Which Rate to cut? Financial Express: 1st November 2008

The CRR and repo rate are two instruments used by the RBI to control growth in money supply. How does one choose between the two and which is the superior tool? This is important in the current scenario where there are phases of low and high liquidity in a scene where interest rates are still sticky.
The CRR impounds/releases cash from banks which reduces/increases their scope to lend thus controlling the money multiplier. The resources impounded earn no interest either in the form of a loan disbursed or a token rate paid by the RBI. This in a way impacts the profitability of banks as interest is being paid without any commensurate income being earned.
In case of the repo (or reverse repo) the RBI changes the interest rate at which banks deal with it. Therefore, to the extent that banks are lending or borrowing from the RBI, the rate change will affect their cost of borrowing which should logically get translated into their interest rate structure. But the efficacy of the rate hike depends on the quantity of money that the banks are borrowing. If banks expect a liquidity crunch then they would revise their basic rates, but if the liquidity situation is expected to be relatively easy, then status quo may prevail. Today a number of banks are offering high rates for 3 years, which means that either there is a mismatch in assets and liabilities or that they would like to lock into such rates for a longer term as the tenure of their assets would be lengthened.
Now, in the current situation, their relative costs could be evaluated. A CRR increase of say 50 bps would mean the impounding of Rs 20,000 crore. As banks would not be receiving any income on these funds, and would be paying an average deposit rate of say 6%, the direct loss would be Rs 1,200 crore, for the entire year. This is intractable unless the CRR is reversed, in which case the gain would be say 200 bps over the deposit rate, or Rs 400 cr. In case of the repo rate, if it is hiked by say 50 bps, then assuming that the banking industry borrows Rs 20,000 cr from the RBI on a daily basis (which never happens since there is an equal spread of borrowings and lending from the RBI throughout the year), then the loss would be Rs 100 crore. Further, banks may choose to rework their interest rates across the board and actually cover up for this higher interest rate. Therefore, they would be better off with a repo rate hike compared with the CRR hike.
The impact on the investment portfolio in terms of losses booked would be the same in both cases as yields on GSecs and bond prices fall when the mark to market valuation is done as interest rates increase.
So, which is the superior tool? CRR is superior when the idea is to curb the lending ability which was the objective when there were large forex inflows which led to monetisation. However, when liquidity is tight, and there is need to restrict growth in credit, then a repo rate is preferable.
An interesting analogy that can be drawn here is with the foreign trade sector, where tariffs are more efficient than quotas. CRR is like a quota while interest rates are like tariffs. Quotas are inefficient as they tamper with the market mechanism, but are effective if the purpose is to physically limit the quantity of funds in the market. Interest rates like tariffs only move the acceptable price schedules laterally. This may not serve the purpose of say increasing interest rates like those on deposits or loans, but then the choice is with the bank to absorb this higher cost.
Therefore, ideally the CRR must be used only when there are excess/shortage funds in the market. The repo rates must be the main tool to control monetary growth. A way out for the RBI is to fix these rates every month based on the state of liquidity. This will address the concern of liquidity changing periodically. As a rule price adjustments should be preferred to quantitative restrictions for the system to be efficient, and the latter should be tinkered with only in case of extreme liquidity situations.