The value of the rupee has generated a lot of discussion and analysis in the Press for the last several months; actually since mid-March, when almost to the day, the government of India gave up its 14-year-old policy to keep the real value of the rupee within a narrow band of its 1993-94 value.
It did not matter which party was in power, or who the finance minister was, or who was the governor of the RBI. As constant as the northern star, the real value of the rupee had not budged.
But then came the ides of March 2007. The authorities gave up their hands, said we are not intervening anymore, and let the rupee appreciate by close to 10 per cent. Way outside any of the previous bands; or previous thinking.
Then came the explanations. Two trial balloons have been floated to defend the inexplicable. First, egged on by some very loose thinking, and even looser analysis, the authorities said they had done the deed in order to control inflation.
Never mind that at the time of the action, the six-month inflation rate was already down to less than 4 per cent on a seasonally adjusted annual rate, and less than 2 per cent on a non-adjusted basis.
But this obfuscation hid another problem with the analysis conducted by GoI experts (and some economists in India and at the IMF). In particular, a question left unanswered was that if nominal rupee appreciation was helpful in bringing down the inflation rate, then nominal rupee depreciation should increase the inflation rate, all other things being equal.
But wait a minute. The value of the rupee was Rs 30 in 1993-94 and the inflation rate had averaged above 9 per cent over the preceding five years.
At the time of the great insight in March, the rupee was close to 44, or a depreciation of close to 50 per cent since 1993-94. And the inflation rate for the most recent five years has fluctuated in the narrow range of 4 to 5 per cent.
Once the hollowness of the rupee's inflation fighting argument was exposed, the advocates of a strong rupee (both within and outside government) have latched onto another presumed benefit of letting the rupee appreciate. This is in terms of costs of keeping the rupee afloat.
If there is pressure for the rupee to appreciate, and the authorities don't allow it to, then the dollar has to be bought and the rupee sold. Rupee assets are converted into dollar assets. However, the risk-free interest rate on dollar assets is about 3 percentage points lower than on rupee assets.
There is also talk of the government not having securities to sell in exchange for dollars, and the budget deficit actually going up because of the interest charges associated with the "market stabilisation scheme". What might this impact on the deficit be?
The total amount in the stabilisation scheme today is around Rs 88,000 crore or around $22 billion. Total loss due to interest costs, at a maximum of 3 per cent per annum, is $660 million, which for a trillion dollar GDP, is only 0.66 per cent of GDP.
In rupee terms, this is Rs 2,600 crore. Rather intriguing that what is pocket money for the great spenders in the UPA government is being offered as an excuse to cover up its tracks of a bad rupee policy.
Recall that its flagship programme for its political followers, also called the national Rural Employment Guarantee Scheme, costs more than ten times as much and not only is there no talk for the deficit, but the UPA actually wants more spending for the benefit of its "poor" workers!
Regardless of the size, if the costs of rupee intervention are present, and there are no benefits, then one should let the rupee appreciate. The presumed benefits of an "undervalued" rupee are higher growth of exports, possibly higher employment, and possibly higher GDP growth.
If the last two results are not obtained, then there is precious little argument for intervention in the rupee market.
Casual empiricism would suggest - and empiricism oriented towards our Asian neighbours from western Japan to eastern China - that more than 2 billion people have seen successful miracle growth for the last 50 years, and miracle growth at least in part attributed to a cheaper currency.
All those in the audience who think that China has benefited the most from its cheap labour policy, and at the expense of other poor developing countries with cheap but more expensive labour than China, please raise their hands? Oh, more than 50 per cent - maybe our authorities are wrong and 2 billion oriental Asians are right.
The above calculations of the cost of keeping the rupee low are for one year only; what about the longer term? In a recent book and paper*, I present evidence on the cumulative gains and losses from undervaluation in terms of growth.
Briefly, on a long-term basis (say 6 years, from 2000 to 2006) each 10 per cent initial (2000) undervaluation of a country's exchange rate allows the country to gain extra 0.2 per cent of growth per annum.
|GDP (in $ billions), in 2000||460||1200|
|GDP (in $ billions), in 2006||880||2560|
|Level of Reserves (in $ billions), in 2006||220||1300|
|Level of Reserves (In $ billions), in 2000||41||172|
|Level of currency undervaluation in 2000 (%)||-11||-18|
|Level of currency undervaluation in 2006 (%)||-28||-56|
|Per year average change in undervaluation, 2000-2006, %||-2.8||-6.3|
|Maximum loss from undervaluation 2000 to 2006
Loss due to sterilization (@ 3% per annum of
Accumulated reserves) (For China, a benefit since interest
rates are 1 per cent lower than the US)
|Loss due to possible appreciation of
domestic currency (@3% per annum)
|Total loss from accumulation of reserves||
|Gains from undervaluation (via higher GDP growth)
Gain in GDP, 2000 to 2006, from low initial
|(@ - .02*initial GDP*initial undervaluation) for 6 years
Gain in GDP (2000 to 2006) from change in
undervaluation since initial years
|(@ - .35*initial GDP*avg. change in undervaluation
each year for 6 years
Total gain in GDP from currency undervaluation
for 6 years ($ billion)
|addition of gains fro level and change in undervaluation
Benefit/cost ratio (ratio of total gains/total losses)
|Source: Based on Surjit S Bhalla, Second among Equals: The Middle Class Kingdoms|
of India and China, draft May 2007; Peterson Institute of International Economics
In addition, if the country cheapens its exchange rate adjusted for productivity and inflation (as China and the East Asians have perfected), then there is an additional growth effect.
The table presents some suggestive figures for India and China for the period 2000 to 2006. Note that the cost of holding reserves are on a 2006 base: $200 bn for India and $1,300 bn for China.
This overstates the true cost for six years by a large margin since reserves in 2000 were much lower; $41 bn for India and $172 bn for China. One can change the particular estimates, and lower the impact of exchange rate depreciation, but the net result is still the same - keeping the exchange rate undervalued provides huge benefits to the economy, benefits that are more than six times the cost for India, and more than 11 times the cost for China.
These numbers provide some explanation for the "miracle" of east Asian, and especially, Chinese growth. It is apparent that the elephant has a lot to learn from the tiger.
*See "There are no growth miracles", paper to be presented at the IMF, Aug. 21, 2007