With a couple of commentaries I opened to discussion what can be done in the context of the monetary policy against the appreciation of the lira. A number of are commentaries on this issue are featured also in many daily newspapers. As you might remember, one of my recommendations was to change the inflation targeting regime after inflation rate is reduced to 5-6 percent. This change was to be made by establishing a framework which considers real exchange rate as well as the difference between the inflation rate and the targeted inflation rate and between the level of production and potential level of production.

We should discuss as an option that real exchange rate is involved in the targets of the monetary policy. The best thing to do for now is to explain what this implies. When doing so, I first want to mention the criticisms about this recommendation of mine. It reads as follows: "Monetary policy cannot have more than one target. The Central Bank has a single tool, which is short term interest rate. You can hit only one bird (inflation target) with one stone (interest rate)."

It sounds perfectly correct at first glance, because it rephrases a fact. After all, if I have one stone and if there are too many birds on the air, I should try to hit just one. Only if two birds are as close as bonded, the stone might hit the second one, too, which would only be a big coincidence. So, the result is, if you have one stone; you should have one goal.

This is correct, as I said. But only at first glance; because if you think on this, you would realize that this argument misses an important point. What I talk about is not that the Central Bank is assigned three duties which are not at all related to each other. The Central Bank will remain to have a single goal which is to maintain price stability. This is what the Central Bank Law says, too. And the Bank will adjust the interest rate policy to fulfill this goal.

The inflation targeting regime is already devised in this line. There is one goal which is to maintain price stability. In this context, you declare an inflation target and try to reach it. Let us say that the expected inflation rate is above the targeted inflation rate. Then, you have to intervene; i.e. increase the interest rate.

I guess we are one the same page up to this point. This is the method followed in all inflation targeting regimes. However, the main feature that distinguishes some inflation targeting regimes implemented throughout the world steps in at this point. This distinguishing feature is about the procedure according to which the interest rate reaction is given. Assume that you need to increase interest rate by four points in order to prevent the inflation rate to diverge from the target. The first alternative is to make the raise over a short period of time. This is called 'rigid inflation targeting' regime. In technical terms, the goal (attention: not the target) function is composed only of the difference between the inflation rate and inflation target.

The second alternative is making the 4-point increase over a long period of time. This is called 'flexible inflation targeting' regime. This time, in technical terms, the goal (attention: not the target) function is composed of two variables: the difference between inflation rate and inflation target and the difference between the production level and potential production level.

So, in short, there is one target and one tool. But there are different goals which will determine the size of the stone you will use to hit the bird. The literature on inflation targeting focuses mainly on this aspect. The most prominent articles in this literature discuss these issues. So to repeat: What is the implication of rigid inflation targeting in terms of the interest rate policy? How will the interest rate policy change if flexible targeting is chosen? What type of fluctuations does this cause in production?

What I recommend is a regime that is addressed in the inflation targeting regime but not discussed widely: If we have to name it; I can call it 'a more flexible inflation targeting' regime. You still have one target and one tool. But this time there are three variables in the function.

The difference with my recommendation and the 'one stone hits only one bird' fact should be apparent to those who are familiar with the inflation targeting literature or who generally deal with control theory. One important phenomenon that maintains the difference is; there exists a correlation between the mentioned three variables. This correlation allows you to start with three goals and set your interest rate policy for a single target. Of course, if the three variables were not correlated at all; the situation would be a three goals-three targets issue.

To get rid of this technical language and leave the poor birds alone, I will explain the issue with a different example on Sunday. The example has an automobile, limited amount of gas, mail(s) to be delivered and different routes to take in it.

This commentary was published in Radikal daily on 22.07.2010