quarter that just ended, on December 20, 2012, I assume no real growth in economic activity and an inflation rate of about 40% between 1390-Q2 and 1391-Q3.  Inflation was about 25% annually, quarter on quarter, but during the last quarter (September-December) prices jumped sharply following the October devaluation.   These assumptions put the value of economic activity during Sept 20-Dec. 20, 2012 at 1,771,056 billion rials.

Now, if this number is divided by 30,000 to get its dollar value, non-oil GDP for the quarter would be about $60 billion, which indicates a whopping 50% decline in economic activity in 15 months! Nothing of this order has happened, so the 30,000 number must be wrong.  Unemployment has gone up and industrial production fell, but people still go to work, and agriculture and services, which together account for about 75% of the non-oil GDP, have not collapsed.

What rate would you get if you assumed a flat rate of economic activity for the period? The answer is 1,771,056/120 = 14759 rials.  This is the exchange rate that is consistent with my assumptions of 40% inflation and zero economic growth in the last 15 months.  Even if my numbers are off this way or that, you will be hard pressed to get a parity rate of 30,000 out of these numbers.   Incidentally, the 14759 is very close to the PPP rate that I calculated using my very rough survey of prices in Iran and the US back in November (see my earlier post here).

I think these two methods narrow down the rial-dollar parity rate fairly well, to somewhere between 15,000 to 20,000.  So let’s stop using the free market rate as the exchange rate, and stop saying that the rial has lost 200-300% of its value; the loss is less than 100%.