« Weber on Financial Globalization |
Main
| WTO and Services Sector in Emerging Markets »
April 03, 2007
Fat Tails in Emerging Markets
With the greater play of market forces, managing and assessing market risks became crucial in many emerging markets. A major challenge facing banks and other financial institutions in emerging economies is to quantify latent market risk accurately and devise suitable mechanisms to contain market risks to tolerable levels. The quality of risk modeling was affected by the fact that among the models widely used in the global financial markets, many rely on the assumption that returns in financial market are normally distributed. These models tend to generate distorted outcomes in the events of non-normality in financial data. For instance, the popular Variance-Covariance model of Value at Risk ( VaR) relies on the assumption of normality. The assumption, which has proved to be quite useful in the deep and liquid markets of advanced countries may not necessarily hold good in emerging market economies which are prone to periodic adverse and extreme market shocks. In the Indian context, as the Indian financial sector prepares for implementation of Basel II and adoption of advanced risk management and measurement models , it is imperative to look into the issue of non-normality and fat tails in the behaviour of returns in Indian financial sector. Non-normality , or the presence of fat tails in data may lead to spurious measures or risk which may in turn lead to inappropriate policy choices by financial entities. The present paper therefore looks into nomality issue and tail behaviour of government debt market in India.
The tests conducted in the paper( available at (http://sunandoroy.googlepages.com/FatTailsinIndianDebtMarket.pdf) confirms that Indian government securities market is characterized by the presence of fat tails in the financial returns data. Daily volatilities, which are essential requisites of a risk management model demonstrate thick tails, high kurtosis levels and departures from gaussian characteristics. This is coupled with negative skewness, which indicates greater downside risks. This makes use of normal distribution based VaR models prone to significant mis-specification of risks in the portfolio of financial entities. The fat tailed nature is evident in all segments of the yield curve. Although recent deepening of the Government securities market has resulted in orderly behaviour of financial markets in line with policy objectives, fat tails should be given adequate attention while measuring market risks of financial institutions. The backtesting results presented in the paper shows that volatilities in the Indian market are influenced by past volatilities and a GARCH model based on conditional volatility may be useful in risk modeling in the Indian market.
Posted by sunandoroy at April 3, 2007 12:34 PM
