Python for Finance Cookbook

Leverage effect

For the fifth fact, run the following steps to investigate the existence of the leverage effect.

1. Calculate volatility measures as rolling standard deviations:

df['moving_std_252'] = df[['log_rtn']].rolling(window=252).std()df['moving_std_21'] = df[['log_rtn']].rolling(window=21).std()

2. Plot all the series for comparison:

fig, ax = plt.subplots(3, 1, figsize=(18, 15), 
                       sharex=True)

df.adj_close.plot(ax=ax[0])
ax[0].set(title='MSFT time series',
          ylabel='Stock price ($)')

df.log_rtn.plot(ax=ax[1])
ax[1].set(ylabel='Log returns (%)')

df.moving_std_252.plot(ax=ax[2], color='r', 
                       label='Moving Volatility 252d')
df.moving_std_21.plot(ax=ax[2], color='g', 
                      label='Moving Volatility 21d')
ax[2].set(ylabel='Moving Volatility',
          xlabel ...