I believe that no one would like a new Demirbank incident nowadays. Would you? No, none of you would. The question in the title is exactly what the recent "What shall we do with the government debt securities (GDSs) in Central Bank's (CBRT) balance sheet?" discussion wants to make us ask. The point "Treasury-friendly central bank" policy will drive us to will be a new Demirbank incident. It is also the Demirbank incident what resulted in the explosion of banks' balance sheets in 2001. This path will end in no good. Let us see what the issue is.

The steps taken by Turkey for the recovery from the 2001 crisis were not in fact a bank bailout operation. New rules launched for banks and implemented precisely do not constitute the core of the reform either. Similarly, inflation targeting policy or the measures implemented by the Ministry of Finance is not the core of the issue. In our consideration, the core of the issue is Central Bank's moving away from setting prices in the GDS market. Central Bank's extensive intervention in the GDS market led to the Demirbank incident. What anesthetized the risk perception of market maker banks was the Treasury-friendly policies of the CBRT. It was also such policies that prevented the Treasury and politicians from feeling a tight budget restraint.

So, what is the essence of this policy framework? It was the assurance given implicitly to market makers that short term financing will be pursued at a certain interest rate for ever. And the essence of this assurance is:  market makers will be provided with overnight central bank finance at a certain interest rate regardless from the increase in the borrowing requirement of the Treasury. This way, the CBRT will continuously finance balance sheets of banks through short term resources. What created the Demirbank incident was the provision of long term GDS finance through CBRT reserves. And so, this maturity mismatch eventually became unsustainable resulting in the loss of Demirbank and the half of the system. What happened in 2000 was then followed by the 2001 crisis.

What is the error of this financing model? It restrains the CBRT from adjusting interest rates upwards. This is not the original purpose of short term CBRT finance. Purpose of providing liquidity is not preventing the rise in interest rates when necessary. Overnight CBRT finance figures rising day by day cumulatively before 2001 indicated that interest rate adjustment was prevented.

So, is this issue what is currently discussed? Unfortunately it is in a way. Treasury used CBRT's reserves as well in transferring liquidity to banks in the bailout operation of 2001. With the operation carried out before the introduction of legal arrangements securing the independence of the CBRT, the Bank transferred a long term fund to the Treasury in exchange for GDSs. This way, long term liquidity was injected into the system. According to the plan, the resultant GDS portfolio of 8.9 billion TL was to be amortized in the mid-2010. What does this mean? In May 2010, the value of this GDS portfolio will be paid to the CBRT and the GDSs will come to maturity. What does this imply? The Treasury must save an extra of 8.9 billion TL in a certain account until mid-2010. To save this amount, the Treasury must borrow an extra of 8.9 billion TL from the market. This operation will not change the total GDS stock but it will alter the method of holding of the portfolio. This amount will be hold by market players instead of the CBRT. In that case, in the short term interest rates must be adjusted upwards.

If the CBRT extends the maturity of the GDS portfolio after this point, only the interest rate adjustment will be prevented.  CBRT could have made this payment together saving a fund in Treasury accounts together with the Treasury starting from 2007 and 2008. But it did not. Moreover, this was not explained in time though it was possible. Now, we are at the last minute. And a senseless "guarantee portfolio" discussion is launched. Given the current economic conditions, is it possible that the CBRT does us a favor? Of course it is possible. As a matter of that this is what can happen at this point. As Uğur Gürses also argued, the going argument is just a cover for a Treasury-friendly central bank.

This is what the experience from pre-2001 period told me: Governments must not be made addicted to heroin. Treasury-friendly central bank policy is bad as it is an incentive to the government for fiscal indiscipline.

Turkey must not risk the fundamental acquisition since the 2001 crisis, the principle of setting of interest rate in GDS market by market players. If this principle is annulled, a new Demirbank incident will be faced at best. We had better be cautious.

Governments addicted to heroin always ask for more.

This commentary was published in Referans daily on 5.10.2009