Model GARCH di Python
Chelsea Yang
Data Science Instructor
GARCH: Generalized AutoRegressive Conditional Heteroskedasticity

Langkah 1: Hitung return sebagai persen perubahan harga $$ return = {\frac{P_1 - P_0}{P_0}} $$
Langkah 2: Hitung rata-rata sampel return $$ mean = \frac {\sum_{i=1}^n {return_i} }n $$
Langkah 3: Hitung simpangan baku sampel $$ volatility = \sqrt\frac {\sum_{i=1}^n {(return_i - mean)}^2} {n-1}= \sqrt {variance}$$
Gunakan metode pandas pct_change():
return_data = price_data.pct_change()
Gunakan metode pandas std():
volatility = return_data.std()
(asumsikan 21 hari bursa per bulan)
$$\sigma_{monthly} = \sqrt{21} * \sigma_d$$
(asumsikan 252 hari bursa per tahun)
$$\sigma_{annual} = \sqrt{252} * \sigma_d$$
Heteroskedastisitas:

Homoskedastisitas vs Heteroskedastisitas

Harga historis VIX:

Model GARCH di Python