The Ultimate Historical and Implied Volatility Rank and Percentile Indicator

Bard · Tue, 22 Oct 2019 15:56:30 +0000

Following on from this topic: https://www.prorealcode.com/topic/a-better-implied-volatility-rank-and-percentile-indicator I would be grateful if we can get this indicator coded. I think this would be an essential addition to the PRC library and am wondering if it is possible to code an Historical Volatility (HV) and Implied Volatility (IV) Rank and Percentile Indicator. It would be particularly beneficial for Currency, Indices and Options traders but can be applied to most markets. IV can only ever be approximated as it is derived from the Black Scholes option pricing model. Black, Scholes and Merton won the Nobel Prize for coming up with their options pricing model because it predicted the future probability and pricing of asset prices. Option pricing, historical and implied volatility are crucial in that understanding of future asset pricing. For thorough explanation on probabilities please see my explanations here: https://www.prorealcode.com/topic/options-probabilities-using-the-black-scholes-assume-a-normal-distribution/ https://www.prorealcode.com/topic/a-better-implied-volatility-rank-and-percentile-indicator/#post-110512 Two things come to mind: 1) A comparison in the same indicator needs to be made between the HV and IV. Example: Please see the screen attached, it’s from ivolatility via the Options Council website: https://www.optionseducation.org/toolsoptionquotes/historical-and-implied-volatility, it’s a blue and gold line graph of the HV v’s the IV. Other vendors also supply IV 30, IV 60 and IV 90 comparisons. I was then hoping we could have an indicator like this one I found on Trading View (linked just below) so it shows not just the IV rank but also the Percentile (so it’s scaled by 100%). The code is locked on Trading View but I am hoping it isn’t too far of the version I linked to above on PRC. https://www.tradingview.com/script/Z30icKNi-IV-Rank-IV-Percentile/ “This indicator is meant to be a substitute for Implied Volatility Rank and Percentile for traders who do not have access to readily available options data. This indicator is based on the William’s VixFix which is an indicator that mirrors the VIX , which charts the implied volatility of the SPX. The great thing about the VixFix is that it can be applied to any security, not just the SPX. IV Rank is calculated by taking the highest and lowest values over the past however many periods you choose, and seeing what percentage of the way between those values the current IV value is. For example if over the past 5 periods the highest IV value was 30 and the lowest IV value was 10, and the current is 20; then the IV Rank would be 50% because 20 is 50% of the way between 10 & 30. IV Percentile is calculated by looking at all of the past values, not just the highest and lowest, and seeing how many of those values were below the current. (IVP tells you the percentage of time that the IV in the past has been lower than current IV). For example lets say over the past five periods the IV values were : 30,10,11,15, & 17; while the current IV value is 20. As stated before the IV Rank would be 50%, while the IV Percentile would be 80%, given that 4/5 of the values were below the current IV value of 20. IV Rank and IV Percentile are often wrongly used interchangeably, but as shown here they are very different. Most people use IV Rank as their main options tool; while IV percentile is a great way to give IV Rank context. Whichever you choose to use, or even both, does not really matter as long as you use either one or both consistently. IV Rank and IV Percentile are mainly used in this way: when IVR/IVP <50, buy options, when IVR/IVP >50 sell options. The reason that you buy options when IVR/IVP is low is because IV is mean reverting, so you would expect IV to eventually start increasing towards the mean, causing prices to move. The reason you would sell options when IVR/IVP is high is because IV is mean reverting and you would expect IV to decrease towards the mean, causing prices to move sideways. PRC Version – High volatility market bottoms (but without low volatility market tops). //Implied Volatility Rank and Percentile Indicator //By Vonasi //Date 20191018 //settings p = 200 vixfixperiod = 22 //Vix Fix wvf = ((highest[vixfixperiod](close)-low)/(highest[vixfixperiod](close)))*100 //Boundaries upper = highest[p](wvf) lower = lowest[p](wvf) wvflev = ((wvf-lower)/(upper-lower))*100 //Percentile rnkCount = 0 count = 0 FOR i = 1 to p-1 count = count + 1 IF wvflev[i] < wvflev THEN rnkCount = rnkCount + 1 endif NEXT PercentileRank = (rnkCount/count)*100 //colours r = 128 g = 0 if wvflev >=50 then r= 0 g = 128 endif r1 = 255 g1 = 140 b1 = 0 if PercentileRank >=50 then r1 = 0 g1 = 255 b1 = 255 endif RETURN 0 as "0%",100 as "100%",50 as "50%",wvflev coloured(r,g,0) as "Williams Vix Fix", PercentileRank coloured(r1,g1,b1) as "Percentile Rank" 2) Can it be programmed so that the IV also can find low volatility market tops as well as high volatility bottoms, — i.e. instead of just having a fear barometer when Put Premiums are rising sharply and the market is due to bottom out, also have a greed barometer for Call Premium rises and potential low volatility market tops) like this indicator here: Here’s a dropbox link: https://www.dropbox.com/sh/bcmihvkefjf98cj/AAADlrHpNFlekHHtIe1vSVjJa?dl=0 to the Ninja Trader 8 Williams Vix Fix code (in #C) shown in the image below and sourced from here: https://ninza.co/product/williams-vix-fix Historical Volatility Please see the formula for HV is from the attached Fidelity Investment screenshot. On PRC we have this version fo HV: indi= HistoricVolatility hh = highest [365](indi) ll = lowest[365](indi) result = 100*((indi-ll)/(hh-ll)) return result and also this version: Length = 20 annualVol = 365 periods = 7// 1 = intraday chart 7 = daily chart Price = log(close / close[1]) sigma = std[length](Price) HVol = (sigma * sqrt(annualVol / periods)) * 100 lowVol = lowest[annualvol](HVol) HVrankUp = HVol - lowVol maxVol = highest[annualvol](HVol) HVrankLow = maxVol - lowVol HVR = (HVrankUp / HVrankLow)*100 return HVR coloured(0,0,255) Both give fairly similar outputs but the log version is more precise. Implied Volatility As noted on the Trading View IV indicator post the formula approximation for IV has been coded by Larry Williams and is called the Vix Fix (with a moving average): https://www.prorealcode.com/topic/indicator-request-for-nic/#post-89764 pd = 22 // "LookBack Period Standard Deviation High" bbl = 20 //"Bolinger Band Length" mult = 2.0 //"Bollinger Band Standard Devaition Up" lb = 50 //"Look Back Period Percentile High" ph = 0.85 //"Highest Percentile - 0.90=90%, 0.95=95%, 0.99=99%" (For Options writers I would recommend waiting until at least the 95th percentile) n = 10 //pl = 1.01 //"Lowest Percentile - 1.10=90%, 1.05=95%, 1.01=99%" mult = max(mult,1) mult = min(mult,5) wvf = ((highest[pd](close)-low)/(highest[pd](close)))*100 sDev = mult * std[bbl](wvf) midLine = average[bbl](wvf) //lowerBand = midLine - sDev upperBand = midLine + sDev rangeHigh = (highest[lb](wvf)) * ph //rangeLow = (lowest[lb](wvf)) * pl if (wvf >= upperBand or wvf >= rangeHigh) then //or (wvf <= lowerBand or wvf <= rangeLow) then color = 1 else color = -1 endif ave = average[n](wvf) RETURN wvf coloured by color as "Williams Vix Fix", ave style(line,1) as " WVF Average" Any help getting this codified would be much appreciated! Cheers Bard

This topic has 21 replies, 3 voices, and was last updated 5 years ago by Bard.

Tagged: fix, Historical, implied, percentile, Rank, VIX, volatility

Viewing 7 posts - 16 through 22 (of 22 total)

← 1 2

10/30/2019 at 4:10 PM #111688

Bard

Afternoon @Nicolas, I got this reply from Ninza about there Implied Volatility Vix Fix code (because the won’t share the complete code with me) :

The original Williams VIX Fix just includes the bottom-catching algorithm.

Our indicator inverses that formula to add the (low volatility) top-catching algorithm, which is our improvement.

For the bottom-catching part, you can compare our indicator with other source-code versions on the internet to check the accuracy.

So.. it appears that we would need to amend the formula for the Williams’ Vix Fix High Volatility Market Bottoms code posted below here? :

wvf = ((highest[pd](close)-low)/(highest[pd](close)))*100

1	wvf = ((highest[pd](close)-low)/(highest[pd](close)))*100

This IV indicator and specifically this line of code I think has been the basis for the confusion here and of which I’d asked a few times about the need to modify it in my Better Implied Volatility post and above (that’s my defence and I’m sticking to it!). I think by not modifying it (wfv) and using this code (wvflex) that seeks to establish extremes that is why we maybe only seeing mirror images between the two indicators from Vonasi’s and your conversion of his code, right? Re: mirroring, please refer to my image in my post #111235 just directly above or see this zoomed in scan of it posted here:

The top indicator is Vonasi’s conversion from my request here: https://www.prorealcode.com/topic/a-better-implied-volatility-rank-and-percentile-indicator/

Cheers,

Ps/ It’s now very tricky adding code here? I got this on the first submit:

<pre class=”lang:probuilder decode:true” title= “Williams’ Vix Fix High Volatility Market Bottom” etc.
Basically it’s only on editing my post that I can truly get the indicator code to display properly in a grey highlighted box using the latest Safari browser.

Bard

Participant

Topics: 138

Replies: 814

Been thanked: 99 times

10/31/2019 at 7:39 PM #111766

Bard

I think this has sorted it @Nicolas: — although red/green IV can rank high for market tops if you use wvflev =100- ((wvf- etc) and IF (wvflev[i]) < (wvflev) THEN

Low Implied Volatility Inverse Vix Fix (for Market Tops) with Rank, Percentile and Historic Rank:

//Implied Volatility Rank and Percentile Indicator — MARKET TOPS
//By Vonasi //Date 20191018

//settings
p = 365
vixfixperiod = 22

//Vix Fix - Implied Volatility Proxy

wvf = ((highest[vixfixperiod](close)-low)/(lowest[vixfixperiod](close)))*100

//Boundaries
upper = highest[p](wvf)
lower = lowest[p](wvf)

wvflev = ((wvf-lower)/(upper-lower)*100)

//Percentile
rnkCount = 0
count = 0
FOR i = 1 to p-1
count = count + 1
IF (wvflev[i]) > (wvflev) THEN
rnkCount = rnkCount + 1
endif
NEXT

PercentileRank = (rnkCount/count)*100

//colours
r = 255
g = 0
if wvflev >=50 then
r= 0
g = 255
endif

r1 = 255  //255
g1 = 160  //160
b1 = 0  //0

if PercentileRank >=95 then
r1 = 0
g1 = 255
b1 = 255
endif

//Historic Volatility

Length = 22
annualVol = 365
periods = 7// 1 = intraday chart 7 = daily chart

Price = log(close / close[1])
sigma = std[length](Price)
HVol = (sigma * sqrt(annualVol / periods)) * 100
lowVol = lowest[annualvol](HVol)
HVrankUp = HVol - lowVol
maxVol = highest[annualvol](HVol)
HVrankLow = maxVol - lowVol
HVR = (HVrankUp / HVrankLow)*100


RETURN 0 as "0%",95 as "95%",50 as "50%", HVR coloured(0,0,255) as "Historic Volatility Rank", wvflev coloured(r,g,0) style(line,4) as "Williams Low IV Vix Fix", PercentileRank coloured(r1,g1,b1) as "Percentile Rank"

//Implied Volatility Rank and Percentile Indicator — MARKET TOPS

//By Vonasi //Date 20191018

//settings

p = 365

vixfixperiod = 22

//Vix Fix - Implied Volatility Proxy

wvf = ((highest[vixfixperiod](close)-low)/(lowest[vixfixperiod](close)))*100

//Boundaries

upper = highest[p](wvf)

lower = lowest[p](wvf)

wvflev = ((wvf-lower)/(upper-lower)*100)

//Percentile

rnkCount = 0

count = 0

FOR i = 1 to p-1

count = count + 1

IF (wvflev[i]) > (wvflev) THEN

rnkCount = rnkCount + 1

endif

PercentileRank = (rnkCount/count)*100

//colours

r = 255

g = 0

if wvflev >=50 then

r= 0

g = 255

endif

r1 = 255 //255

g1 = 160 //160

b1 = 0 //0

if PercentileRank >=95 then

r1 = 0

g1 = 255

b1 = 255

endif

//Historic Volatility

Length = 22

annualVol = 365

periods = 7// 1 = intraday chart 7 = daily chart

Price = log(close / close[1])

sigma = std[length](Price)

HVol = (sigma * sqrt(annualVol / periods)) * 100

lowVol = lowest[annualvol](HVol)

HVrankUp = HVol - lowVol

maxVol = highest[annualvol](HVol)

HVrankLow = maxVol - lowVol

HVR = (HVrankUp / HVrankLow)*100

RETURN 0 as "0%",95 as "95%",50 as "50%", HVR coloured(0,0,255) as "Historic Volatility Rank", wvflev coloured(r,g,0) style(line,4) as "Williams Low IV Vix Fix", PercentileRank coloured(r1,g1,b1) as "Percentile Rank"

Ignore first image.

Bard

Participant

Topics: 138

Replies: 814

Been thanked: 99 times

11/02/2019 at 6:54 PM #111911

Bard

Creating an Inverse Low Volatility Vix Fix…

Larry Williams’ original High Volatility Vix Fix IV proxy is:

wvf = ((highest[pd](close)-low)/(highest[pd](close)))*100

1	wvf = ((highest[pd](close)-low)/(highest[pd](close)))*100

I initially only changed the original High Vol Vix Fix denominator from highest to lowest, but logic tells me that the inverse of that high volatility wvf is not correct in line 10 above and should in fact be: (where pd = 22 periods)

wvf = ((high-lowest[pd](close))/(lowest[pd](close)))*100

1	wvf = ((high-lowest[pd](close))/(lowest[pd](close)))*100

..yet that produces a worse result for spotting market tops with green spike bars? Pls see images, the first being the correct code ( ((high-lowest[pd](close))/(lowest etc ). The top indicator (of four) is the one in question in both the attached images. Do you have any ideas please @Nicolas because these volatility indicators are great but they’re driving me nuts!? Lol.

Bard

Participant

Topics: 138

Replies: 814

Been thanked: 99 times

11/04/2019 at 9:20 AM #111985

Nicolas

Normal behavior as these peaks mark the candle when the new highest value of WVF is, so that the normalization is doing its job.

100% WVF is on that candle, so the oscillator is marking it with a spike in the curve.

Nicolas

Keymaster

Topics: 81

Replies: 20863

Been thanked: 4773 times

11/04/2019 at 2:34 PM #112042

Bard

Thanks @Nicolas, in your opinion though which would be the correct code to inverse the wvf code and find low volatility extremes:

wvfinverse = ((highest[pd](close)-low)/(lowest[pd](close)))*100

1	wvfinverse = ((highest[pd](close)-low)/(lowest[pd](close)))*100

 wvfinverse = ((high-lowest[pd](close))/(lowest[pd](close)))*100

1	wvfinverse = ((high-lowest[pd](close))/(lowest[pd](close)))*100

Cheers,

Bard

Participant

Topics: 138

Replies: 814

Been thanked: 99 times

11/04/2019 at 2:37 PM #112043

Nicolas

The second one, as it is the almost exact same one as a Stochastic.

1 user thanked author for this post.

Bard

Nicolas

Keymaster

Topics: 81

Replies: 20863

Been thanked: 4773 times

11/04/2019 at 2:40 PM #112044

Bard

Cheers, that’s what I thought although for some reason the first produces far better indications of the market topping out.

Bard

Participant

Topics: 138

Replies: 814

Been thanked: 99 times

Author

Posts

Viewing 7 posts - 16 through 22 (of 22 total)

← 1 2

Find exclusive trading pro-tools on