pairs trading

None. First I would recommend that you try to trade as many as you can for diversification.

Second I would make sure that returns on the 5 pairs I choose to trade are not correlated. This too would improve diversification. That is the biggest.


After that I might look into risk/reward as a distant second measure.
 
Quote from TradeNYSE:

It's a hypothetical question, trying to discern what you consider to be most important factor when picking pairs to trade. Assume you had 30 pairs, traded them all the same way, and the statistics for each (net profit, avg profit, etc...), and no fundamental background for the pairs, just the profit numbers. How would you pick the 5 best with that data?

I can already say that total net profit would not be a great idea b/c you could have one day (like a buyout) that you may have been in the right side of (or wrong side of) which would skew the net profit in one way or another.
More...

Okay, but in the event of a big news engine it wouldn't really matter that much because after the event i'm no longer to trade it (whether i took a beating or was paid off). To be more blunt I will trade as many pairs as I can at a time as long as i'm profitable on each one. Like the last poster in some ways it's trial and error to see which pairs suit your particular style best.

I have specific filters for screening new candidates and there's been times the ones that would look great according to my filters are not, and vice-versa. One thing about pairs is that each is unique and you have to be flexible in trading each one individually to some extent.
 
What if you had one strategy that you used for pairs, fundamentals aside (for arguments sake let's say they are all similar in that sense), but you had a list of 5000 pairs. You can't trade them all, so how do you determine which you would trade given profitability statistics?
 
Quote from TradeNYSE:

What if you had one strategy that you used for pairs, fundamentals aside (for arguments sake let's say they are all similar in that sense), but you had a list of 5000 pairs. You can't trade them all, so how do you determine which you would trade given profitability statistics?
More...

Well i would probably look at the pairs that had the greatest frequency of success IF i was applying the same criteria for entering/exiting trades. That's a big if, because there's some pairs I have that I trade more frequently throughout a single trading than others I swing trade over a few days. Now, I try to keep the targets and capital/risk as equal as possible so that two different profitable trades on pairs is generating a similar risk/reward. I'm probably not the best pairs trader to be speaking with on this subject because i apply a blend of strategies for my methodology.

There's a definite art form to this like all trading and once you start trading you soon find out which ones you like and which you don't.
 
Quote from teun:

Why? 50% of the times they are profitable for the pairs trader.
More...

Have you ever thought about the other 50% chance that a takeover will go against you?
 
Quote from reg:

Have you ever thought about the other 50% chance that a takeover will go against you?
More...

Sure, it happens, but since we tend to not be short the smaller (thus more likely takeover candidate), we have a pretty good record overall.

Don
 
Quote from Don Bright:

Sure, it happens, but since we tend to not be short the smaller (thus more likely takeover candidate), we have a pretty good record overall.

Don
More...

I'm pretty sure that RIG was a larger cap co than DO at the time of the takeover.

HLT, one of the largest casino co in the world before being aquired.

The reality is simple. You must be well diversified if you even want to try out pairs (20 minimum). 20 pairs is 40 positions. 500 shares each puts you at around 20,000 shares carry overnight. If your average share price is $35.00 than you are holding about $700,000 worth of equity.

What is the cost of holding $700,000 worth of stock? Let's say the first $100,000 is free. The next $600,000 will cost you around 6-7% averaged in minimum.

Okay, now you're paying a minimum of 6% on $600,000/year in interest. You now have a $36,000 cost to overcome in trading your pairs.

Pairs trading is deceptively costly for a small "prop trader" type of account. The truth is, you may not blow up your account, but it won't be easy to cover your costs.

And don't fall for that "reversion to the mean" crap. Ask the guy who went long GE and short HON when it "should" have reverted to the mean. It's only gone from -4 to -20. Try to keep adding on to that one and see how long your account lasts.

Good Luck.
 
I can shed some light on the carrying costs with Bright. There is a 1.75% differential between long stock (pay) interest, and short stock (receive) interest. 12 times equity (Say $360,000 with $30,000 equity) adds another 1% to the mix, so that would 2.75% per year, pretty cheap overall by my calculations.

To go another 6 times equity would be 4% per year. We go up to 30 times before addressing the issue with the trader.

I agree with a lot of what you say, and so we do encourage the newer trader to keep things small (of course), and do a lot of "crutch" pairs trading where they often times do not even end up with both sides of the pair on, preferring to "lean" on the other side only when their initial trade goes the wrong way.

Trading the noise of the market, intraday, is a must IMO for successful pairs trading as well.

As I've said before, just one strategy of many that our traders decide to trade. Keep in mind that each trader is running his/her own trading business within our framework, not "following orders" or anything.

All the best,

Don
 
Don,

You do take a lot of "flame throwing" on this forum, but whatever they say about you, you do lay your cards on the table.

Thanks for the explanation of the charges.

Keep wearing that flame retardent vest.

LOL
 
You must log in or register to reply here.
Back
Top