i hear you, man. this thread seems generally for the folks that are trading using pairtradefinder.com or something similar that will alert you to a correlated pair's divergence without much concern for a bias based on fundamentals. all short term, no focus on intraday production and definitely not layering, etc. however, i have gotten a few good pair swings/scalps from this site over the past few months so wanted to throw something out there that seems to go with their trading style, i.e. short term overextension of correlated stocks' differential. $2 layers seems good for .7 ATI - RTI. what is the 20 day ATR, about 80 cents? ahh, whatevuh. best of luck with your trades, bro
Great discussions going on guys, enjoying reading it all especially other pair trade idea's, nice to see others perspectives. New trade Long AHL @ 25.24 Short PTP @ 33.47
Generally speaking, I try to put pairs on with strong correlation and in the same industry. If not in the same industry, at least the same or similar sectors. In regards to WHR/MAS, obviously that didn't apply. Coincidence with that correlation? Quite possibly - but I would speculate it is no different than taking a sector long vs a sector short. Example, perhaps one is long the XLK against a short in the XLY. 85% correlation...is it coincidence? Don't know for sure, but given that stocks from different groups generally move in tandem, until the relationship breaks down I think it can be playable. With WY/MAS, 69% correlation even tho in same industry. This probably speaks to the ongoing, changing fundamentals, which is something I always consider in a trade, especially when I might be holding it for more than a day or two. Fundamentals DO matter.
ATI/RTI looks worth a shot to me. It has been trending, but looks short term overdone. Which, btw, there are good pairs to be found that are trending. One example might be long LM, short TROW. Still not back to parity and slightly less than -1.0 Std deviation. On pullbacks it makes sense to me to see 0.70-.74 in the ratio. Thanks for the idea, and trade 'em well!
One of the problems with PTF is that signals are generated with incorrect prices. For example, today I got a signal in the RY/TD pair with a buy of 46.11 in RY. As far as I can tell,looking at real time time and sales, the low in RY was 46.85. In reviewing performance for the system in general, these kind of errors lead to a distorted upward bias which needs to be looked at.
I agree. But the problem is that there is no realistic way to trade this pair at those prices. In reviewing the performance of the system this"trade" among others will show better results than reality.
You get what you pay for. Choose the IQFeed datasource, and alot of these issues dissappear. I did a test the other day between the two feeds, backtesting 'Business Services', and I ended up with 20% more 'good looking' pairs that were bogus from the Yahoo feed than from IQfeed. 60 bucks a month is not a bad price to pay to remove these headaches imo. Adrian
I have Trade Station and went back and looked at RY. It traded as low as 45.82 after market hours so Yahoo showing 46.11 at 9:30 next mornign for last trade is not out of line.
Even if there was a trade of 100 shares at that price overnight, it still makes you wonder how realistic it is to match the performance of these simulated trades in real time trading.