Conditional Volatility Models in Financial Markets and Its Application

Sudden and rapid changes in the economy leads to an increase in volatility. The fact that high volatility in financial markets brings along an increase in risk made it necessary to model it. Modeling volatility, which is accepted as a measure of risk, will benefit investors in their attitudes towards risk. The volatility of financial variables such as exchange rates, interest rates, and stock market indices is a measure of how far these variables deviate from their expected values.ARCH-GARCH models, which are used in order to understand the dynamics of financial markets and to predict the changing volatility over time, have been expanded within the framework of some additional needs. Conditional volatility models are used extensively in modeling financial series. In general, ARCH models are models that relate the variance of error terms to the square of previous period error terms. The main feature of these models is that the variance of the error term in period t depends on the square of the error term in the period (t-1). In GARCH models, the error’s variance is not only associated with previous period errors but also with its own variance. The main feature of these models is that they allow varying variance of both autoregressive and moving average components. In order to model the volatility in financial time series, first it is necessary to test whether there is an ARCH effect in the model. If there is no ARCH effect in the model, the OLS estimation method can be used. However, if there is ARCH effect in error terms, the stage of estimating the variance model is started. Since the variance function is not linear, some iterative algorithms are used to maximize the likelihood function. In this study, conditional volatility models that used in modeling the sudden and rapid changes in financial markets are discussed within the framework of other models that have been introduced in recent years. Then, conditional volatility models are applied on real data and obtained results are discussed