The GSec market is probably considered to be the only liquid one where secondary transactions take place. But, actually there are only a handful of securities that are traded and more importantly once the security loses tenure - a 2023 implicit 10-year bond becomes a nine-year bond in 2014 - it loses market interest. But still the 10-year GSec becomes the barometer of interest rates in the system for want of an alternative. How should this security be priced? The implicit yield in the past three months has varied from 8.18% to 9.15% - a differential of around 100 bps. The crossing of the 9% mark has caused commotion in the market, and quite understandably so, as the basic underlying factors appear not to have changed. Can any theory be fitted here?
One way is to look at the state of liquidity in the past three months. The repo auctions are more or less fixed with around Rs 40,000 crore being the daily inflows from the RBI based on the 0.5% NDTL mark. The MSF has become the effective indicator of liquidity. Here too, the amount averaged around rs 40,000 crore in August, rs 66,000 crore in September and Rs 37,000 crore in October. Yet, the rate has remained virtually unchanged at a weighted average of 8.46%, 8.5% and 8.57% respectively. Therefore, liquidity in the system cannot be an explanation.
Another way is to just look at the differential between the repo rate and the yield in different phases. When the repo rate was 7.25%, the average variation of the 10-year GSec was 125 bps above this rate. It came down to 110 bps when the repo rate was raised to 7.5% and decreased to 100 bps at 7.75%. Quite clearly, this theory cannot explain a yield above 9% as 8.75% should be the upper limit.
Bringing in some statistical tools, a regression analysis, linking changes in yields to changes in repo rate on a monthly basis for the past five years, gives a rate of around 8.65% to be the ideal rate when the repo rate is 7.75%, which is close to the crude calculation done earlier. Adding inflation (WPI inflation as CPI data is not available) as another explanatory variable brings the yield to 8.67%. But both these relations explain between 22% and 30% of the variation in the 10-year GSec rate. Intuitively, this also means that the rest is being driven by non-quantifiable factors. What could these factors be?
Sentiment is one factor that can be driving prices of government paper and hence yields. The market believes that the RBI will probably continue increasing rates since inflation appears to be still a hard nut to crack. Also, the RBI has made it clear that inflation control is the primary aim and the CPI, which is higher than the WPI inflation number, is of more consequence. Therefore, expectations of rate hikes can keep yields higher. Significantly, no fresh news on inflation came in when the yields jumped up.
Second, the Fed tapering programme has its role to play in our daily lives. Any good news, like the announcement that more jobs have been created in the USA, leads to the conclusion that the tapering will start soon, even though the unemployment rate has inched up to 7.3%. But this is interpreted as an indicator of further interest rate tightening by the RBI, which, in turn, drives up the rates. But the impact of these measures would be more or less transient for a couple of days and rates should return to equilibrium as these are typically single day sentiments.
Third, movement in exchange rates also impacts the interest rates. While a long-term coefficient of correlation between exchange rates and GSec rates is low at 22%, on a daily basis, the two do move together, especially if the exchange rate moves faster. Therefore a sudden movement of the exchange rate from 61-62/$ to above $63, automatically gets reflected in the interest rates moving up. This sounds plausible, but should again be mean reverting once the exchange rate corrects. A part of the reason for the rupee to fall in November has been the shifting of part of the OMC purchase of dollars to the market from the RBI. This, combined with the prospect of the closing of the swap window for FCNR deposits, has caused a modicum of panic in the money market too.
Fourth, the reaction of banks with these rates works both ways. When banks raise interest rates, which have been done by some of them, the tendency is for the market to follow suit, which can explain partly the increase in rates. If this is sustained, it will tend to feed back into the system and cause banks to reconsider their options.
Fifth, an explanation given often is profit booking where some entities, be it banks or FIIs are selling and thus making a profit. Excess selling leads security prices to fall and rates to rise, which can contribute to the upward movement. This explanation, however, is not borne out from the trading volumes in GSecs, which has been stable at Rs 30,000-50,000 crore in the past 8-10 weeks. In fact, in November, when the yields moved up, the total volumes traded were lower.
The conclusions that may be drawn are that yields can only partly be explained by factors such as RBI rates or inflation and around 70% of variation is caused by a variety of factors - each one stepping in and out in short phases. While it is hard to pin-point which factor is working decisively from outside, the market will have to factor all of them to cover all options.
One way is to look at the state of liquidity in the past three months. The repo auctions are more or less fixed with around Rs 40,000 crore being the daily inflows from the RBI based on the 0.5% NDTL mark. The MSF has become the effective indicator of liquidity. Here too, the amount averaged around rs 40,000 crore in August, rs 66,000 crore in September and Rs 37,000 crore in October. Yet, the rate has remained virtually unchanged at a weighted average of 8.46%, 8.5% and 8.57% respectively. Therefore, liquidity in the system cannot be an explanation.
Another way is to just look at the differential between the repo rate and the yield in different phases. When the repo rate was 7.25%, the average variation of the 10-year GSec was 125 bps above this rate. It came down to 110 bps when the repo rate was raised to 7.5% and decreased to 100 bps at 7.75%. Quite clearly, this theory cannot explain a yield above 9% as 8.75% should be the upper limit.
Bringing in some statistical tools, a regression analysis, linking changes in yields to changes in repo rate on a monthly basis for the past five years, gives a rate of around 8.65% to be the ideal rate when the repo rate is 7.75%, which is close to the crude calculation done earlier. Adding inflation (WPI inflation as CPI data is not available) as another explanatory variable brings the yield to 8.67%. But both these relations explain between 22% and 30% of the variation in the 10-year GSec rate. Intuitively, this also means that the rest is being driven by non-quantifiable factors. What could these factors be?
Sentiment is one factor that can be driving prices of government paper and hence yields. The market believes that the RBI will probably continue increasing rates since inflation appears to be still a hard nut to crack. Also, the RBI has made it clear that inflation control is the primary aim and the CPI, which is higher than the WPI inflation number, is of more consequence. Therefore, expectations of rate hikes can keep yields higher. Significantly, no fresh news on inflation came in when the yields jumped up.
Second, the Fed tapering programme has its role to play in our daily lives. Any good news, like the announcement that more jobs have been created in the USA, leads to the conclusion that the tapering will start soon, even though the unemployment rate has inched up to 7.3%. But this is interpreted as an indicator of further interest rate tightening by the RBI, which, in turn, drives up the rates. But the impact of these measures would be more or less transient for a couple of days and rates should return to equilibrium as these are typically single day sentiments.
Third, movement in exchange rates also impacts the interest rates. While a long-term coefficient of correlation between exchange rates and GSec rates is low at 22%, on a daily basis, the two do move together, especially if the exchange rate moves faster. Therefore a sudden movement of the exchange rate from 61-62/$ to above $63, automatically gets reflected in the interest rates moving up. This sounds plausible, but should again be mean reverting once the exchange rate corrects. A part of the reason for the rupee to fall in November has been the shifting of part of the OMC purchase of dollars to the market from the RBI. This, combined with the prospect of the closing of the swap window for FCNR deposits, has caused a modicum of panic in the money market too.
Fourth, the reaction of banks with these rates works both ways. When banks raise interest rates, which have been done by some of them, the tendency is for the market to follow suit, which can explain partly the increase in rates. If this is sustained, it will tend to feed back into the system and cause banks to reconsider their options.
Fifth, an explanation given often is profit booking where some entities, be it banks or FIIs are selling and thus making a profit. Excess selling leads security prices to fall and rates to rise, which can contribute to the upward movement. This explanation, however, is not borne out from the trading volumes in GSecs, which has been stable at Rs 30,000-50,000 crore in the past 8-10 weeks. In fact, in November, when the yields moved up, the total volumes traded were lower.
The conclusions that may be drawn are that yields can only partly be explained by factors such as RBI rates or inflation and around 70% of variation is caused by a variety of factors - each one stepping in and out in short phases. While it is hard to pin-point which factor is working decisively from outside, the market will have to factor all of them to cover all options.