say i have a good trading system for both the long and short sides of the market. the time frame is from a few days to a few months. i usually have anywhere from 5-20 positions at once. i'm constantly asking myself, "how many should be long and how many should be short?" i could go all long, all short, a long for every short, or some other method. i'm wanting to come up with a system in which the market itself tells me how many i should have on each side. i'm just wondering if anyone else has been through this and has any good ideas? thanks
Another interesting thread from Gekko! Good stuff! I've thought a lot about doing this type of trading (multiple positions 20-40, few-10days) myself. I haven't come up with a great, or an exact, answer on long/short weighting but I obviously it would be very dangerous to go all long or all short. I think it really depends on how well the stocks you are trading at the time correlate to broad market moves. Personally I think this type of trading might be better suited to small cap issues that aren't really affected by market moves. I'm looking at doing some very basic support/resistance stuff combined with relative strength vs the sector (not wilder's RSI - why the hell did he pick that name!). I'm quite the novice when it comes to multiday swings, so I don't really have a lot to say on how to use the market to "tell you" how to divide your longs and shorts. This could be a very interesting thread, I hope we get some input.
You could go long stocks in the strongest sectors and short ones in the weakest sectors and let the probability of trend continuation work in your favour. Lately you'd be carrying hmo's homebuilders and restaurants long and chips, telecom and utilities short. And you'd be cleaning up. You could also move in and out of spiders and qqq's as a hedge based on your net long/short stock position.
If you are trying to trade a modified hedged portfolio; we use 66% short / 34% long when we have a downward bias, 50% short / 50% long when we have a no bias, 34% short / 66% long when we have an upward bias, based on capital while trying to balance all positions. Only works in a portfolio with a lot of positions; i.e plus fifty. You have to stop out your losers before they run away and obviously gaps hurt the most. We use this with options but it would work with equities just as well.
metooxx, would you say that the bigger winners are a result of a favorable gap? and if so, what percentage of the profit on a successful trade do you find a favorable gap accounts for?
the key is finding the optimal combination of hedge vs directional bias so that you have protection when you are wrong but upside when right sometimes the optimal hedge is just getting out quick instead of trying to be ambidextrous to hedge effectively you need super low execution costs, lest you get gummed to death by fakie reversals that wind up flat i prepare to go either way each day and then let the market determine the mix, like a DJ who takes requests
You want to know everything? This type of trading only has a slight directional bias, you can use almost any traditional trend indicators; worst case you are trading a 50/50 hedge.
If you want to be long and short at the same time, it would be wise to stay away from stocks that follow the rest of the market. I've found that the cheaper the stock, the more likely it will move based on it's own fundamentals rather than what the overall market is doing. edit: this would solve the problem of ending up witha 50/50 hedge.