Daily Commentary - Posted on Sunday, November 23, 2014, 7:10 PM GMT +1

17 Comments


Nov Sunday 23

Trading The S&P 500 Index (via Implied vs. Historical Volatility)

A couple of weeks ago I started a series of postings, all dealing with trading volatility ETNs / ETFs like XIV® (VelocityShares Daily Inverse VIX Short-Term ETN) and VXX® (iPath® S&P 500 VIX Short-Term Futures™ ETN) and respective trading strategies. One of those strategies was DDN’s VRP Strategy (Double-Digit Numerics , Volatility Risk Premium) due to its exceptional performance – at least until December 2012 – and its compelling approach (from the paper Easy Volatility Investing from Double-Digit Numerics).

Since then there not only has been a discussion (comments section of the blogs mentioned later on) about data mining and curve-fitting (among others), which has been adressed and investigated by Ilya Kipnis on his (highly recommended) blog QuantStrat TradeR and the guy behind Volatility Made Simple (likewise highly recommended), but about the shrinking premium of VIX® front month futures (1st and 2nd month) vs. the VIX® index and the (future) doubtful tradability and potential trashiness of the Volatility Risk Premium Strategy as well (please note: DDN’s VRP Strategy (original) is currently undergoing a severe drawdown).

At least with respect to the time frame under review, I absolutely disagree.

Trading XIV® and VXX® not only means trading volatility as an underlying (VIX® front month futures, 1st and 2nd month), but trading a higly volatility asset itself, leverage (the ratio of the average daily movement of the S&P 500® index and VIX® front month futures (1st and 2nd month) is at/around 4, between S&P 500® index and VIX® index at/around 5.5) and time (risk) premium/time decay (regularly selling volatility/buying XIV® in order to harvest the premium associated). So how would VRP systems have performed and results be different if the S&P 500® index had been traded (instead of XIV® and VXX®) using VRP system signals for timing purposes, and therefore which part of VRP’s strategy performance(s) can be accounted to the strategies quality of forecast (S&P 500® index movements), and which part to buying/selling time (risk) premium/time decay ?

“How would VRP System results be different if SP500 L/S were to be traded using the VXX and XIV signals? (For VXX, go Short SP500, and for XIV go Long SP500).”

“Future returns will depend on the shrinkage/expansion of this premium. If the cake is gone ( or the pie much smaller so to speak ) , it’s gone ! No clever optimization will change this fact.” “The VRP has been zero or negative most of the year. … So VRP strategy returns should be zero.”

But first of all the original and revised Volatility Risk Premium (VRP) Strategy rules (always market on close):

  • Long S&P 500® : 5-day average of [VIX® | VXMT® – (10 | 2-day historical volatility of SPY | S&P 500 * 100)] > 0
    (Please note: Before 2008, VIX® instead of VXMT® is used)
  • Short S&P 500® : 5-day average of [VIX® | VXMT® – (10 | 2-day historical volatility of SPY | S&P 500 * 100)] < 0
  • Hold until a change in position.

Image I shows the respective equity curves (always trading the S&P 500® index, no leverage used, and always being ‘all in‘), DDN’s Volatility Risk Premium Strategy (original strategy) (red line, using a 10-day historical volatility), the VXMT® index version with a 2-day historical volatility (blue line), my own strategy, see PORTFOLIO & TRACK RECORD (black line), and a buy&hold (S&P 500® Index) approach (grey line). Please note: The CBOE provides VXMT® historical data back to 1/8/2008. Before 2008, VIX® has been used instead.

 

Image I – Trading the S&P 500® Index – Total Equity Curve(s)
(03/25/2004 – present)
(slippage, fees and transaction costs are assumed to total 0.05% per trade, 0,1% per round-trip)

From 3/25/2004 to 09/2008, all strategies were more or less dead even, while out-/underperformance startet (continuously accelerating) at the same time when the S&P 500® went into free fall during the financial crisis. Although this time DDN’s VRP Strategy trading the S&P 500® does not undergo a severe drawdown like it currently does trading VXX® and XIV® , it startet to flatten out one and a half year ago while the S&P 500® as well as all other counterpart strategies (trading the S&P 500®) went straight up (reason for the flat performance: a subpar 10-day historical volatility as already mentioned in my previous postings).

Image II shows the respective statistics for those strategies mentioned above.

 (click on image to enlarge)

Image II – Trading the S&P 500® Index via Implied vs. Historical Volatility
(03/25/2004 – present)

 

Next step/question: Having been stripped off it’s leverage and time (risk) premium (just up and down index movements), how would VRP systems have performed and results be different if a leveraged S&P 500® index ETF/ETN had been traded (instead of XIV® and VXX®) using VRP system signals for timing purposes, in order to work out (approximation) which part of VRP’s strategy performance(s) can be accounted solely to buying/selling time (risk) premium/time decay ?

For simplification purposes I calculated the respective leverage as the ratio of the average daily movements of the S&P 500® index vs. VIX® front month futures (constant maturity, 1st and 2nd month futures), which is at/around 4 . And for demonstration purposes, a combination of the Volatility Risk Premium Strategy (trading the S&P 500® index) – VXMT® index version with a 2-day historical volatility and the S&P 500 RSI(2-day) indicator for indentifying oversold/overbought conditions (blue line) has been added as well in order to show how performance could have been improved (boosted) by adding a well known technical indicator.

The Volatility Risk Premium (VRP) Strategy rules in combination with the S&P 500 RSI(2-day) indicator are as follows:

  • Long S&P 500® : [ 5-day average of [VXMT® – (2-day historical volatility of S&P 500 * 100)] > 0 OR S&P 500®  RSI(2-day) < 1 ]
    AND S&P 500®  RSI(2-day) < 95
    (Please note: Before 2008, VIX® instead of VXMT® is used)
  • Short S&P 500® : otherwise …
  • Hold until a change in position.

Image III shows the respective equity curves (strategies trading the S&P 500® index – blue line and red line – have been leveraged by factor 4), the Volatility Risk Promium Strategy (trading VXX® and XIV®) – VXMT® index version with a 2-day historical volatility – (black line), the Volatility Risk Premium Strategy (trading the S&P 500® index) – VXMT® index version with a 2-day historical volatility in combination with the S&P 500 RSI(2) indicator – (blue line), the Volatility Risk Premium Strategy (trading the S&P 500® index) – VXMT® index version with a 2-day historical volatility – (red line), and a buy&hold (XIV®) approach (grey line).

 

Image III – Volatility Risk Premium Strategy (Market Timing)
(03/25/2004 – present)
(slippage, fees and transaction costs are assumed to total 0.05% per trade, 0,1% per round-trip)

 

Image IV shows the respective statistics for those strategies mentioned above.

 (click on image to enlarge)

Image IV – Volatility Risk Premium Strategy (Market Timing, Leverage Factor 4)
(03/25/2004 – present)

From my perspective even with a shrinkage of time/risk premium, the Volatility Risk Premium Strategy rules can be useful (for market timing purposes) and highly profitable trading the S&P 500® index instead of trading XIV® and VXX® due to its (superb) quality of forecasting major market index movements. Although I’d still prefer trading XIV® and VXX® , for example adding the S&P 500® RSI(2-day) shows that trading the S&P 500® performance figures could be (further) improved not necessarily by adding any additional rule(s) and/or parameter(s), but by playing around with the existing set of parameters as well (x-day historical volatility, y-day moving average, the threshold/cutoff, …).

 

__________________

Have a profitable week,

Frank


Disclosure: I’am long/short VXX, and long/short VIX, RVX and EURO STOXX 50 volatility futures.

________________________________

Remarks: Due to their conceptual scope – and if not explicitly stated otherwise – , all models/setups/strategies do not account for slippage, fees and transaction costs, do not account for return on cash and/or interest on margin, do not use position sizing (e.g. Kelly, optimal f) – they’re always ‘all in‘ – , do not use leverage (e.g. leveraged ETFs), do not utilize any kind of abnormal market filter (e.g. during market phases with extremely elevated volatility), do not use intraday buy/sell stops (end-of-day prices only), and models/setups/strategies are not ‘adaptive‘ (do not adjust to the ongoing changes in market conditions like bull and bear markets). Index data (e.g. S&P 500 cash index) does not account for dividend and cash payments.

________________________________

Disclaimer

The information on this site is provided for statistical and informational purposes only. Nothing herein should be interpreted or regarded as personalized investment advice or to state or imply that past results are an indication of future performance. The author of this website is not a licensed financial advisor and will not accept liability for any loss or damage, including without limitation to, any loss of profit, which may arise directly or indirectly from use of or reliance on the content of this website(s). Under no circumstances does this information represent an advice or recommendation to buy, sell or hold any security.

I may or may not hold positions for myself, my family and/or clients in the securities mentioned here. Actions may have been taken before or after information is presented, and any opinions expressed in this site are subject to change without notice.

(Data courtesy of MetaStock and Pinnacle Data Corp., and for data import, testing, surveys and statistics I use MATLAB from MathWorks)

Comments (17)

 

  1. Dave says:

    This is really interesting. Thanks Frank for all you do. Have a good week.

  2. John says:

    The volatility risk premium strategy has been around since the 90’s. Don’t forget LTCM employed the short volatility strategy but their problem arguably was overly large leverage. A lot of papers were written back then saying that short volatility strategy is done and returns will never be the same, guess what happens from 2000 to 2008? Once again the short volatility play gets a huge drawdown in 2008 and from then it was once again profitable up until now (wonder when the next reset will be.)

    I see this strategy as a recurring theme. It’s tied to the whole greed and fear theme for investors. Fear will always return, so will greed. As long as you aren’t too greedy to collect the premium, you should be fine. Use stop losses and keep an eye on IVTS and the roll yield.

    I’ve been following this strategy along with my two other models and sometimes they produce contradicting trades. I try to make sure to enter large trades only when all three models have similar output for the trades.

    • Krystian says:

      I agree, but I don’t think this strategy was designed to be traded ‘as is’. Any strategy with a +-50% drawdown would probably not stand a chance of real life test, as seeing half of your gains evaporate would send everyone but the least risk-averse home. But it is a nice addition to one’s portfolio of trading ideas and a visual depiction of how collecting pennies in front of a steamroller could work in practice. Also adding stop losses would need to be well tested before, as I’ve seen many strategies that loose their shine when armed with stop-losses and other money management techniques

  3. Effem says:

    Don’t you have a slight problem in these strategies since the VIX value is a 4:15pm value not a 4pm close?

    • TradingTheOdds says:

      No, I don’t. VIX is regularly more or less flat during the respective quarter of an hour, and in the long run ups and downs averages out. And even if there is some movement, it doesn’t have any impact on the respectives day’s position due to the fact that HV – IV is part of a 5-day average, and regularly well above or below the respective threshold/cutoff (any movement wouln’t trigger a contrary signal, making the difference between > 0 or < 0).

  4. […] Trading The S&P 500 Index (via Implied vs. Historical Volatility) [Trading the Odds] A couple of weeks ago I started a series of postings, all dealing with trading volatility ETNs / ETFs like XIV® (VelocityShares Daily Inverse VIX Short-Term ETN) and VXX® (iPath® S&P 500 VIX Short-Term Futures™… […]

  5. Marco says:

    I think these strategies to have still much to give in the future… Just be careful with leverage…
    http://nightlypatterns.wordpress.com

  6. Konstantinos says:

    how many trades your VRP (2DAY HV and VXMT) version took?
    since you wrote 27 round trips,do you mean 54 trades per year ?

  7. ERic says:

    I am new to the your site and I am wondering which VRP version/strategy is used for the nightly signals via Twitter? Optimized, Revised, VXMT or one that is not public?

    Thanks,

    Eric

  8. RQ says:

    It would be interesting to see the results for using your optimized VRP strategy and SP500 leveraged by factor 4. Can you show them?

    • TradingTheOdds says:

      Of course I could, but it takes some time and effort, and due to the fact that I’am always short of time I prefer to focus on those items on top of my to-do list. If you’re really interested in those stats/equity curves, have you thought about making a donation to my charitable fund first? (the add on the top right of my blog) ;-)

      • RQ says:

        No problem, Frank. Just asked if you already had the results somewhere but forgot to mention them here. If you are busy you don’t have to run the backtests for me :)

        I just wondered would there be big differences between trading XIV/VXX and 4x (or maybe 3x) leveraged S&P 500 ETF with your optimized VRP strategy. I guess trading the S&P 500 ETF instead of XIV/VXX would remove the main risks involved with XIV/VXX, which I think are the XIV termination event (> 80 % drop during a day) and the credit risk involved with ETN’s.

Leave a Reply

Your email address will not be published. Required fields are marked *