Featured -
Posted on **Tuesday, January 27, 2015, 11:14 AM GMT +1**

## Volatility Risk Premium Strategy – And The (Preliminary) Outperformer Is …

*My insights have always been – and will probably be – offered for free, but if the information provided is helpful for your own trading business, any donation to my* **Be it! Children’s Charitable Foundation** *is much appreciated (donations can be sent via PayPal*, see button on the right).

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 (original version) – and its compelling approach (from the paper Easy Volatility Investing from Double-Digit Numerics).

Since then a couple of bloggers have taken up this issue (VRP) as well ( e.g. Volatilty Made Simple , Trading Volatility , QuantStrat TradeR , Evolution Trading , among others), some offering their services (including a VRP strategy) and insights for free while others are subscription based. If you’re interested in the topic, please check for their blogs (highly recommended).

Most recently Volatilty Made Simple had an interesting article ( here ) about the different measures of implied volatility (e.g. VIX^{®}, VXMT^{®}, VX 30-day constant maturity) I present over the course of the series of postings mentioned above, and he is completely correct when concluding ‘*that any advantage of one over the others is likely the result of random chance*‘. Fortunately I’ll show you that there is a huge but: ‘average’ does not mean ‘average’ , especially in combination with those different measures of implied volatility mentioned before.

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

- Long
**XIV**: 5-day*average*of [VIX^{®}– (2-day historical volatility of S&P 500 * 100)] > 0 - Long
**VXX**: 5-day*average*of [VIX^{®}– (2-day historical volatility of S&P 500 * 100)] ≤ 0 - Hold until a change in position.

Image I shows the respective equity curves when – this time – utilizing different kind of *averages* instead of different measures of implied volatility (with one exception to the rule):

**black line** : **I**nverse **D**istance **W**eighted **M**oving **A**verage (IDWMA)

**grey line** : **D**istance **W**eighted **M**oving **A**verage (DWMA)

**blue line** : Simple Moving Average (SMA)

**orange line** : Exponential Moving Average (EMA)

**green line** : Exponential Moving Average, utilizing VIX^{®} futures VX_{1}|VX_{2} , merged into a continual time series as a 30-day constant-maturity futures price ( the underlying for VXX^{®} and XIV^{®} ), and using **1** instead of **0** as cutoff.

Utilizing the VX_{1}|VX_{2} cont. contract in combination with an (5-day) exponential moving average and 1 as cutoff outperforms all other averages by a wide margin and way beyond any S&P 500 / XIV^{®} buy-and-hold strategy (but please note: these are all hypothetical – not actual trading – results, based on partly simulated VXX^{®} and XIV^{®} data), despite the fact that it – up to now – doesn’t use any kind of position sizing and/or money management (the strategy is always all-in), no intraday buy/sell stops (end-of-day prices and orders only), do not utilize any kind of abnormal market filter (e.g. during market phases with extremely elevated volatility), the strategy is not adaptive (do not adjust to the ongoing changes in market conditions like bull and bear markets), there is no over-optimization (e.g. by applying additional rules for seasonalities like FED announcement days, the last 2 days before maturity, S&P 500 overbought/oversold conditions, among others), and none of those findings presented before on this blog are applied, and and and.

**Image I – Total Equity Curve(s)**

(03/26/2004 – present)

**Image II – Drawdown Curve(s)**

(03/26/2004 – present)

Image III shows the respective statistics (key figures) side by side.

(*click on image to enlage*)

**Image III – Statistics – Sidy by Side Comparison**

(03/26/2004 – present)

And last but not least monthly and annual returns (despite the German term for the respective months, I think you get the picture).

(*click on image to enlage*)

**Image IV – Performance Summary – Monthly and Annual Returns**

(03/26/2004 – present)

This will for sure not be my last posting dealing with the Volatility Risk Premium Strategy. From my perspective still the biggest hurdle to overcome are those periods in time when implied volatility (investor fear) is on the rise while realized volatility remains relatively calm, forcing the strategy to stay long XIV^{®} (short volatily) way too long and driving the net asset value into a severe drawdown, like it happened most recently (since December 2014).

* to be continued* … (means more on this to come, stay tuned)

__________________

Have a profitable week,

**Frank**

**Disclosure**: I’am long/short XIV, 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. The results are regularly based on simulated data and/or hypothetical performance and do not represent real trading.

*________________________________*

**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 (86)

I like the idea of using the constant 30d maturity future price as I find that a better measure of implied Vol than the VIX (no weekend/holiday effect, its actually trade-able etc)

Do you have to calculate that yourself or does someone publish that as an index?

thanks

Marco.

Marco,

you may find daily weightings here …

http://www.ipathetn.com/US/16/en/details.app?instrumentId=259118 (Index Components) ,

but in order to have the respective prices at hand before the close, I’am afraid you’ve to do the math yourself (which I do, but it is relatively easy).

Best,

Frank

Frank,

Could you show a sample of your VIX® futures VX1|VX2 series?

Best regards,

Alex

Alex,

an example: The current cycle has 19 trading days: 1/20/2015 is the last trading day for January futures, 2/17/2015 will be the last trading day for February futures. On the close of Monday, January 26, 2015, there are 15 days left.

Calculate:

[15 * 17.225 (= settlement of VIX Feb futures) + (19-15) * 17.625 (= settlement of VIX Mar futures)] / 19 = 17.31 or

(15/19) * 17.225 + (4/19) * 17.625 = 78.95% * 17.225 + 21.05% * 17.625 = 17.31

Best,

Frank

Frank,

Your math looks suspicious. Your weights do not match those on http://www.ipathetn.com/US/16/en/details.app?instrumentId=259118: 78.95% yours vs 78.56% theirs. 78.56% weight can only be got as 22/28. 28 is days between Feb and Mar expirations and 22 is days between 26 Jan and 18 Feb (expiration) minus 1.

Best regards,

Alex

Alex,

VolocityShares (XIV) shows 79% / 21% (machting my figures).

http://www.velocitysharesetns.com/xiv

I think iPath shows the actual dollar weigthings, while my calculation shows an ideal price (assumes one would’ve bought the all futures at the settlement price).

“… the weight of each index component is also adjusted every day to ensure that the change in total dollar exposure for the index is only due to the price change of each contract and not due to using a different weight for a contract trading at a higher price. …” (see http://us.spindices.com/indices/strategy/sp-500-vix-short-term-index-mcap, “Methodology”; only the current roll period is accounted for, not the then following).

Best,

Frank

Frank,

I will check the methodology later but now could you show a sample of the series anyway (for the last week, for example), so I could reproduce your results?

Best regards,

Alex

Alex,

I’am truly sorry, and for sure I don’t want to sound unpolite, but I’am running out of time, and blogging and providing every kind of support is not my business. Everybody is asking for this and that, but only one (in several years of posting) has had the generosity and made a donation to the charitable fund (see my recent posting).

Best,

Frank

Alex,

THANKS for your donation to the fund!!!

1/26/2015 17,31

1/23/2015 18,40

1/22/2015 17,90

1/21/2015 18,88

1/20/2015 20,10

1/16/2015 20,15

1/15/2015 20,84

1/14/2015 19,97

1/13/2015 19,81

1/12/2015 19,42

1/9/2015 18,66

1/8/2015 17,74

1/7/2015 18,82

1/6/2015 19,76

1/5/2015 19,17

1/2/2015 18,01

I hope that helps. Let me know if you need more.

Best,

Frank

Alex,

There are indeed 15 trading days – 22 calendar days less 6 Sat/Sun, less 1 holiday.

Cap

Is your portfolio page going to switch to this method?

The portfolio page is already based on this strategy, but with a couple of improvements …

Frank,

Do you mean the portfolio page and Twitter updates are not anymore dealing with the optimized VRP strategy (http://www.tradingtheodds.com/2014/10/volatility-risk-premium-trading-volatility-part-ii/)?

Even though this new strategy seems amazing, I think the optimized VRP strategy still looks better (Sharpe ratio, max drawdown, Ulcer Index, > 70 % yearly profit).

RQ,

the optimized strategy IS based on the strategy presented in the posting (of course the EMA VX1|2 cont. contract strategy).

Best,

Frank

Ok, thanks for the clarification :)

I must be missing something because the portfolio page has terrible results. How can you start in October with already 25% profit?

The portfolio page started on 10/16/2014 with 0% profits. The first trade was closed out on 10/28/2014 with 28.54% profit. The strategy is currently in a severe drawdown due to the jump in implied volatility, but relatively calm realized volatility.

So back to VXX now. Unless my spreadsheet is completely wrong the original VRP strategy is still at 2.8.

Quite a difference.

The original VRP strategy utlizes a 10-day historical volatilty instead of a 2-day HV. Even a 2-day HV is (too) slow to react to a rising VIX while realized volatility keeps relatively calm (what happenend in December 2014 and last month), leave alone a 10-day HV.

sutluc,

My calculations for EMA5(VX1|2 – HV2) gives -2.44 -> long VXX.

Frank,

Have you found any nice way to detect periods of high IV and low HV in order to avoid massive drawdowns?

Also, if the markets falls in a steady decline with no up days, 2D HV will actually fall not rise. You will have a long XIV signal as the VIX keeps rising.

[…] Volatility Risk Premium Strategy – And The (Preliminary) Outperformer Is… [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™… […]

Thanks for the reply.

Yes, I realize that you are using a modified strategy, I was just commenting on there being much difference. I have not been calculating your modification, only the original with HV10.

Did you stop trading the VRP strategy? e.g. the “discontinued” note on the track report page.

Just curious… the current disconnect between realized vol and implied vol is really tough on the strategy right now…

The VRP strategy on the portfolio page is more or less a study (it doesn’t account for slippage, commissions, fees, …). The formula behind is present in my last posting (utilizing an 5-day EMA of VX1/VX2 futures – 2-day HV of $SPX). Everybody could do the math hisself/herself. I am trading VIX futures and options for a living, and if I’am already long/short volatility (VIX futures and options) I don’t need any additional position and leverage in VXX/XIV ETF/ETNs.

Hi Frank,

Appreciate your work with VXX/XIV trading. Just a question on whether you think XIV trading is still viable. You VRP strategy is approaching its max drawdown and so am I with my realtime trading. I have been trading using a combination of MR, intermediate MR (DVI), VIX futures term structure, VIX MR, and long term trend following. My backtest equity curve looks great, but I’m thinking that the popularity of VXX/XIV trading has eroded any edge with these trades. Any thoughts?

Thanks,

James

James,

it would be really surprising (but not impossible) if a strategy which has had great returns over the course of the past 10 years suddenly stops working. There have been similar periods when implied volatily rose while realized volatily has kept relatively calm, forcing the strategy to continuously short volatility (wrong side) and driving the net asset value into a severe drawdown (see my posting http://www.tradingtheodds.com/2014/12/trading-mid-term-volatility-periods-of-low-historical-volatility/ ).

From my perspective its only a question of time that the market will return into its ‘regular’ modus. I don’t think that we undergo a ‘regime shift’. Patience is a virtue.

Best,

Frank

Hi Frank,

Maybe this was addressed elsewhere and I haven’t seen it. How have you dealt with the fact that VIX/VXMT close at 3:15CST/CDT but XIV/VXX and SPX close trading at 3. If not addressed it gives the strategy 15 minutes of “time travel” ability to go back and execute at a price that’s no longer available.

Dan,

most of the time the difference between the price at 4:00 and 4:15 is small, and from my perspective in the long run any ups and downs are balanced out so it doesn’t matter. It may of course have a (significant) impact in the short run, especially if any news came out between 4:00 and 4:15 (moving the S&P 500 and VIX futures).

Best,

Frank

If my numbers are correct, then the 4:00 – 4:15 90 day average range of the futures contract is 1.27. Since 01/01/2014 the largest range is 4.85 on 10/16.

Upon further review, I discovered that range is incorrect. I wrote a 4:00 – 4:30 session template for Ninja trader, but I forgot to change the bar time to 15 minutes from daily bars.

Therefore, the correct 4:00 – 4:15 90 day average range of the futures contract is .32 – NOT 1.27. I apologize for any confusion I might have caused.

[…] is a test of another “Volatility Risk Premium” (VRP) strategy from the always excellent Trading the Odds (1). The strategy is similar to the Brute Force VRP, DDN’s VRP, and original TTO’s VRP […]

Hi,

Nice work!

I am curious if this is possible to trade with Vix futures?

/Sam

Great work Sir. Brilliant work.

One question: How do you calculate the HVT of the S&P and the moving average of the difference to $VXMT each day – on a spreadsheet manually? I have most of my systems set up in Stockcharts scanners, but this is too complicated for that. When you are travelling do you find it easy to keep on top of the calculation?

Thanks

Rod

Rod,

thanks. I do it manually (Excel speadsheet). And no, I don’t have a problem keeping on top of the calculation when I’am on vacation and/or on a business trip. It takes a couple of minutes only to put the data into the spreadsheet, and when trading for a living it is simply a must in order to be (keep) successful and profitable.

Best,

Frank

Dear sir,

Are the tweets send before the close?

How many minutes before?

Best regards

Xavier

Xavier,

you can do the math yourself (the respective formula is presented in my posting). On apprimately 9 out of 10 days – due to the -day EMA – there is no doubt about what to do (at the close) long before the close (except there would be any exceptinal move in VIX, VIX front month futures and XIV/VXX right before the close). I tweeted the respective move (position maintained or long XIV/VXX) a couple of minutes before the close.

Best

Frank

Is there a way to test this out-of-sample in order to test the robustness? Maybe on another volatility ETF (oil?)?

Sorry, last question.

How would this look with a fixed n% stoploss (10?).

Br.

Xavier,

good question. Please get me informed about your results (please take into account that you’ve to take into account intraday data in order to know if – and when – a stop has been triggered; I don’t have intraday data).

Best,

Frank

Thank you Frank,

My excel table is set up.

I’m going in with the 5d EMA VIX-2d hvolSPX at the next signal change with a small account. We’ll see.

Best to you.

xavier, please can you let us know the moment you get the first signal, and at which price you’re filled? thx

What ‘first signal’ (date, asset, …) ?

Arocep,

currently I have the VRP based on 5EMA (VIX-2dHVOL SPX) at 0.68, which is almost neutral (0). Tomorrow (at close) could be a good time to start with the system.

Regardless, I don’t think it matters when you start as long as you keep following the rules and don’t quit at the first larger DD.

TTO, what’s the rationale if using 1 as cutoff instead of 0 when using the VX1/VX2?

A better return of investment / performance figures !

Frank, the VIX has a new method of calculation since October (including weekly options). The new VIX and the ‘old’ VIX (ticker VIXMO) differ significantly sometimes.

Any ideas how this would affect the strategy using “VIX”?

Up till now I couldn’t find historical daily values of VIXMO to test the difference since October.

Xavier,

I don’t use the VIX as a measure of implied volatility for the VRP strategy (I use the VIX futures cont. contract), and I don’t know a source of historcial data for VIXMO before October 2014. Sorry.

Best,

Frank

Hi Frank,

The results (2010-today) look even better when you add a contango rule.

Say, you buy XIV when VRP reaches your treshold OR when contango is very strong.

You buy VXX when VRP reaches your treshold OR backwardation is very strong.

It improved my backtested results.

Basically, VRP is the main trigger except in case of extreme contango / backwardation (e.g. >7% or <-7% between F1 and F2).

Unfortunately I can't test this earlier than 2010.

Best to you.

Xavier

Xavier,

I’am afraid you’re on the way to overfitting. I know a lot of rules which let the system look even better (e.g. never buy VXX on close of a session before a FED announcement day, …), but the question is if those rules will stand the test of time (among others).

Best,

Frank

Hello Frank,

I enjoy your insights but i’m getting vastly different results. Would you be able to confirm the calculations and hard data in the linked google doc?

https://docs.google.com/spreadsheets/d/1HJrcnYvfTTeWuR6KC3N-2aYUd-RG0_lxbEZhaqx7DnA/edit?usp=sharing

Thank you,

CFAInvestor,

it seems that there are at least 2 “deviations”:

1) Your “weight” is delayed by 1 day. E.g.: Volatility ETFs – as well as the continuous contract concerning the strategy – are invested 100% in XIV (VXX) on close of the session immediately before expiration, not on expiration.

2) Your 5-day EMA is calculated as: X2425*2/3+Y2424*(1-2/3) (e.g. cell Y2425 on 11/7/2013). That seems not to be the correct formula for calculating the 5-day EMA. It should look like: ((X2425-Y2424)*2/(5+1))+Y2424.

Best,

Frank

Good afternoon Frank,

Thanks for getting back to me so quick on this. Can you expound upon the weight being off a day? The XIV site says that May 14, 2015, weight is 89% second month and 11% front month. Is that as of the close on 5/14/15? If so then I’m not following how my numbers are incorrect?

Thanks for helping us all understand this strategy better!

CFAInvestor,

according to your data sheet you assume that XIV and VXX are holding a small position of VIX front month futures into expiration (according to your data sheet weight is 100% / 0% on close of expiration). But that is not the fact. It should be 100% | 0% on close of the session immediately before expiration (no position in front month futures on close of the session immediately before expiration), while it is at (estimated for a 20-day maturity) 95% | 5% ON (close of) expiration.

But that is the least problem and won’t change much with respect to the final result. Much more important is the 5-day EMA calculation.

Best,

Frank

I think i understand, I’ll attempt a change in the formula shortly.

The 5 day EMA formula has been corrected though. Thanks!

So to confirm, the holdings at the close should look like this?

Assuming a 20-trading day expiration period:

Day before Exp: 0% Front Month (May) 100% Second Month (June)

Expiration Day: 95% Front Month (June) 5% Second Month (July)

Day after expiration: 90% Front Month (June) 10% Second Month(July)

Thanks!

Yes, should be working now.

Good morning Frank!

I pulled up the Velocity Share XIV website to see what the weightings of the futures were at the close yesterday and it’s telling me it was 96% second month and 4% front month which is in line with my current calcs.

Please let me know if i’m misunderstanding something here.

http://www.velocitysharesetns.com/xiv

Thanks!

CFAInvestor,

I’am afraid that is a misunderstanding (my fault). In my database (Excel) the settlement date is marked bold, while in your sheet the final trading day is marked bold. Your weightings should be correct.

Best,

Frank

Frank,

I graphed my data and I’m still drastically off somehow, the strategy should’ve been near ~500 million around 9/26/14 but i’m showing a peak around $88 million around that time. Would you be able to transfer your backtest metrics to a google doc to see where I’m off?

I didn’t change the weightings to be 100% second month at the close the day before the front month expiration day since the XIV site says (I’m assuming as of the close) the weightings yesterday (the day before today’s May VIX expiration) were 96% second month and 4% front month. Do you happen to have a source that describes when XIV/VXX rebalance? Perhaps it’s at the open or pre-market trading thus explaining why they still held a position going into expiration day.

Thanks again! Really looking forward to getting this data right to be able to run some of my own ideas on it. One of which i posted on your more recent article about finding when VIX futures are under and overvalued using VIX spot as a reference from a historical standpoint.

Best regards,

CFAInvestor,

I’ve to correct myself: Your weightings are at least partly incorrect. It doesn’t start with 100% on 3/26/2014 (there’re only 14 days left to maturity on that date). This affects every single date (due to the EMA calculation). There might be other issues (e.g. incorrect maturity dates) as well.

Best,

Frank

Ah, I see the confusion then. Great to hear we’re getting closer to the same calculations.

Per your comment that the weighting doesn’t start at 100% front month (May 2004) on 3/26/2004 – There was no April VIX future contract to reference at that time so the May and June contracts had to be used correct? And the first recorded trade date of the May’04 futures was on 3/26/2004 according to CBOE’s historical VIX data here:

http://cfe.cboe.com/products/historicalvix.aspx

On a separate note, when i was building the data from the above CBOE website into the google doc I ran into a few issues on days right after expiration in the years 2004-2008 that the next month futures were not available so I used estimated values for the 2nd month futures for those days based on movement in VIX relative to the value of the days where the 2nd month futures were quoted. Perhaps you have a different method of correcting for those gaps in the simulated data?

Would you be willing to share your simulated XIV and VXX closing values prior to their inception dates? I feel like that may be where we are deviating as I got mine from an unreliable source.

Thanks!

CFAInvestor,

this is my source for simulated XIV and VXX closing values prior to their inception dates:

http://sixfigureinvesting.com/downloads-2/

Best,

Frank

Great post. You might also find this interesting – Top 6 Reasons to Trade Volatility http://www.vixstrategies.com/top-6-reasons-to-trade-volatility-2/

Hello Frank,

Congratulations for the excellent presentation of your strategy on this blog.

I combed the article and the comments & remarks and was still at a loss to found out how to calculate the 2-day historical volatility of the SPX.

I would go for hourly intraday returns, unless you have a better idea?

Best regards,

Cedric

Cedric,

thanks a lot for your kind words.

The 2-day HV is calculated as: STDEV [ LN (today’s close / yesterday’s close) : LN (yesterday’s close / day before yesterday’s close)] * SQRT (251)

Best,

Frank

Frank,

Many thanks for your quick reply.

Best regards,

Cedric

Hi Frank,

shouldn’t it be

STDEV [ LN (todaysclose / yesterdaysclose) : LN (yesterdaysclose / daybeforeyesterdaysclose)] * SQRT (251)

Best,

Xavier

Xavier,

thanks, and you’re absolutely correct. I just copied my Excel formula without thinking twice … -:(

Cedric: My comment and respective formula has been corrected.

Best,

Frank

Dear Frank,

just wanted to keep you informed. I started with this VRP strat (using VIX, not the 30-d constant maturity) 3 months ago.

Went +20% up, -18% dwn, and today almost at break-even (+1%) with 2.5 round trips (buy/sell XIV, buy/sell VXX, buy XIV).

I’d like to keep this strategy going for at least a year and hope for the best.

Best to you.

Hi Frank,

Are you still following this (or any) strategy?

Didn’t see any comments or tweets from you.

Best regards.

Xavier,

up to now this is the best-performing strategy trading XIV and VXX (even it is significantly under water at the moment) I could come up with, but I’am still working on it (there is always room for improving, especially with respect to maximum drawdown …). As soon as there is something worth publishing, I will be back with another posting …

Best,

Frank

Thanks Frank

No problem. I am not worried about the current strategy. I just wondered if you were still around.

Best to you.

Hi Frank,

Is your strategy still long XIV for the moment?

The spreadsheet I am using trying to duplicate your strategy still is, quite to my surprise. (Fortunately still did not put any money in ;-) )

Best regards.

Hi, where can I find VX30 or VX45 data please? I would backtest optimized VRP strategy. Regards Martin

Hi,

what do you mean with “VX30” and “VX45” data ?

Best,

Frank

I think, It´s 30-day constant maturity price of VIX futures and 45-day constant maturity price of VIX futures. Or not?

For example, for “Optimized VRP Strategy” I need calculate the following: the 5-day exponential moving average of [30-day constant maturity price of VIX futures – (2-day historical volatility of SPY * 100)]

And where can I find 30-day constant maturity price of VIX futures?

I don’t know anybody using a 45-day constant maturity price of VIX futures. I calculate the 30-day constant maturity price of VIX futures myself, otherwise it can be found here (not free of charge): http://investing.kuchita.com/

Thank for your response. And how did you calculate 30-day constatnt marturity price of VIX?

It is possible in Excel? On Investing.kuchita.com is a pack with historical data of VIX futures, but how can I get weights (for example, 6 day do maturity/30*price of first futures + 24/30*price of second futures) ?

Take a look at Peter Jacobsen’s comment (below) and link to his workbook:

https://docs.google.com/spreadsheets/d/1ivrUnfxMWJ5pRFzWVbf4VleysyucKaDGU_CLfX1yheU/edit?usp=sharing

I filled Peter’s workbook out up to September 10, 2015. Interesting to see how two methods – Frank’s and Peter’s start to differ once VIX jumps around August 19 (if of course I put everything correct into the worksheets):

https://doc.co/zsf3w4

Hi Frank, Really interesting strategy and thanks for sharing Peter Jacobsen’s work. Just like Peter and other people who left a comment, I’m trying to replicate this strategy and am suprised the results are different. Would it be possible for you to share your thoughts on where the difference is coming from?

Also is the assumption (as per Vadis’ and Peter’s work) that your strategy has remaines long VXX since 6/30 or haven’tt you been able to share the updated results since?

Thanks!

Differences – if any – should be very very rare (e.g. long XIV on a FED day -no questions ask-, or long XIV/VXX when the S&P 500 is significantly ‘oversold’/’overbought’). Otherwise any differences (as discussed in the coment section before) may be due to simulated data sources (before XIV/VXX were released for trading), the method for calculating the EMA, data issues (e.g. utilizing SPY w/dividend or wo/dividends instead of S&P 500, among others), calculating days to expiration, …

Due to the fact that I trade for a living and the blog is completely free of charge, I don’t like the idea to be committed to update data and performance on a daily basis. It’s just a hobby, not a business.

Updated Google Doc Spreadsheet earlier mentioned here:

https://docs.google.com/spreadsheets/d/1gt1pwM5thSqVpg0UsPMrjWaCaFcAcAX84Wqkvnf_DBU/edit?usp=sharing

Happy to discuss, i’ve enabled commenting in my google doc if you see an error or have a question.

Thanks!

The Tab with the data/tracking is “VXX/XIV Strat”