Original paper
Abstract
This study examines the efficiency of VIX option trading strategies that exploit the VIX futures roll and the often substantial VIX futures volatility premiums from January 2007 through March 2014. The study first assesses the related issue of whether VIX options typically are overpriced by examining long VIX option delta-hedged returns and demonstrates that average losses on front contract calls and puts over 5-business day horizons either are not statistically significant or are economically small. In light of the evidence that VIX option buyers on average do not overpay at all or by much for the limited risk associated with VIX options, the study then turns to whether long VIX option positions can be used to exploit the well-documented tendencies of VIX futures to rise and fall when the VIX futures curve is in backwardation and in contango, respectively, as well as the tendency of VIX futures to build in large ex-ante volatility premiums. The results demonstrate that these defined-risk strategies are highly profitable and offer attractive risk-reward tradeoffs. Moreover, the systematic tendencies of VIX futures have far more power for predicting attractive VIX option returns than the ex-ante volatility premiums built into VIX options. The study also shows that long VIX option strategies importantly benefit from a strong tailwind that owes to the tendency of VIX option implied volatilities to rise with increases in the actual volatilities of underlying VIX futures contracts, as VIX futures move toward settlement and their volatilities rise to the typically higher volatility of the VIX.
Keywords:Â VIX Options, Volatility Premiums, Trading Strategies
Trading rules
- Target assets involve VIX Index options.
- Trading decisions hinge on the daily roll, computed by the differential between front VIX futures and the VIX, normalized by business days remaining to the VIX futures contract settlement.
- For a situation of backwardation, the daily roll should fall below -0.10 VIX futures units; for contango, it should exceed +0.10 units.
- When the VIX futures curve indicates backwardation, buy the nearest to maturity at-the-money VIX calls; when indicating contango, buy the closest to maturity at-the-money VIX puts.
- Trades are entered and exited at the mid-point of closing bid-ask quotes, and new trades are entered as soon as they are triggered.
- Trades have a holding period of 5 working days.
- Dedicate 5% of the trading portfolio for every transaction.
- Backtested results from the academic paper are scaled accordingly with a multiplication factor of 1/20.
Python code
Backtrader
# Import necessary libraries
import backtrader as bt
# Define the options on VIX index as investment universe
class VIXOptions(bt.Strategy):
params = dict(
rolldays=bt.num2date(7),
atmf=0.5,
contract='VX1')
def __init__(self):
self.rollover = btind.RollingFuture(
self.data1, period=self.p.rolldays, calendar=btind.BusinessDay,
name='FC')
self.roll = self.rollover - self.data0.close
def next(self):
business_days = len(btind.businessdays(
self.datas[self.p.contract].datetime.datetime(0).date(),
self.datas[self.p.contract].datetime.date(0)
))
# Calculate daily roll
daily_roll = self.roll[0] / business_days
# Buy at-the-money VIX calls in backwardation and VIX puts in contango
if daily_roll < -0.10 and self.data0.close[0] > self.p.atmf:
order = self.buy(size=0.05 * self.broker.getvalue() / self.data1.close[0],
exectype=bt.Order.Market)
elif daily_roll > 0.10 and self.data0.close[0] < self.p.atmf:
order = self.sell(size=0.05 * self.broker.getvalue() / self.data1.close[0],
exectype=bt.Order.Market)
# Exit trades after 5 business days
for order in self.broker.open_orders:
if (self.datas[order.params.contract[0]] <= order.params.expiry.datetime(0)) and order.params.expirycheck:
self.cancel(order)
for p in self.broker.positions:
if p.data == self.data0 and len(btind.businessdays(p.opened.dt.date(), self.datetime.date(0))) > 5:
self.close(p)