You initiate 2 baskets of FX pairs, let's say 5 contracts long and 5 contracts short. (or stocks/indices) All of these instruments have correlation of 0.8 and above. After a while, (1) what do you do if the net P/L is negative, Scale in when net P/L down by 1%, 3%, 5%, 8% (any other suggestion)? how much (in %proportions) of no. of contracts do you scale in on winning side, how much on losing side? I propose 0.8 contracts for every winning instrument, 0.5 for every losing instrument? Any other suggestions? (2) if net P/L is positive, when/at how much %gain do you scale in/out? scale in: 0.5 contract on both winning and losing? scale out/profit taking: 50%, 30%, 20% of total number of both winning and losing contracts? (3) or simply add equal amounts on both sides when P/L is up or down at 1%,3%,5%,8%? Is there a better system to scale in and out? Please share your analysis. how do you apply this concept to hedging spreads with spreads
i propose you find a different direction to think in unless you just have to exhaust this current idea in order to move on. concentrate on finding a better method, man.. really. your like someone who wants to make money by conserving your losses, you need to make some money to have something to conserve. mb
Yes theoretically, its called a universal portfolio. Read up on it, though there isn't any simple explanation available. Transactions costs add up. Thomas Cover- http://www.stanford.edu/dept/news/news/2000/april12/universal-412.html
Check out http://getfolio.com/. AFAIK he only goes on the long side. He's also big on position sizing and money management. Of course using this method you will buy the Enrons and Worldcoms, but overall it seem to be profitable. I tried it around 1999-2000. I bailed when the market tanked (and yes I did buy some Worldcom). I had put the stocks into a Yahoo portfolio. When I checked it recently it was actually showing a profit. Too bad I sold the stocks in real life. And if (when) the market crashes and doesn't come back for 20 years you will lose.