This project is about volatility models comparison, you can expect:

• working on the real market data (from stooq.pl)
• time series data EDA (logarithmic returns transform, realization plots, ACFs, PACFs, descriptive statistics)
• statistical tests (ARCH LM, Jarque-Bera)
• GARCH modelling with prior assumptions about epsilon distribution (hypotheses)
• 4 GARCH models (standard, exponential, threshold, component) with 4 different epsilon distributions each (normal, t-student, skewed t-student, generalized error)
• 3 naive models (random walk, historical average, moving average)
• 9 performance metrics (ME, MAE, RMSE, AMAPE, TIC, MME(U), MME(O), DCP, DCPU)
• quality paper-style report in Polish

• econometrics
• finance
• quality report
• R
• statistics
• time series analysis