Kudos to MMs

Quote from nitro:

Almost every single position I have put on in this terrible selloff has shown a decent profit from the get go, only for it to go red. Other than saying that when VIX is high, we need to take profits at regular intervals based on some simple function of VIX, I have no theory as to what is correct.'

The conclusion that I have come to is that,

A) In quiet longer term Bull markets with a tame VIX ( < 23 ish), and long positions to take advantage of them, it pays to buy pullbacks against the short term trend.

B) On the contrary, when markets speed up and go lower fast with a climbing VIX (> 23.5 ish) and you want to go long, fading and buying support (catch a falling knife) should only be done when those fast markets slow down, with a collapsing VIX, and only on crossovers from below (instead of having standing limit bids at support, have stop limit on crossovers). You can buy crossovers from below on a high VIX with a tight stop at the support.

Those two simple rules would cover 90% or more of the correct experience required. Taking profits also has this asymmetry, but I am not sure what it is yet.
More...

It looks to me that your rules A and B can be consolidated into a single one: buy pullbacks from a trend up when volatility is subsiding. I am going to backtest this theory and report here.
 
Quote from nonlinear5:

It looks to me that your rules A and B can be consolidate into a single one: buy pullbacks from a trend up when volatility is subsiding. I am going to backtest this theory and report here.
More...
That is nowhere near rich enough, and it is not even correct since you can buy on a short term downtrend on declining VIX too on a crossover from below. The distinction is subtle, but non-trivial. One buys pullbacks in a strong uptrend market with low VIX (sleepy bull markets). The other once "catches falling knifes" on even a high VIX, but a VIX that is collapsing fast, and only at support and only on crossovers from below of that support.

Also, you have to know where R/S is (and in my case, what FV is since you don't want to buy a market with a FV below the market). Still, it would be interesting to hear the result, especially if you showed the code (I have no "interest" in your strategy, only the assumptions to see if they lead to a strong conclusion.)
 
For example, I am long ES from 1125. At 1145 I had 20 handle gain. If I followed the rules above, I should take profits there (I didn't, I sold it and converted to a 55 delta option - not the same thing. I know the whole position is not profitable no need to point it out.)

In other words, compute Implied move from VIX, and in a high VIX environment, if you are playing against the (short term) trend, take profits "early" whatever early means in your time frame.

The point is that high VIX (vola) makes more things happen in the same tick-time.
 
Quote from nitro:

That is nowhere near rich enough. You have to also know where R/S is. Still, it would be interesting to hear the result, especially if you showed the code (I have no interest in your strategy, only the assumptions.)
More...

Certainly. I manage an open source project and that's what I am going to use for this test:
http://code.google.com/p/jbooktrader/

In terms of R/S, I use exchange limit order book balance, instead of price.
 
Quote from nitro:

For example, I am long ES from 1125. At 1145 I had 20 handle gain. If I followed the rules above, I should take profits there (I didn't, I sold it and converted to a 55 delta option - not the same thing. I know the whole position is not profitable no need to point it out.)

In other words, compute Implied move from VIX, and in a high VIX environment, if you are playing against the (short term) trend, take profits "early" whatever early means in your time frame.

The point is that high VIX (vola) makes more things happen in the same tick-time.
More...

I haven't read 90% of the thread, so indulge me when I ask what synthetic did you convert to (55D option)? Did you convert the futures to a synthetic long call at edge?

If so, wtf are you buying the vol-line here, especially if you're bullish?
 
Quote from atticus:

I haven't read 90% of the thread, so indulge me when I ask what synthetic did you convert to (55D option)? Did you convert the futures to a synthetic long call at edge?

If so, wtf are you buying the vol-line here, especially if you're bullish?
More...
No, I just sold ES and then bought a 55 delta call option. To protect from a vola implosion on rising markets, I also buy VIX put when I buy NTM SPY calls.
 
Quote from nitro:

No, I just sold ES and then bought a 55 delta call option. To protect from a vola implosion on rising markets, I also buy VIX put when I buy NTM SPY calls.
More...

You're mixing terminology. Are you buying the natural call? Because "converting" implies the synthetic. There is no "converting" if you're long at n and cover at n+x. That's a round-trip and then you're buying a 55D call.

I am trying to understand wtf you're doing, but it seems like you're intentionally trying to complicate matters.

Why not simply buy the risk-reversal? Tons of edge if you're right on direction. Uh, whoops...
 
Quote from nonlinear5:

It looks to me that your rules A and B can be consolidated into a single one: buy pullbacks from a trend up when volatility is subsiding. I am going to backtest this theory and report here.
More...

I put together the corresponding strategy:
-- buy when a) price is at the support and b) volatility is low
-- sell when a) price is at the resistance or b) volatility is high

I ran this through the optimization engine using Oct 2010 to Aug 2011 as my test period, on the ES data. I found no evidence that this strategy can provide any meaningful edge. I did make various assumptions, so feel free to validate with your own back test.
 
You must log in or register to reply here.
Back
Top