Brooks, Robin, and Kenichi Ueda, 2007, User Manual for the Corporate Vulnerability Utility, International Monetary Fund, manuscript.
Cavallo, Eduardo, and Patricio Valenzuela, 2007, “The Determinants of Corporate Risk in Emerging Markets: An Option-Adjusted Spread Analysis,” IMF Working Paper WP/07/228.
Chan-Lau, Jorge, and Toni Gravelle, 2005, “The END: A New Indicator of Financial and Nonfinancial Corporate Sector Vulnerability,” IMF Working Paper WP/05/231.
Chan-Lau, Jorge, and Andre Santos, 2006, “Currency Mismatches and Corporate Default Risk: Modeling, Measurement, and Surveillance Applications,” IMF Working Paper WP/06/269.
Goldman Sachs, 2000, Asia Bank NPLs III: How High, How Structural?, Goldman Sachs Global Equity Research Report (Hong Kong SAR: Goldman Sachs).
Heytens, Paul, and Cem Karacadag, 2001, “An Attempt to Profile the Finances of China's Enterprise Sector,” IMF Working Paper WP/01/182.
Jones, Matthew T., and Meral Karasulu, 2006, “The Korean Crisis: What Did We Know and When Did We Know It? What Stress Test of the Corporate Sector Reveal,” IMF Working Paper WP/06/114.
Prepared by Hiroko Oura and Petia Topalova.
While Indian companies finance the majority of their investment using retained earnings, Oura (2008) finds that they had been increasing their use of external funds (including domestic bank and capital market financing as well as overseas financing) to finance considerably larger investment during the recent period of 9 percent economic growth. As a result, India's corporate sector is increasingly exposed to global financing conditions.
We thank Kenichi Ueda for providing the Matlab code used in the IMF's Corporate Vulnerability Utility (CVU) to estimate default risk indicators. While the CVU also provides BSM default risk indicators, the sample for India is smaller and the aggregated and annual nature of the data prevents us from updating indicators to incorporate the latest equity market information and from relating individual firms' default risks to economic activities. Unlike Moody's KMV, our default risk indicators are theoretical (risk-neutral) indicators, and hence not comparable to actual default frequency. Still, the trends and sensitivity of the estimated default risks capture how the health of the corporate sector evolves over time as well as the relative vulnerabilities of the companies it comprises.
Indeed, cases of corporate bankruptcy in India are extremely rare partly owing to the cumbersome legal framework. Bankruptcy procedures under the Sick Industrial Companies Act, which governs financial reorganization of distressed companies, continue to be time consuming and burdensome, owing to indefinite stays on creditors' claims. Liquidation under the Companies Act is even more complicated and long court delays are common. Since the early 2000s, out-of-court corporate restructuring mechanisms such as the Corporate Debt Restructuring forum and the SARFAESI Act (2002) have facilitated the restructuring of distressed assets. Unfortunately, data on corporate debt restructuring undertaken by banks are not publicly available.
In this section, we focus on the non-financial corporate sector as the stress tests in this section consider the impact of non-financial firms' distress on banks' non-performing loans.
Market capitalization weighted averages are best suited for cross country comparisons. By assigning higher weights to the economically more important companies, the market cap weighted averages focus on systemic risk and mitigate cross country differences in coverage.
First, ICRs do not necessarily account for all the resources that the companies have at their disposal to meet debt servicing obligations. For instance, companies may acquire additional funds from shareholders, take credits from other non-financial companies, draw down reserves, and sell assets. Therefore, as long as poor financing conditions do not persist for too long, the theoretical default may not translate into actual default and a rise in bank NPAs. Second, while credit to corporates accounts for a significant share in total bank credit, banks do lend to other sectors. Similarly, not all firm debt comes from banks. Third, loans that are restructured are not classified as NPAs according to RBI guidelines. Banking sector NPA data augmented for the restructured debt (or disposal of distressed assets) would likely be more closely related to our vulnerability measures; however such data are not publicly available. A study by Goldman Sachs (2000) points out that for Korea, Taiwan Province of China, and Thailand, the reported NPA ratios for the financial system were 18, 5, and 25 percent, respectively in 2000, while their implied NPAs calculated using ICR were 37, 16, and 44 percent, respectively.
Following the CVU, the default barrier includes short-term debt, one half of long-term debt, and interest payments.
Under the normality assumption for the asset returns in BSM, an event measuring 3 standard deviations from the mean is extremely rare, with cumulative density of one percent.
However, it should be noted that the BSM methodology depends on the normality assumption of asset returns. If the true return distribution has fatter tails, the likelihood of severe corporate distress could be larger than what BSM default risk indicators suggest.
At the time of the Asian crisis in 1997, the market capitalization weighted average DtD for emerging Asia was about 7.
This approach could be considered as a sensitivity test similar to the stress tests of the previous section.
This is also a shorthand way to analyze the impact of a sharp depreciation. The BSM default risk analysis is built on the assumption that investors are pricing default risks—including the impact of depreciation—correctly in equity prices. A fuller analysis of the impact of depreciation could include a factor analysis linking equity prices and the exchange rate: an exchange rate shock could be translated in terms of equity valuation and volatility shocks, and then fed into default risk indicator calculations. Furthermore, a volatile exchange rate is likely to require relaxing the normality assumption in the benchmark BSM model, and would require modifying the pricing model as in Chan-Lau and Santos (2006). Having said this, our simple analysis can still give some idea about how the direct impact of rapid depreciation compares to what is implied by actual changes in equity prices and volatility, and whether an exchange rate depreciation is cause for worry or not.
As discussed in the next section, the simple average DtD seems to be the best predictor for macroeconomic performance.
However, our analysis does not include losses owing to derivatives and other contingent liabilities, which could underestimate the overall rupee depreciation impact.
All the analyses maintain the same expected returns from assets as in the baseline. Adjusting expected returns in line with what is implied by the shocks would further increase the impact.
Ex ante, one would expect the market capitalization weighted average to be better correlated with macro variables as it tends to reflect the trends for larger, more economically important companies. However, in India, the share of the formal sector is very small (employing only 10 percent of the labor force). The simple average might do a better job as it can better represent the trend for a large number of small companies that, nonetheless, represent a large share of total economic activity.