The VIX is the volatility expectation for the next 30 days. My question is can it be
used to get an estimate for today using these steps (can't type question mark)

As of 8am PST 3-21-24 (can't use forward slash for date)
ES = 5318.50
VIX = 12.79

1) VIX divided by 30 = 0.43 (can't use forward slash to divide)
2) Multiply #1 by ES = 5318.5 * 0.0043 = 22.87 = 23
3) Take ES, then add & subtract #2
(High) 5318.50 + 23 = 5341.50 & (Low) 5318.50 - 23 = 5295.50
4) Would this work for an estimate of today's range (can't use question mark)

I've done this manually for the past few days and it seems to hold up fairly well
but I'm not sure if it's just coincidence or it's valid.

On a side note, if anyone has any idea why I can't type a question mark or
forward slash (same key) but instead get teleported to the search bar on the top
right, I'd love to hear it. These work outside nexusfi, so I don't see how it's my
keyboard.

I thank you all in advance, JM

VIX is reported as an annualized number. Since volatility is statistically defined as the square root of variance, the monthly volatility implied by VIX can be calculated by dividing its level by the square root of 12 because there are 12 months in a year.

In your case, you might try to find the square root of 365 and use that to find today's expected range.

Ex.: if VIX level is 15 then divide by square(12) (monthly expected range) = +/-4.3%
Ex.: if VIX level is 15 then divide by square(365) (daily expected range) = +/- 0.52%

https://www.spglobal.com/spdji/en/vix-intro/

On a day to day basis, i would not use this technique. Not accurate enough.

trendisyourfriend 

VIX is reported as an annualized number. Since volatility is statistically defined as the square root of variance, the monthly volatility implied by VIX can be calculated by dividing its level by the square root of 12 because there are 12 months in a year. In your case, you might try to find the square root of 365 and use that to find today's expected range. Ex.: if VIX level is 15 then divide by square(12) (monthly expected range) = +/-4.3% Ex.: if VIX level is 15 then divide by square(365) (daily expected range) = +/- 0.52% https://www.spglobal.com/spdji/en/vix-intro/ On a day to day basis, i would not use this technique. Not accurate enough.

Thank you. Your post was very helpful.

Well if you look at the weekly option in /ES or SPY this is already factored in the expected move. VOL plays a large part in options. Especially the indexes.

This video talks about the research and how often the weekly expected move was accurate. So if your your trading outright, this could be very useful. Maybe the expected move on zero day options can help with daily trading in futures.

The idea is a good one, but it has limitations. I wrote my MA thesis on range prediction and the problem goes like this.

1/ there is some "true" volatility underlying the price generating process
2/ you can estimate this true volatility using different approaches. One (imprecise) is the stdDev of returns. Anoher (imprecise) is the daily range (+/- some multiplication constant). Another (the best you can get) is the realized variance or realized range using intraday data, sampled at 5-minute intervals for best accuracy.

Option prices and VIX reflect the true volatility. To estimate, how well the daily range captures this true volatility, you do a regression like

RNG[t] = alpha + beta * RV[t] + epsilon

estimated for days, where RNG[t] is the daily range of day t and RV is the realized variance from 5-minute data of the same day. The estimated alpha will be close to zero, beta will be close to one (which is fine and in line with expectations), but the R-Square of this is going to be around 65%. Roughly speaking, this means that daily range is composed of the true volatility, as captured by VIX, plus some random term that is "approx half as large as the size of true volatility".

Why the daily range contains such a large error term can be seen when you imagine two days. A choppy one and a trending one. If both days were to have the same realized volatility, the choppy day would have a much smaller range than the trendy one.

There are some econometric models used for range prediction, back in the day HAR and cointegration of highs/lows was used, but neither can explain the random "epsilon" term, as its random.