Original paper
Abstract
Prior research documents that volatility spreads predict stock returns. If the trading activity of informed investors is an important driver of volatility spreads, then the predictability of stock returns should be more pronounced during major information events. This paper investigates whether the predictability of equity returns by volatility spreads is stronger during earnings announcements. Volatility spreads are measured by the implied volatility differences between pairs of strike price and expiration date matched put and call options and capture price pressures in the option market. During a two-day earnings announcement window, the abnormal returns to the quintile that includes stocks with relatively expensive call options is more than 1.5 percent greater than the abnormal returns to the quintile that includes stocks with relatively expensive put options. This result is robust after measuring volatility spreads in alternative ways and controlling for Â…firm characteristics and lagged equity returns. The degree of announcement return predictability is stronger when volatility spreads are measured using more liquid options, the information environment is more asymmetric, and stock liquidity is low.
Keywords:Â cross-section of equity returns, volatility spreads, equity options, information flow, put-call parity.
Trading rules
- Targeted Stocks: Focus on NYSE, Amex, and Nasdaq companies with actively traded options.
- Calculate volatility spread: End of the day before earnings announcement.
- Volatility spread measure: Weighted difference between implied volatility of matched put and call options, considering average open interest.
- Refine options selection: Discard options with zero or negative open interest and those lacking trading volume.
- Monthly analysis: Group stocks into five categories based on the put-call volatility spread.
- Real trading consideration: Compare stock’s volatility spread to last month’s reporting stocks to determine quintile placement.
- Assess option liquidity: Classify option pairs, the day before the earnings release, into three segments based on the mean bid-ask difference.
- Trading strategy: Go long on stocks with highest implied volatility spreads (measured by low bid-ask spread options) and short on stocks with lowest implied volatility spreads on the pre-announcement day
- Investment duration: Hold for two days, encompassing the announcement day and the subsequent day.
- Portfolio strategy: Maintain equal weightage, rebalance daily, and commit 50% of funds to ensure volatility is manageable.
Python code
Backtrader
import backtrader as bt import pandas as pd
class EarningsAnnouncementStrategy(bt.Strategy):
def __init__(self):
self.earnings_announcements = self.get_earnings_announcements()
def get_earnings_announcements(self):
# Load earnings announcement data here, assuming DataFrame format
return pd.DataFrame()
def get_volatility_spread(self, stock):
# Calculate and return the weighted difference between implied volatility of matched put and call options
pass
def prenext(self):
self.next()
def next(self):
date = self.data.datetime.date(0)
pre_announcement_day = self.earnings_announcements[self.earnings_announcements['date'] == date]
if not pre_announcement_day.empty:
stocks = pre_announcement_day['symbol'].tolist()
volatility_spreads = [self.get_volatility_spread(stock) for stock in stocks]
quintiles = pd.qcut(volatility_spreads, 5, labels=False)
long_stocks = [stocks[i] for i in range(len(stocks)) if quintiles[i] == 4]
short_stocks = [stocks[i] for i in range(len(stocks)) if quintiles[i] == 0]
# Rebalance portfolio
self.rebalance_portfolio(long_stocks, short_stocks)
def rebalance_portfolio(self, long_stocks, short_stocks):
# Get target position size for each stock
position_size = 1.0 / (len(long_stocks) + len(short_stocks))
# Close positions not in the new target stocks
for stock in self.getdatanames():
if stock not in long_stocks + short_stocks:
self.order_target_percent(stock, target=0.0)
# Open or adjust positions for target stocks
for stock in long_stocks:
self.order_target_percent(stock, target=position_size)
for stock in short_stocks:
self.order_target_percent(stock, target=-position_size)
if __name__ == '__main__':
cerebro = bt.Cerebro()
cerebro.addstrategy(EarningsAnnouncementStrategy)
# Add data feeds for stocks in the investment universe
# ...
cerebro.broker.setcash(100000)
cerebro.run()
print(f"Final Portfolio Value: {cerebro.broker.getvalue():.2f}")