I read an article the other day claiming "The VIX is Very Predictable"
It claimed the method to buy/sell "the VIX" (which itself can't be bought or sold) is to take the 11 day ema and sma, plot them and buy when the ema goes under the sma, and sell when the ema goes above the sma.
Here is a chart of the vix and a little indicator that I coded up using the method described. The red line is the ema and the white is the sma. I colored the background "green = buy" and "red = sell".
Seems there needs to be a bit of refinement for this method to work ... but it's a start. Would welcome other ideas in the comments.
I am happy you mentioned you can't trade the VIX which is just a statistic calc'd by the CBOE -- many people don't realize this!
One thing to point out, which is really neither here nor there but FYI -- I am familiar with that author, and he was big into mean-reversion strategies. During the period of 08-09 - mean reversion strategies had an extraordinary period of outperformance relative to baseline results.
You could literally sell short, don't know, SKF on *any* 3 days in a row up, or RSI 2 > 90 and close 1 or two days later with like an 80-90% win rate and low drawdown. I remember them being particular popular in the blogosphere at that time. The issue is that markets being cyclical, this strategy fell out of favor and began to underperform mid 2009 to more or less now, as the markets got more trendy.
So what is my point? The only point is that this strategy was written and tested in 2008, a highly mean-reversionary period in the market, by an author who's entire line of work was centered around mean-reversion strategies. I think this would be important background info to have to test the efficacy of it today.
codypd
⋅
That is very helpful. Don't want to imply here with my attempt to implement the indicator that the original author "had it wrong." Good of you to point out that context. Thank you.
SPYderCrusher
⋅
Oh of course, that wasn't my intent either -- just thought it would be some back history worth mentioning is all :D
flappinfish
⋅
My small experience of trying to trade the VIX via VXX was far less then predictable. VXX is NOT a predictable product. My experience was aprox. 30% correlation when VIX was positive and near 1 to 1 correlation when VIX was down. Occasionally the two moved in opposite directions. Be very careful with VXX.
codypd
⋅
My experience is when you are playing the VIX derivatives you are playing two things: the VIX and the derivative's contango/backwardation behavior. I have found that when youtrack them seperately the success rate improves a great deal. I trade primarily UVXY and SVXY because I have modelled their contango structure which is pretty consistent once the variables are understood.
codypd
⋅
Here is an example of a ratio chart that shows the consistency and predicatability of contango for UVXY. When in contango, the rate will be fairly consistent over time for a given "run" tradingview.com/v/Zs32QmmO/
mike1v
⋅
codypd: I don't know whether or not you read the follow-up article to "The VIX is very predictable" or not. The article (link below) looks at the VIX and historical volatility in more detail. I also don't know if the article will help in any refinements you might make, but thought I'd post it just in case. The comments section has some good stuff also. Good luck!
The easy way to trade VIX is to wait for a peak, and keep shorting because of the decay involved with futures (VXX).
The only risk being the fat-tailed distribution (eg: another lehman brother), so risk management is equally important.
LastBattle
⋅
I'll never trade/make decisions based on a single indicator.
I can see that VIX EMA is smoothed which is clearly a flaw to me. They often lag behind by several candles, and by the time you get a buy signal, risk reward would have already been low. Would you want to be the herd of sheeps following the indicator or would you want to get ahead of the sheeps? Remember, there'll always be 90% losers and 10% winners.
There are also exceptional conditions where the market may remain constantly overbought for an extended period of time due to fears from events like grexit if should it happen. Can you still trust that indicator to sell short?
One thing to point out, which is really neither here nor there but FYI -- I am familiar with that author, and he was big into mean-reversion strategies. During the period of 08-09 - mean reversion strategies had an extraordinary period of outperformance relative to baseline results.
You could literally sell short, don't know, SKF on *any* 3 days in a row up, or RSI 2 > 90 and close 1 or two days later with like an 80-90% win rate and low drawdown. I remember them being particular popular in the blogosphere at that time. The issue is that markets being cyclical, this strategy fell out of favor and began to underperform mid 2009 to more or less now, as the markets got more trendy.
So what is my point? The only point is that this strategy was written and tested in 2008, a highly mean-reversionary period in the market, by an author who's entire line of work was centered around mean-reversion strategies. I think this would be important background info to have to test the efficacy of it today.