| TAKING STOCK OF THE RUPEE’S RISE |
Click here for Analysis Report Archives
May 2007
SUMMARY
|
The rupee has appreciated by more than 14 percent from its low of 47.04 touched in July 2006. More than half this appreciation has come over the past two months. While RBI has capped the rupee's earlier bouts of appreciation (at around 43.25), this time, pressured by their goal of keeping inflation contained, they stepped back from the market and there is now a growing apprehension that the rupee may be embarking upon another phase of a secular up trend.
To test
this belief, we decided to look at different measures of the long run (intrinsic) value of
a currency to try and assess how far the rupee is from these long run values.
This report explores three important questions in this context
How much is the rupee over/undervalued from its theoretical long run value?
To what extent do these models reflect reality?
What will or can reverse the rupee's current up trend?
We also attempt a subjective medium term forecast, within the context of the overall analytic framework.
How much is the rupee over/undervalued from its theoretical long run value?
We have used three measure of exchange rate valuation to assess the intrinsic (long-run) value of the rupee.
Our own long run model of fair value: As described in our special report of November 2006 "Forecasting the Rupee", we developed a model that explained movements in USD/INR since 2001. We found the two long-run variables that most effectively explained the movements in the rupee were (1) the 12-month sum of FII inflows into the capital market, and (2), the ratio of the 3-month rolling sum of oil imports to total imports. We updated the model with data to April 2007 and found that the rupee is currently overvalued by 10 percent as compared to the model's long-run forecast.
The model, which had seemed quite robust at the time, also recognized the impact of short-run (or cyclical) forces, which could create variances from the long-run value. The maximum variance that we saw over the period 2001 to 2006 was 3-4%.
The fact that the rupee has shot out of the model's variance range so dramatically suggests that the model is no longer working. Indeed, we note that RBI intervention had been a key component of the market environment for the entire period of study in developing the model; now that RBI is out of the market (at least currently), it would seem reasonable that we need to look at a different model to explain the rupee's movements.RBI's REER: RBI has often referred to the real effective exchange rate of the rupee, based on six-country trade weights; it has been widely assumed that RBI maintains the nominal value of the currency within a 5 percent band about this REER. With RBI having stopped intervening, the rupee is now overvalued by more than 10 percent compared to the REER.
Clearly, here, too, there is a case for reassessing the existing model, which, in any case, has two structural flaws (which, to be fair, infect most economic models) - the selection of a base year and the assumption that inflation and exchange rate differences are the only drivers of trade competitiveness.
Purchasing Power Parity: PPP estimates the long-run equilibrium exchange rate of two currencies by equating each currency's purchasing power. It is based on the law of one price, which states that in an efficient market, identical goods must have only one price. Absolute measures of PPP are readily available from the IMF's World Economic Outlook, which show the PPP for the rupee is currently around INR 10 per USD. Deviations from the PPP are long lived and currencies may take many years to - and may never - converge to their PPP.
We have constructed our own PPP model, using the approach of "relative PPP", which is described in greater detail in Box 1. This model suggests that the rupee is currently overvalued by 15% from its PPP.
| Box 1: Fair value measures of the rupee based on "relative
PPP" The concept of PPP is based on the notion that over time, prices of traded goods would equalize across borders. Thus, the nominal exchange rate should move over time to equalize relative prices across countries for a basket of goods and services. Countries that experience high levels of inflation will see their currencies depreciate in order for PPP to be restored over time. Thus, the higher inflation in India relative to the US would push the PPP higher, as, indeed, has been seen - see graph. Of course, the PPP approach implicitly makes a number of assumptions: it assumes that the basket of goods and services is tradable, that there are no barriers to trade, that there are negligible transaction costs, and that the basket is homogenous across countries. Calculation of PPP also requires two explicit assumptions: first, the choice of the domestic and foreign price levels used and second, the choice of base year. We have constructed a relative PPP model as follows: Starting in April 1994, we used the existing exchange rate and multiplied it by the ratio of the US price index to the Indian price index to come up with a unit-less number; we created this number for each month till December 2005, and then calculated the average. The monthly PPP values were then generated by multiplying this baseline by the monthly ratio of the Indian price index to the US price index. The chart below shows how the actual values of USD/INR for each month compared with these PPP values. Empirically, we found that USD/INR remained mostly within a band of 7 percent about the PPP over the last 13 years. From 1998 to 2002, it traded close to the top of this band; it reversed its course in 2002, and moved downwards swiftly and, from 2004, has traded at the lower end of the 7 percent band. It broke through the band in mid-2006 and, according to this model, the rupee is now overvalued by more than 15 percent against the dollar. |
While all three theoretical approaches indicate that the rupee is currently overvalued by 10-15% from its long run value, the rest of this paper attempts to assess whether this is, indeed, reasonable, or whether this conclusion is a construct resulting from constraints of the models and, of course, the continuing deregulation of the Indian market.
To what extent do these models reflect reality?
Our current analysis suggests that, contrary to the conclusions of the theoretical models, the rupee's recent strength may merely be taking it back to its true fair value. The following explains our analysis:
Broad dollar weakness: We studied the correlation between USD/INR and the US Fed's broad dollar index, which is the weighted average of 7 major currencies and 19 emerging market currencies (including, incidentally the Indian rupee, which currently has a weight of about 1.25%). The correlation has been very strong - for the entire period since 1995, it has been over 88 percent - suggesting that the dollar's strength or weakness overseas is a strong indicator of USD/INR. The chart shows that, since the dollar index started falling in 2002, the correlation began to weaken.
This is
doubtless because RBI began to intervene to protect export competitiveness. Intervention
was stronger since 2004, as reflected in a further weakening of the correlation.
We also note that this period of the past few years has been characterized by increasing openness of Indian markets and a dramatic rise in global investment interest in India. This chart then suggests that current burst of rupee strength, as RBI stepped away from the market, is simply be a catch-up to global dollar weakness, and, indeed, that it (rupee strength) has some further way to go.
RBI absence from the market: In the model we had developed last year, we had recognized that there are several short run factors that could keep the currency away from its intrinsic value, and that these short run factors could be in play for a long time. One of the key short run factors we had identified was RBI intervention. However, it does appear that we had not been able to fully assess the significance of its impact till, as recently, the factor was taken out of play.
As argued above - and, as is intuitively obvious - the huge amount of RBI intervention over the past several years has prevented the rupee from strengthening fully in response to the dollar's global weakness, and that the resulting rupee value was much stronger than its "fair value".
From 2002, and even more so since 2004, RBI intervened aggressively in the forex market buying dollars to stem rupee appreciation. During the 3 years to March 2007, India's foreign currency reserves rose by about USD 86 bn, indicating the scale of RBI dollar buying. This intervention released Rs. 364,000 cr into the market. In addition to raising the CRR by a total of 1.5% (impounding about Rs. 65,000 cr), RBI did attempt to sterilize the intervention, by selling bonds to suck out the rupee liquidity generated. However, given that the total amount of bonds available to RBI under the market stabilization scheme (MSS) was a mere 80,000 cr (which was increased up to 110,000 cr last month), it is hardly surprising that money supply has surged way beyond target and that inflation, too, has crossed the 5.5% limit.
With inflation control still its primary objective, RBI has had little choice but to step back from the market, which they did towards the end of March 2007, which resulted in the rupee's sudden surge.
This change in RBI's approach has resulted in a huge hue and cry, as many exporters, particularly those from the SME sector who have little expertise in currency management, have found their profits being squeezed, or, indeed, vanishing. Trade figures confirm that export growth has slowed to single digits each month since December 2006. We note, however, that the current account deficit remains comfortable at under 2% of GDP.
Thus, it
appears that the reality of India's positioning in the globalization cycle has come home
to roost. Huge capital inflows, more than partly driven by the weak dollar overseas, has
pushed RBI's command and control approach to the wall, and, indeed, brought the rupee on
to its appropriately appreciating path against the weakening dollar.
[As an aside, the current hysteria as a result of the rupee's sudden surge, which has
certainly left most exporters in pain, may well have been avoided if RBI had let the rupee
steadily respond to dollar weakness over the years. This would have enabled companies to
adjust their costings and their expectations on a continuous basis, rather than having to
deal with it in one fell swoop.]
What will or can reverse the current trend?
Let us take each of the two factors in turn:
Broad dollar weakness: It would
seem clear that only a turnaround in the dollar's fortunes globally could prevent further
rupee strength. Indeed, given the limitations on RBI's activity [see below], any further
dollar weakness overseas could lead to a yet stronger rupee.
Thus, more than ever, we need to have a sound prognosis on the dollar's future. The dollar
is clearly very weak, but we note that the index (on a monthly average basis) is still
some ways from its all-time low. In any event, we believe that current dollar weakness is
driven by different forces than those that drove its weakness from 2002 to 2004. At that
time, it was the fear that capital inflows would be inadequate to finance the huge US
current account deficit that pushed the dollar lower. However, over the past year or so,
the US current account deficit has started to show signs of improvement (possibly as a
result of dollar depreciation) and capital flows in the US remain robust. Indeed, these
parameters hardly find a mention in any analyses these days.
Nonetheless, the dollar remains on the ropes and it is more likely that the current dollar weakness is being driven by more structural factors like the rotation of growth away from the US towards Europe, Japan and the emerging markets. Till 2005, the USD 13 trillion US economy provided nearly 50% of the incremental growth in the world economy; by 2006, it had fallen below 40% and in 2007 it is expected fall further. While it is unclear whether the current slowdown in the US economic growth is "cyclical" or "structural", the fact remains that there is increasingly a new game in town for growth and investment.
Indeed, with the Euro-zone continuing to surprise on the up side, we could be seeing the start of a significant diversification into Euro-denominated assets, which could render a step-change blow to the US dollar. There is considerable anecdotal evidence of oil invoicing in Euros and of several emerging market central banks diversifying their reserves holdings our of dollars. Indeed, China, which is the largest player in this game, is believed to have set up vehicles to pull all the fresh reserves - which are in the range of USD 200 bn a year - into new investment vehicles (rather like the investment vehicles of the Government of Singapore) both to improve returns and to minimize the risk on their already hugely overweight dollar portfolios. There is already evidence of this trend towards diversification with the IMF reporting that the share of US dollar in total world allocated reserves has fallen from 72 percent in 2001 to 64 percent more recently. Total allocated reserves are USD 3.3 trillion; incidentally, there is USD 1.7 trillion of reserves, which are "unallocated" - i.e., the IMF does not know their currency composition.
If this
structural shift does indeed pan out, it looks like dollar weakness, and hence rupee
strength, could be here to stay in the medium to long term.
Having said that, there will, of course, be cyclical forces that could interrupt this trend from time to time. One of the most important of these is the interest rate differential between the US and its trading partners. The chart confirms that at this time the correlation between the dollar index and partner country interest rate differentials is very high, indicating that it is currently an important force in the market.
With the weakening US economy, the US Fed, while vigilant on inflation, may not be able to raise rates too soon. In fact, the betting is that the next move in US interest rates will be down, whereas in Europe, Japan and the emerging markets interest rates are expected to either hold steady or rise.
Thus, interest differentials point to short-term dollar weakness, which, of course, suggests continued short-term upward pressure on the rupee.
| Box 2: Monetary Condition Index (MCI) The intention of a MCI, as the name suggests, is to provide information about the stance of monetary policy. While not all central banks publish a MCI, there has been a fair amount of work done in this area by the IMF, the OECD and, of course, several global investment banks. Monetary policy influences inflation primarily through two channels: interest rates and exchange rates. A rise in interest rates or exchange rates would, in general, cause the economy to slow down and lower inflationary pressures. Similarly, a fall in interest rates or a decline in exchange rates would tend to stimulate the economy and could lead to higher inflationary pressures. The monetary conditions index MCI t, is defined as the weighted sum of changes in the exchange rate (E) and in the interest rate (R), from their levels in a chosen base year MCI t = W1(Et-E0) + W2(Rt-R0) where W1 and W2 are weights for the interest rates and exchange rates respectively. The MCI can be constructed in terms of the effect of the interest rates and exchange rates changes on either "aggregate demand" or on "prices". Since there is no meaningful measure of aggregate demand in the Indian economy and also because RBI is clearly focused sharply on price stability at this time, we have constructed our MCI to focus on the impact of these variables on domestic prices. We used one-year interest rates (gilts), WPI (as a measure of inflation) and RBI's REER (6 country trade weights) for our analysis. Recognizing that
monetary transmission is very low in India as compared to developed economies, we used
coefficients of 2.5 and 1 for changes in real interest rates and exchange rates,
respectively. The ratio of coefficients in developed economies varies from 4 (in Europe,
as estimated by the OECD in 1996) to 10 (for the US during the same assessment). Our
assumptions translate to mean that our domestic markets today are somewhat less liquid
than European markets were ten years ago. |
Of course, the nature of the market is to prove most analysts wrong, so we would expect that at some time in the future - probably when the fewest people expect it - there will be a sudden turnaround in the dollar's fortunes, the reasons for which will, after the fact, become obvious to all.
All in all, we would have to say that the medium term prognosis for the dollar is not good. Which, in turn, confirms our belief in continued rupee strength over the next several years.
RBI's stance: There may, however, be certain local extenuating factors that could provide some respite from the strong rupee in the near term.
First of all, it is becoming clear that monetary conditions in India are already becoming restrictive. We have developed a measure of "monetary conditions index" for India (Box 2), which suggests that the stance of monetary policy in India is currently "restrictive", largely because of the recent sharp appreciation of the exchange rate.
This suggests that we may begin to see some slowing growth numbers by the second half of the year - as noted above, exports growth has already slowed quite sharply. This, together with the turning base effect, would allow inflation to cool a bit, which may enable RBI to take its fist from the monetary noose and resume buying dollars to boost competitiveness. We may already have seen some evidence of this earlier this month.
However, we do not believe that RBI will return to their blind dollar buying - the inflationary spurt and the fear that it may once again get out of control will keep them deterred. In any event, it is clear that faster deregulation is a more sensible - and certainly more sustainable - answer to the combination of India's increasing attractiveness and the dollar's continuing weakness in global markets.
Indeed, unless the dollar does surprise and turn stronger, we believe that monetary policy will remain restrictive for most of the remaining part of the year, since there is what appears to be an inexhaustible pool of savings targeting India. Note the huge (by our standards) surge of FII inflow into Indian debt markets in the last year, directly a result of the improvement in the yield differential in favor of India.
It is also significant that RBI has de-emphasized interest rates (repo/ reverse repo) as the policy instrument in favor of tightening liquidity measures (using CRR) - clearly, they recognize that the narrowness of Indian financial markets makes interest rate increases a highly ineffective method of monetary transmission. This leads to the (hopeful) belief that RBI will move much more swiftly than is its wont to further deregulate financial markets - more open markets would enable the current level of monetary tightness at lower (and, possibly, much lower) interest rates.
While it will be a tight tightropewalk, we believe that RBI will be spared having to go for the jugular - i.e., capital controls. Given that our global credibility has been very hard won, we believe that even if the dollar decline push comes to a capital inflow shove, RBI will let the rupee take the brunt.
| Box 3: Market Stability Index We have developed a Market Stability Index, which provides a measure of the effectiveness of the market for risk management. Managing market risk is really about making judgments about market volatility, which is why options, which are priced based on market volatility, are the ideal hedging instruments If the volatility is largely stable - irrespective of how high it is - it is relatively easy to manage risk; however, if the volatility itself is unstable and jumps around, it is much more difficult to manage risk. This second condition often prevails in markets where volatility is kept artificially low by regulation and/or market intervention. We have been tracking this index for the rupee (and a range of other currencies) for some time now, and have found, unsurprisingly, that the index (at 78%) is at its highest level since April 2004 - surprise, surprise! In fact, it has reached levels higher than those reached by the Thai baht back in December 2006, when the Thai government, terrorized by the appreciation of its currency, imposed some fairly draconian capital controls (which were reversed in a couple of days).
The index confirms what we all know - that our market is currently in a fairly delicate situation. We also note, incidentally, that the index for the Euro is also running a bit higher than average, suggesting that some part of our market discomfiture may have to do with global conditions. The good news is that this immediate term of trial may be passing - the indicator has topped out (on May 9) suggesting that things may be cooling down. In any event, we do not believe that capital controls - other than the soft ones, like not increasing the ECB cap, the technically incorrect shifting of preference capital to the debt bucket, etc. - would be on the cards this time around. However, unless deregulation accelerates, the next episode of instability will most likely be worse. Another useful aspect of this indicator is that it can provide breakout signals for trading. On each of the 13 times the index has peaked since January 2006, the rupee has moved by an average of 2.2% over the following two months. |
Subjective forecasts:
It is well known that only fools make forecasts that provide both a rate and a time horizon. Nonetheless, emboldened by our 2003 forecast - that the rupee will reach 38 in 5 years - we will take another plunge.
Over the near term, it is beginning to feel like the rupee's rise has reached a peak, even if temporary. The fact is that since it broke above 42 to the dollar on April 22, the rupee has seemed much less a wild bull - it has steadied about a mean of 41.15, and the volatility, which has been soaring, also appears to be stabilizing. [We note that our peak to average volatility indicator, which had hit levels not seen in several years, indicating intense market instability, also appears to be moderating (Box 3).] Further, with more and more companies buying structured products to hedge their heavily out-of-the money exports, we may see a slowdown in dollar selling from this source.
Capital inflows, too, appear to be moderating, both as a result of high valuations in the equity markets and the increasing uncertainty over restrictions that may be imposed on FDI and ECBs.
Further the dollar is showing some signs of having reached a (temporary) bottom. Gulping strongly, we forecast a near term range of 40.50 to 41.50.
Over the medium term, however, things are likely to get volatile again. If the dollar corrects seriously overseas - the likelihood of which increases the longer that it doesn't - we could see the rupee dip sharply, certainly breaking 42, and, possibly 43. On the other hand, if the dollar does go into another downward spiral, we would see the rupee under pressure to appreciate again. Assuming inflation is under a bit better control - certainly likely given the base effect - RBI may be able to man the barricades for some time, unless, of course, the dollar's weakness was intense.
We would forecast a very wide range of 41.00 to 43.50 for the medium term.
And finally, we stand by our long term forecast that the rupee will hit 38 some time during 2008.
