Estimation of Critical Level of Exchange Rate to Manage Corporate Default and NPAs

In addition to this, for the purpose of forecasting the exchange rate and default rate for a particular year, diffusion process with geometric Brownian motion (GBM) framework was employed. This process may follow either Martingale or Markov process properties. In the former case, the changes in every period are independent from each other while the latter case, on the other hand, views that there is no correlation between stochastic processes and stochastic realizations, but they are not entirely independent from each other. As to Duffie and Glynn (2004) if a distribution has \(\Omega , \Im , \rho\) parameters \(\left\{ {X\left( t \right)} \right\}_{t \ge 0}\) and if the stochastic process is defined as \(\left\{ {\Im \left( t \right)} \right\}_{t \ge 0}\). and if \(\left\{ {X\left( t \right),\Im \left( t \right) } \right\}_{t \ge 0}\), then it can be said that the stochastic process shows martingale properties. And if \(\left\{ {X\left( t \right),\Im \left( t \right) } \right\}_{t \ge 0}\); \(P\left\{ {X \left( t \right) \in B\backslash \Im \left( s \right)} \right\} = P\left\{ {X \left( t \right) \in B\backslash X\left( s \right)} \right\} \quad 0 < s < t, \forall B \in B\left( \Re \right)\), then the mentioned stochastic process is considered to show the properties of Markov process. As to make assumptions, it is important to determine whether the stochastic processes show properties of the martingale or the Markov process. So that the use of stochastic processes in exchange rate forecasting becomes possible.

Accordingly, if a S_t—valued asset follows a stochastic process is as shown below, it is called the geometric Brownian motion:

be the wiener process, µ denotes the percentage of drift, and σ shows the percentage volatility. When the equation is solved, the process is defined as

Using the above specification, exchange rate can be forecasted for a particular period of time. We can find the number of default companies for the same period. We also tried to forecast foreign as well as domestic liabilities and assets of firms to calculate critical exchange rate. However, due to insufficient data with lot of missing values and errors in the data, we found that variance was coming zero for many companies. As a result, the Eq. (7) has become S_t = S₀ exp (ut) and thus the forecasting score for the selected indicators moved in to single path, either decreasing or increasing without giving any clarity about how these indicators are heading to. Therefore, due to this reason, we could not carry forecasting process. The same reasons apply for not employing diffusion method in the analysis.

The Bureau of Industrial and Financial Reconstruction (BIFR)’s criteria to determine sickness in an industrial company are (i) the accumulated losses of the company to be equal to or more than its net worth, i.e., its paid-up capital plus its free reserves (GOI, 1985). We use the criteria of PBIT and PBDIT for defining the ability of the firm to repay the debt to arrive at the firm at the risk of inability to repay.