Risk Forecasting on Leading Balkan Stock Exchanges Through Unconditional Variance Models
Authors:
Stefan
Simeonov
D. Tsenov Academy of Economics – Svishtov
Teodor
Todorov
D. Tsenov Academy of Economics – Svishtov
Daniel
Nikolaev
D. Tsenov Academy of Economics – Svishtov
Pages:
166-
175
Abstract:
Forecasting the variability in the profitability of investment instruments is central to the evaluation and management of market risk. In the first section we present a synthesized overview of current research in the field of forecasting the daily volatility of investment instruments. In the second point we focus on the models of unconditional variance, presenting in chronological order a critical review of the methodology of the models: Simple Volatility; Parkinson’s; Garman and Klass; the modified Garman and Klass model; Rogers and Satchell; Yang and Zhang. In the third point we empirically test the six models on the main stock exchange indices of nine Balkan stock markets: SOFIX, BGBX-40, BELEX-15, MONEX, BIST 100, SBITOP, BIRS, BET and ATG. The most accurate forecasts for the daily return are shown by the Parkinson’s model, in which for eight of the nine stock indices surveyed the results have the smallest percentage error. The modified version of the Garman & Klass model, in which weighting is applied to the price parameters used, occupies the second position in terms of objectivity for each of the studied stock exchange indices. Significant differences in actual profitability are reported by the Simple Volatility model, which is not surprising due to its basic philosophy and the fact that it does not take into account the impact of some components of stock prices.
Keywords:
market risk forecasting; unconditional variance models; Parkinson’s model; Garman & Klass modified; Rogers & Satchell; Yang & Zhang.
Download
271 downloads since 1.3.2025 г.
NA