Comments and Discussion

Peter Garber: When Brookings asked for a comment on Özatay and Sak's paper, I tried to remember how we were thinking about Turkey going into the crisis. Our analysis of the Turkish situationwas not unusual. We thought about the issues in Turkey in the standard way, considering economic fundamentals, debt dynamic issues, and financial sector problems.

First we looked at the fundamentals, which were seen as weak at the time, with low growth, bad fiscal picture, bad external picture, and high (though declining) inflation. Though we had held this view for a couple of years before the crisis, we thought that fundamentals were moving in the right direction.

After reviewing fundamentals,of course, we had to overlay confidence in the sustainability of the debt, since debt was still growing and the real interest rate was high. So we wrotethe debt dynamic equation, which is the standard method of doing things when analyzing a country's economic condition. We often modify our descriptions of this methodology, calling it the "mindless debt dynamic equation," because one has to plug in several parameters that are assumptions about the growth rate of the economy—inflation rate, real interest rate, and primary surplus—to get whatever result one wants. Only in horrendous cases will this methodology imply an obvious unsustainability, as Morris Goldsteindescribes in this volume. The rest of the time, arbitrary assumptions will imply either debt explosion or convergence, according to one's prior preference.

Together, these methods would provide some indication about whether the situation was going to blow up. In our view, Turkey was not obviously going to blow up with certainty.

Finally, one adds what one knows about the financial sector. As a relatively large foreign player lending into the banking sector, we knew that there was a set of weak banks.

As the crisis unfolded there was sudden information about the magnitude of banking system losses. That tells one more about what will happen to debt dynamics and sustainability.

The official sectors' response to the crisis tells where the various imbalances in the system are going to blow up. For instance, is it going to be strictly an interest rate crisis? It would have been if the central bank had stuck to the limits on the net domestic asset position. That will hit the banking system in a particular way: some banks will win if they are liquid, others will bleed, and others will be cut off in a liquidity crisis.

Or, will it be a foreign exchange crisis? As Turkey's central bank managed the November crisis, it provided domestic credit to the banks. This threatened the foreign exchange policy and meant that the foreign exchange imbalances in the banking system were going to be important.

So, one immediately starts to analyze who is going to get hit under a particular crisis policy and who is not going to get hit.

These considerations require a detailed look at the financial system.I amglad that Özatay and Sak'spresentation basically centered on the financial system. Of course, we knew, and as everyone did, that there were some rarified players in the system, of the sort that can exist only in high inflation. That is, in an environment of high inflation, financial institutions emerge that have no reason for existing at all in a relatively stable, noninflationary environment.The business of these institutions is simply to offer the public ways of avoiding the inflation tax or to speculate on the direction of the inflation.

Anyone with Latin American experience knows about these institutions. When the system starts to move toward a more stable situation—and we were theoretically in a disinflationary policy environment—such institutions are going to be swept from the scene as their profitability evaporates. They exist purely on spreads that will be eliminated in a more stable environment.

Demirbank was exactly one of these institutions, a purely prop-trading, speculative operation. But the bank effectively had a deposit...