I am interested in whether pairs trading is a directional strategy or not. What are others experiences with this? I'm finding that ever since the S & P turned downwards in mid June that it has become harder to find profitable pair trades, and I keep getting stopped out of positions. In the experiences and opinions of others here, is there a tendency for mean reversion to occur less when markets are turning down?
i was also wondering about this. how does pairs trading fit into market structure? if you have both stocks that are more or less in a range, the whole sector has no direction, is that the best setting for a pairs trade? i mean, what have relevant traders noticed about the behavior of ratios, pairs in a directional market, falling market etc? i guess the recent crisis is probably not a good example, because that's what it is, a crisis, a black swan, so to speak . .. everything is out of whack but if, in an otherwise calm market, a sector is showing a strong momentum in either direction, are there any analomies to be aware of etc?
Hi TT, In my experience I see the same thing. It seems like opening positions in a bullish period provides profits when the market turns bearish, and vice versa. I generally have about 10 open positions at any one time, and a month ago I did well until the bullish period started. Now it seems my portfolio tends to do better on the bullish days. I'd like to investigate further to see if I can somehow neutralise this experience, without over complicating things - the answer may be to just continue to open positons through all market movements to equalise things out. This I mostly do anyhow. Each time I check my portfolio has beta equivalence for longs and shorts, though I don't do this through any intention, and I don't bother to check beta's prior to choosing my pairs. I'm still seing heaps of opportunities, and struggle not to open too many positions. Perhaps I'm not being picky enough... Adrian
A few things that might help you guys in pairs... When looking at a chart of pair ratios, determine the time frame for reversion. I have found the 13 day average is the best. When a pair deviates from the 13 day mean, understand that it will always come back to that mean in time. Sometimes in a day, sometimes in a month. Be patient. Also look at the long term chart of the ratio. Is it flat and trading up and down from the mean, or is the pair ratio trending in a direction. If it is going up than the best time to enter a pair is to trade the direction of the longer term average. For example, the 50 day mean is heading higher, but the pair ratio just dipped 2 stdev's from the 13 day mean. Go long the pair. The ratio will find support along the trend line. If the pair deviates to the upside, be careful and more cognizcant of the fact that the trend is working against you so shorting the pair is not a great idea. It can still be done, but unless the longer term trend turns, your profit potential is less. The point i guess is use shorter time frames. Trying to catch a reversal using 200 day averages is just too tough and you might confuse the deviation as a point of reversal, when the shorter term moving average is still going against you. Hope this helps.
The long MRX, short PMTI has been taken off the sheets. In 8 trading days the trade yielded $3.31 in the differential, or +9.96% in the capital employed. Does it have further to go? Probably...but given the relatively low correlation for 2 stocks in the same industry, I am comfortable with taking the gain and moving along. The long HNZ, short CL is virtually flat. It would be down a bit over 0.5%, except that I was able to collect the $.42 dividend. I still have the trade on and think it works nicely. HNZ went out today at $35.34 and CL closed at $70.19. Long KNDL, short PRXL has come all the way back in. It currently sits at -2.2 std deviations from the mean. Given the fundamentals, valuation, and growth profiles for the companies, this appears like it's time to up the size in the trade again. KNDL closed at $11.90 and PRXL went out at $13.15 today, giving us a differential of -$1.25. I did add back to the position today. Be aware both will be presenting at the Wachovia conference on Friday. One that I may be a touch early on, but I like just because the fundamentals and valuation are so compelling for the former is a long in SUN, short in TSO. Currently at -1.5 std deviation. Normally I would volatility adjust this to be short more TSO against the SUN, but given this month's performance in TSO and it now sitting in an oversold condition, I am choosing to use a 1:1 ratio, which obviously gives me a long bias. Lastly, I see the discussion on beta adjusting and couldn't agree more with those who have said beta is meaningless in pairs trading. Yes, I volatility adjust the pairs, but beta is measured against the S&P 500 and is worthless relative to what one is hoping to accomplish with their pairs trading. Beta is only useful in the context of a larger portfolio with many positions; as well as for opening only orders. That's all I got...trade your pairs well, folks!
I know its probably been posted before... but how large of a portfolio is being used and how much leverage. And how much per each leg of the position is used.. Dont feel like going through over a thousand posts..lol thanks Nick
Anyone engage in AMEX stocks? I have avoided that exchange for years on account of getting horrible fills. Anyone have any good experiences lately? Looking at GIFI/PDC, however PDC is an AMEX stock. Adrian
Dividend Play: stocks on the left have dividends due i.e. RSG, AXP, GD, MON, BMY, JPM. Two plays either collect the dividend or benefit from the surge leading to the dividend. RSG/WMI 100X100 (06/29) AXP/WU 100X100 (06/30) GD/LMT 300X200 (06/30) MON/SYT 200X300 (06/30) BMY/PFE 100X100 (07/01) JPM/WFC 100X100 (07/02)
I know some people who have had decent luck with this strategy. The logic is sound. There seems to be an edge. Good luck.