Just a poll -- Assuming you routinely found such pairs to trade. How much of a capital base would you require to say, average $200/day? Informal guess -- I know it depends on random variables, but a guess with regard to the current market would be nice to poll.
I guess you mean currency trading. But for stock, the idea of "must take money home every day" is wrong. Think about the ebb tide and flood tide that didn't occur every hour during a day. Opportunities do not occur every day during a year.
I didn't say must-make-money-everyday, I said average, meaning some days will be up, some days will be down, but the end result is an average of $200. Though, your point about opportunities not occurring is definitely valid -- I'm asking, based on people's experience with opportunities occurring and not occurring, what it takes to make $200/day.
how much are you willing to risk every day to make that $200 per day? my guess would be have $40,000 on hand if you want to play it safe. Maybe $20,000 as a low end figure but probably need more if you get any substantial draw down.
I'll assume you mean the real way, which has no relation to currencies. <a href="http://www.collective2.com/cgi-perl/systems.mpl?want=p&systemid=25716110"> <img src="http://www.collective2.com/cgi-perl/badge.mpl?systemid=25716110"/> </a> Hmm, being the best pair trader in the world: I've made $58,855 on $100k in the last 750 days. This would be $78.47/day, but, it's an unfair comparison when you consider that we've only been open for about 500 days. So, under that assumption $117.71/day. I would say starting capital is $200/117.71*(100,000)=$169909 to start. You could just save time and subscribe to me, b/c it takes years to get a system that works that well. It's not really that the returns are exciting, but that the system is even possible that makes it best. Most people developing systems discard valuable ideas b/c they don't like certain risk characteristics, when it is actually producing "viable, tradeable" systems consistent with your style that is the whole objective of rule based trading. As a side note, using the systems that "scan for pairs" doesn't work. It's really only in the last several years with double leveraged and double leveraged inverse ETF's that this works consistently. Before the pairs would only last a couple months, now you have an eternally negatively correlated pair to arb the hell out of.
You are talking unleveraged, right? If I get 10x leverage, we're talking about 17k. I will definitely check out your work.
crap, why am I even trading. I should just give money to someone on C2. i'd rather take 20k and blow it in the market and then start over. at least you would learn something in the process. If you lose it on C2 then all you learned was not to rely on anyone else to do your trading which you could probably have guessed in the first place.
I'm assuming 1.92:1 leverage, but with the double leverage, we're talking 4:1, but certainly viable with 1.92:1 leverage. If you don't learn anything else from my crass self-promotion, just do some experimenting yourself, and for anyone else interested in pairs trading. Plug in QID and QLD, and make that your watchlist to start developing. It is not that pair trading produces small returns, more than it is that finding the pair is the reason for the small returns. You can do it on SDS, SSO QID QLD MVV MZZ And I guarantee there's shell programs to start with that can be your guide into this. Seriously, pair trading isn't technical analysis. Done correctly, and accounting for volatility as you would in an option pricing model, it's very effective, and that system is proof enough for me. Anyway, if you don't subscribe, well, I guess it's no skin off my back, but experimenting by yourself with a shell program that's already out there might be something to try. And that's not just for the OP, but for anyone interested in real quantitative analysis. Ex-Post Edit: There's one further than this, too. Using options gives a bit of a cushion through the time decay component of options pricing. Doing it this way, and certainly leveraging to 10:1 and beyond is possible, but don't forget, I drewdown 33%, so at 10:1 drawdown is greater than 100%, so keep your expectations tempered. 4:1 is absolutely the max anyone should trade, but only after years of trading. It's laughable people go straight to forex to trade at 100:1 or 20:1 or whatever stupid amount beyond 5:1 that will just nuke their account.