Original paper
Abstract
The shape of the volatility smirk has significant cross-sectional predictive power for future equity returns. Stocks exhibiting the steepest smirks in their traded options underperform stocks with the least pronounced volatility smirks in their options by around 10.9% per year on a risk-adjusted basis. This predictability persists for at least six months, and firms with the steepest volatility smirks are those experiencing the worst earnings shocks in the following quarter. The results are consistent with the notion that informed traders with negative news prefer to trade out-of-the-money put options, and that the equity market is slow in incorporating the information embedded in volatility smirks.
Keywords:Â stock return predictability, option-implied volatility smirks, cross-sectional asset pricing
Trading rules
- Select stocks that have liquid tradable options for your investment focus.
- Compute skewness by analyzing the variance between implied volatilities of OTM put options and ATM call options.
- OTM put: The ratio of strike price to stock price that's nearest to 0.95 (without surpassing it).
- ATM call: The ratio of strike price to stock price that's nearly identical to 1.
- Focus on options that have an expiration period ranging from 10 to 60 days.
- Calculate a weekly SKEW by taking the mean of daily SKEW values over a seven-day span.
- Categorize companies into quintiles drawing from the preceding week's SKEW average (from Tuesday's closing to the next Tuesday's closing).
- Group 1: Contains firms with the lowest skew.
- Group 5: Represents those with the highest skew.
- Adopt a long position in Group 1 and a short position in Group 5 (with portfolios based on the value).
- Adjust portfolios on a weekly basis.
Python code
Backtrader
import backtrader as bt
import numpy as np
class OptionSkewnessStrategy(bt.Strategy):
def __init__(self):
self.liquid_options = self.get_liquid_options()
def get_liquid_options(self):
# Implement your logic to obtain liquid options data
pass
def get_skew(self, stock):
options = self.liquid_options[stock]
otm_puts = [opt for opt in options if opt.type == 'put' and 0.95 < opt.strike / stock.close[0] < 1]
atm_calls = [opt for opt in options if opt.type == 'call' and 0.99 < opt.strike / stock.close[0] < 1.01]
valid_options = [opt for opt in options if 10 <= opt.days_to_expiration <= 60]
otm_puts = sorted(valid_options, key=lambda x: abs(x.strike / stock.close[0] - 0.95))
atm_calls = sorted(valid_options, key=lambda x: abs(x.strike / stock.close[0] - 1))
skew = atm_calls[0].implied_volatility - otm_puts[0].implied_volatility
return skew
def next(self):
if self.datetime.weekday() == 1: # Tuesday
skews = {stock: self.get_skew(stock) for stock in self.getdatanames()}
sorted_skews = sorted(skews.items(), key=lambda x: x[1])
quintile_size = len(sorted_skews) // 5
long_portfolio = sorted_skews[:quintile_size]
short_portfolio = sorted_skews[-quintile_size:]
# Rebalance long positions
long_weight = 1 / len(long_portfolio)
for stock, skew in long_portfolio:
self.order_target_percent(stock, long_weight)
# Rebalance short positions
short_weight = -1 / len(short_portfolio)
for stock, skew in short_portfolio:
self.order_target_percent(stock, short_weight)
cerebro = bt.Cerebro()
# Add data feeds and set initial cash
cerebro.addstrategy(OptionSkewnessStrategy)
cerebro.run()