The Hindu Business Line : Structural boost to capital flows

There’s only so much the Government and the Reserve Bank of India can do to stem the rupee’s appreciation against the US dollar in the face of surging capital inflows, the Finance Minister, Mr P. Chidambaram said the other day. The immediate trigger for that comment which almost suggested a hands-off approach to the forex markets was the pronounced pick-up in FII flows after the US Fed cut interest rates on September 18.

Indeed, short of capital controls (Malaysian style), there may be few policy options which could work in the short-term (or even in the long term) to arrest the flood of incoming capital and slow down the rupee’s appreciation. Capital controls may work but will likely have long-term adverse implications.

Even an about-turn by the Federal Reserve to hike rates after, say, six months — which has happened in the past — may not alter the overall environment which appears strongly supportive of increasing levels of private capital flows to emerging markets such as India. The bottomline, therefore, seems to be: more investment capital inflows and therefore higher stock market levels and a stronger rupee for the medium term.

Low correlations provide structural push: More than the superficial push to capital flows by interest rate developments in the US/other advanced markets, the real, structural driver of capital flows is the fact that Indian and global markets are still only loosely or weakly related.

In finance jargon, this may be stated as Indian and global markets having low correlation. And such low correlation between different markets (assets) is precisely the foundation for portfolio diversification.

It is the diversification strategy of international investors which is the fundamental underpinning for capital flows. And neither this strategy nor the low market correlations that drive this strategy seem as if they will change anytime soon. In fact, the correlations seem to have become weaker in the last five years (2002 to 2007) compared to what they were in the previous five years (1997 to 2002) bucket. Therefore, interest rates may go up or down in the US, as it has indeed happened over the last 10 years, but the big picture structurally favours stronger capital flows into India/other emerging markets.

This weakening of the correlation is significant because it has occurred in a time period when the Indian economy has opened up more to the global economic system. Barriers to trade in goods and services have come down further, the financial markets (be it equity or foreign exchange) have been liberalised further to allow greater foreign participation.

More generally, there is a broad consensual movement towards freer markets — goods and financial — in all countries. One would expect that a natural consequence of freer markets would be greater strength in the relationships between national financial markets.

Break-up of the correlation between Nifty and S&P 500: An analysis of the relationship between the Nifty and the (US) S&P 500 over the past 10 years — from July 1997 to August 2007 — shows the weak relationship between the two. The monthly US dollar returns on the two indices over the 10 year period show that the two data series have a correlation of 0.3.

This means that just around 9 per cent of the performance of the Nifty is influenced by what is happening in the S&P 500/the US and more than 90 per cent is determined by economic conditions obtaining specifically in India. Also significantly, as mentioned above, the correlation is even weaker in the five years from July 2002 to August 2007 at just around 0.11.

In this context, it may be interesting to find what keeps the relationship between the Nifty and the US S&P 500 weak? The break-up of the correlation estimate provides the answer to that question and actually throws light on the extent of the divergence between the two indices.

The correlation measure between the two data sets is arrived at by “normalising” what is known as co-variance between the data sets. Co-variance gives a basic estimate of the extent to which the data sets move in tandem (up or down). But it matters little if such joint moves are only a small proportion of the total moves in the two data sets.

It is necessary, therefore, to divide the covariance estimate by the level of variability (risk or standard deviation) in each of the data sets, to find out how closely the data sets move together relative to their volatility. In mathematical terms,

Correlation = (Co-variance / product of Volatility of the two data sets).

It can be seen from the above formula that correlation will be low if either co-variance is low or if the volatilities are high.

The Table shows that volatilities in the Nifty and the S&P in the past 10 years have broadly held stable at 6-7 per cent and 3-4 per cent respectively. Against that, it can be seen that the co-variance in the last 5 years (2002 to 2007) has fallen sharply to 0.000217 from 0.001047 — a fall of around 80 per cent. Therefore, the steep fall in the correlation in the 2002-2007 period to 0.11 from 0.35 in the 1997-2002 time period is explained almost solely by the substantially weakened co-variance between the Nifty and the S&P.

For a given level of co-variance, correlation could still be high if the volatilities are suppressed — either by policy action or otherwise. For instance, a part of the volatility in overall returns will be contributed by volatility in exchange rates.

If exchange rates can be kept stable, it will have a moderating impact on overall volatility. But, if the co-variance itself is weak and is declining, it points to a fundamental divergence in the structural make-up of the markets and consequently their return potential. Such seems to be the case between the Indian and global markets.

More Stories on : Insight | Stock Markets | Forex

Article E-Mail :: Comment :: Syndication :: Printer Friendly Page