Many interesting comments appeared on my previous blog post, Mythbusting: Current account deficit edition, and I thought it made sense to respond to all of them in this post.
Ambarish: I don't think there has been a sudden rise in rupee trading outside India. It was always there; we weren't seeing it. As Jayanth Varma has emphasised, we used to think the NDF market was in Singapore. But the BIS data on rupee trading shows significant rupee trading at many places worldwide, not just in Singapore. Overall, the picture is roughly one with \$20 billion of onshore trading a day and another \$20 billion of offshore trading, giving an overall market size of \$40 billion a day.
One can think of many good reasons for domestic and overseas economic agents to do INR trading outside India. E.g. foreigners are, presently, not permitted to trade on the onshore currency futures. Given that gross flows across the Indian border, on the BOP, are now at \$1.3 trillion a year, it is not surprising that there is a lot of rupee trading going on. Until big changes to the capital controls take place, I believe there will be an increasing shift of INR trading away from India.
Neeraj: I agree with you that capital controls can potentially change the situation significantly. So one can think at two levels. First, for a given set of capital controls, a central bank can float as in not trading. That gives you a float, but yes, the price that comes out of this is distorted because there are capital controls. As an example, the government can have non-interference in the domestic market for DRAM chips, but the domestic price can itself be distorted through quantitative restrictions or customs duties on DRAM chips. So even though the government is not manipulating the domestic price by directly trading in the market (as it does with foodgrain or currency) the observed price is a distorted one. Then comes the second level where you have full convertibility. Once again, here the central bank could choose to trade in the market or it could choose to not trade in the market. Only when there are no capital controls + no trading by the central bank do you get to the true floating rate and the market's price.
Durga: Modulo the issue raised by Neeraj and touched upon above, I think we're a fairly flexible rate today. If INR trading globally is \$40 billion, then RBI trading of anything less than \$2 billion per day would have a negligible impact on the price. RBI has to either hit the market with very big trades (over \$2 billion a day, i.e. over \$40 billion a month, i.e. Chinese style currency manipulation) or RBI has to sit back and accept the price. Small trades are pointless, and actually make you wonder what the strategy there is.
That said, the rupee is still a small currency. India is a GDP of only \$1.25 trillion and there are a lot of restrictions on cross-border commerce. So there is a long way to go before the INR becomes a serious international currency. It does not, hence, surprise me to see that the spreads on the INR are much worse than those seen for the big international currencies.
However, what I talked about in the blog post -- that when a central bank stops trading on the currency market, the CAD = capital flows -- is not an equilibrium condition. It is an accounting
identity. It requires nothing about market microstructure on the currency market, or about the capital controls, in order to hold. As long as RBI trading on the currency market is zero, CAD will be exactly equal to capital flows.
Finally, Anonymous, you ask: Is there a point where the CAD becomes so big that it becomes dangerous? We should think in two parts.
First, in a place like the US, there has been a lot of concern that the imbalance (= the very large CAD) is too big in the sense that under reasonable assumptions, the US is not going to be able to service all the capital coming into the country. After all, all the equity / debt capital that comes into (say) India today inexorably requires that at future dates, dividends and coupon payments and debt repayment have to happen in dollars, which will require purchases of foreign exchange by residents. In order to service the borrowing of the US today, substantial exports growth will be required, which is unlikely. Hence, when this borrowing of today is to be repaid in the future, a huge dollar depreciation will have to take place.
As long as there is an environment of high growth in exports of goods and services, there is no problem. If, hypothetically, you see a country with a big CAD but you also have a WEAK pace of exports growth then you know that at future dates, there will be pressure on the currency which will give sharp depreciations. Odds are, the financial system will see that and these fears will translate into a depreciation right away! Conversely, if you see capital inflows going into investments which will bolster growth of exports of goods and services, then you feel comfortable that there is no problem.
I believe that's a fair description of the present Indian situation. Over the last 15 years, the gross inflows on the current account into India (which can be roughly interpreted as the total revenues from exporting goods and services) grew by 8.1 times, from \$42 billion in 1994-95 to \$343 billion in 2009-10. This was an average annual growth rate of 15%. This is a huge pace of growth, and gives me confidence that the CAD coming in today will be serviced tomorrow without large currency depreciation. If, hypothetically, you disagree with my optimism about future growth in exports of goods and services, then you would think that this large CAD today increases the odds of INR depreciation in the future, and you would go short the rupee.
For smaller emerging markets, there is a risk of sudden changes in international financing conditions, which is rooted in the lack of information in the hands of foreign investors about the country. Then a large CAD could mean that if something goes wrong and a lot of capital leaves the country, then it could yield a large currency depreciation. I believe this is less and less an issue for the large emerging markets like India, where problems of asymmetric information and lack-of-attention in the global community are not a problem.
Ambarish: I don't think there has been a sudden rise in rupee trading outside India. It was always there; we weren't seeing it. As Jayanth Varma has emphasised, we used to think the NDF market was in Singapore. But the BIS data on rupee trading shows significant rupee trading at many places worldwide, not just in Singapore. Overall, the picture is roughly one with \$20 billion of onshore trading a day and another \$20 billion of offshore trading, giving an overall market size of \$40 billion a day.
One can think of many good reasons for domestic and overseas economic agents to do INR trading outside India. E.g. foreigners are, presently, not permitted to trade on the onshore currency futures. Given that gross flows across the Indian border, on the BOP, are now at \$1.3 trillion a year, it is not surprising that there is a lot of rupee trading going on. Until big changes to the capital controls take place, I believe there will be an increasing shift of INR trading away from India.
Neeraj: I agree with you that capital controls can potentially change the situation significantly. So one can think at two levels. First, for a given set of capital controls, a central bank can float as in not trading. That gives you a float, but yes, the price that comes out of this is distorted because there are capital controls. As an example, the government can have non-interference in the domestic market for DRAM chips, but the domestic price can itself be distorted through quantitative restrictions or customs duties on DRAM chips. So even though the government is not manipulating the domestic price by directly trading in the market (as it does with foodgrain or currency) the observed price is a distorted one. Then comes the second level where you have full convertibility. Once again, here the central bank could choose to trade in the market or it could choose to not trade in the market. Only when there are no capital controls + no trading by the central bank do you get to the true floating rate and the market's price.
Durga: Modulo the issue raised by Neeraj and touched upon above, I think we're a fairly flexible rate today. If INR trading globally is \$40 billion, then RBI trading of anything less than \$2 billion per day would have a negligible impact on the price. RBI has to either hit the market with very big trades (over \$2 billion a day, i.e. over \$40 billion a month, i.e. Chinese style currency manipulation) or RBI has to sit back and accept the price. Small trades are pointless, and actually make you wonder what the strategy there is.
That said, the rupee is still a small currency. India is a GDP of only \$1.25 trillion and there are a lot of restrictions on cross-border commerce. So there is a long way to go before the INR becomes a serious international currency. It does not, hence, surprise me to see that the spreads on the INR are much worse than those seen for the big international currencies.
However, what I talked about in the blog post -- that when a central bank stops trading on the currency market, the CAD = capital flows -- is not an equilibrium condition. It is an accounting
identity. It requires nothing about market microstructure on the currency market, or about the capital controls, in order to hold. As long as RBI trading on the currency market is zero, CAD will be exactly equal to capital flows.
Finally, Anonymous, you ask: Is there a point where the CAD becomes so big that it becomes dangerous? We should think in two parts.
First, in a place like the US, there has been a lot of concern that the imbalance (= the very large CAD) is too big in the sense that under reasonable assumptions, the US is not going to be able to service all the capital coming into the country. After all, all the equity / debt capital that comes into (say) India today inexorably requires that at future dates, dividends and coupon payments and debt repayment have to happen in dollars, which will require purchases of foreign exchange by residents. In order to service the borrowing of the US today, substantial exports growth will be required, which is unlikely. Hence, when this borrowing of today is to be repaid in the future, a huge dollar depreciation will have to take place.
As long as there is an environment of high growth in exports of goods and services, there is no problem. If, hypothetically, you see a country with a big CAD but you also have a WEAK pace of exports growth then you know that at future dates, there will be pressure on the currency which will give sharp depreciations. Odds are, the financial system will see that and these fears will translate into a depreciation right away! Conversely, if you see capital inflows going into investments which will bolster growth of exports of goods and services, then you feel comfortable that there is no problem.
I believe that's a fair description of the present Indian situation. Over the last 15 years, the gross inflows on the current account into India (which can be roughly interpreted as the total revenues from exporting goods and services) grew by 8.1 times, from \$42 billion in 1994-95 to \$343 billion in 2009-10. This was an average annual growth rate of 15%. This is a huge pace of growth, and gives me confidence that the CAD coming in today will be serviced tomorrow without large currency depreciation. If, hypothetically, you disagree with my optimism about future growth in exports of goods and services, then you would think that this large CAD today increases the odds of INR depreciation in the future, and you would go short the rupee.
For smaller emerging markets, there is a risk of sudden changes in international financing conditions, which is rooted in the lack of information in the hands of foreign investors about the country. Then a large CAD could mean that if something goes wrong and a lot of capital leaves the country, then it could yield a large currency depreciation. I believe this is less and less an issue for the large emerging markets like India, where problems of asymmetric information and lack-of-attention in the global community are not a problem.
Dear Ajay,
ReplyDeleteWhile I agree that 40 bn a day is a respectable number for INR trades, the number in itself is not so large to leave RBI as helpless as you made it out to be.
RBI needs to only hit 2 big dealers (say CITI and SBI) with 10 trades of 10 mn a piece in rapid successions when it sees low volumes and the price impact is huge and then the volume charters in the market will do wonders.
Dear Chirag,
ReplyDeleteI don't agree with your characterisation.
There is a good deal of expertise in market manipulation on the equity market and I have had long discussions with people who know that game. To get a sustained impact on the price, you have to be ready to buy 10% to 20% of the overall turnover every day. Anything less than that is just pointless.
E.g. it's incredibly hard to do a market-based manipulation of Nifty given the huge activity on Nifty futures and options. It is also very hard to do this with the top 50 stocks, the members of Nifty, because they are quite active. It is cheaper to manipulate information : hence you see a big focus these days among the bad guys to go after media management (with both lucre and threats).
10 trades of 10M each is $100M. Yes, it might generate a small movement, but that small movement will tend to revert back. If you are after a sustained and economically significant shift in the price (e.g. Rs.1 or Rs.2) then this will just not do it.