The ACD Method

Quote from Shanb:

I havent tested this pair for a cointegrating relationship, but as you can tell there does seem to be a tendency to revert as it gets to certain levels. Here its like buying the dip into the longer term trend!

It's been trending the entire year! This last dip it had was the first one of any consequence for the year.
 

Attachments

Quote from Maverick74:

It's been trending the entire year! This last dip it had was the first one of any consequence for the year.

A pair can be trending and still have a co integrating relationship tho!
 
Quote from Shanb:

These are for mean reversion

OK. I have NO interest in mean reversion. LOL. I don't want to trade in and out of this stuff looking for pennies. HFT's do that pretty well at zero cost. I rather ride that trend. :)
 
Quote from Maverick74:

OK. I have NO interest in mean reversion. LOL. I don't want to trade in and out of this stuff looking for pennies. HFT's do that pretty well at zero cost. I rather ride that trend. :)

Ok :)

These are all multi-day spreads, just thought it was a cool tool. Maybe one can even look for spreads that totally suck with the mean reversion approach and flip it! I haven't found any other easy accessible tools for this and a little creativity can even help you find some super trending spreads!

There are some stocks that lend themselves well to the MV approach tho. Check out AGNC V BMR

http://www.pairtradinglab.com/backtests/TuEP2VWQ_HCx1cHN

3.5 PF with 80% winners :)
 
Quote from Shanb:

Ok :)

These are all multi-day spreads, just thought it was a cool tool. Maybe one can even look for spreads that totally suck with the mean reversion approach and flip it! I haven't found any other easy accessible tools for this and a little creativity can even help you find some super trending spreads!

There are some stocks that lend themselves well to the MV approach tho. Check out AGNC V BMR

http://www.pairtradinglab.com/backtests/TuEP2VWQ_HCx1cHN

3.5 PF with 80% winners :)

Shan I use to sell options for years. I can show you 98% winners. It's the one bad trade I am worried about. That is why co-integration is such bullshit.

I don't want to find trending pairs either! I want to find stuff you can't back test for!!!!!! That is where ACD comes in. Anything that shows up on a backtest for you will show up at every quant trading desk in the country. I want to find the moves "before" they happen.

I'm not knocking mean reversion for other people. God knows Don Bright and I have been in the octagon debating this shit for years. It's not a matter of what works, but what longevity it has.
 
Man this market is not letting off...had a bet that BAC would close above 5.50 by the end of the month, looks like we may get below that print relatively soon!
 
Look at the TLT. If we close above today's OR we'll have a 9 on the number line... for December alone, (not the 30 day I typically use)!
 
Back
Top