I wouldn't trust ANY of those results. Your positive results are most likely due to several bad ticks not visible when performing this type of analysis. Positively correlated pairs above .90 are hell to trade imo.
You have a look at the tradeslist, then. I see no bad ticks, and it's hard to screw up the open price. In fact, I don't recall any problems with open or closing prices with Fidelity, just highs and lows. Position Symbol Shares Entry Date Entry Price Exit Date Exit Price % Change Net Profit Bars Held Long IVV 484 7/2/2004 113.22 7/20/2004 110.54 -2.4 -1,313.12 11 Long SPY 453 12/2/2004 118.16 12/3/2004 119.1 0.77 409.82 1 Long SPY 450 12/22/2004 119.6 12/23/2004 120.38 0.62 334.95 1 Long IVV 428 11/29/2005 126.67 12/5/2005 126.72 0.01 5.4 4 Long SPY 435 1/4/2006 125.1 1/5/2006 126.86 1.38 749.6 1 Long SPY 426 4/19/2006 128.95 4/20/2006 130.72 1.34 738.02 1 Long SPY 450 6/16/2006 123.92 6/19/2006 125.19 1 555.5 1 Long SPY 430 9/13/2006 130.56 9/14/2006 131.64 0.8 448.4 1 Long SPY 424 10/5/2006 133.23 10/6/2006 134.92 1.24 700.56 1 Long IVV 401 1/26/2007 142.9 2/16/2007 145.74 1.96 1,122.84 15 Long SPY 404 4/13/2007 143.74 4/18/2007 146.97 2.22 1,288.92 3 Long SPY 391 5/31/2007 151.46 6/1/2007 153.66 1.43 844.2 1 Long SPY 407 6/28/2007 148.13 6/29/2007 150.38 1.49 899.75 1 Long SPY 478 3/12/2008 130.72 3/13/2008 132.73 1.51 944.78 1 Long SPY 449 5/2/2008 138.39 5/5/2008 142.33 2.82 1,753.06 1 Long SPY 474 6/16/2008 135.17 6/17/2008 135.55 0.26 164.12 1 Long IVV 498 6/27/2008 128.55 7/8/2008 125.2 -2.63 -1,684.30 6 Long SPY 498 8/6/2008 126.02 8/7/2008 128.02 1.56 980 1 Long SPY 498 8/11/2008 126.58 8/12/2008 129.47 2.26 1,423.22 1 Long SPY 519 9/9/2008 128.04 9/10/2008 125.1 -2.32 -1,544.42 1 Long IVV 524 9/16/2008 117.4 9/17/2008 119.8 2.02 1,241.60 1 Long SPY 780 4/13/2009 84.67 4/14/2009 84.92 0.27 179 1 Long IVV 771 4/21/2009 83.03 4/22/2009 84.36 1.57 1,008.04 1 Long SPY 765 4/30/2009 86.52 5/1/2009 88.55 2.32 1,536.95 1 Long SPY 752 5/27/2009 88.36 5/28/2009 91.44 3.46 2,300.16 1
no, you don't. you just need a blip of deviation and the infrastructure to catch it. just because your margins are only viable down to 30, doesn't mean others aren't down to ms. better to not speak/think in absolutes, or in subjective bubbles relative to your abilities/timeframe... you will almost always limit yourself if you do.
It's obvious you don't know what you're talking about, but it's ok, I've yet to hear someone from Wall Street explain how to do a pairs trade correctly, and that includes on this board, too, so don't take it personally. You missed the point, and I'm sure you're not the only one. The premise of a pairs trade is that there is a sufficient deviation from the respective value that allows you to arbitrage the mispricing. The problem at any level below daily is that it doesn't make any sense to arbitrage out just a penny of inefficiency for obvious reasons. The daily time frame allows prices to move further away from the correct value, and you don't have to worry about commissions or transaction costs if you have a sufficiently large mispricing. Does it make sense to you to pay 99.99 if you know the price should be 100? No, it doesn't, and if that wasn't your answer, you have far greater things to be worried about. That's why millisecond strategies don't work. Generally, there's two securities perfectly correlated either positively or negatively and if they deviate from their value you take a position. The problem as the op is running into is the transaction costs and other incidental costs to running the strategy. The strategy can be corrected at the daily level for all the reasons I've spoken about previously because at the intraday level there is not a sufficiently large mispricing to take advantage of the arbitrage opportunity. The ratio will always be the same no matter if we are at the minute, second, tick or daily level. The point is, it doesn't make any sense to put the position on when there's an insignificant amount of value to pull from the market, and that's the OP's main problem.
Bw, I think that's good news that a daily timeframe is more profitable for all the reasons stated. However, what about all the trading that Ed Thorp's, aka mr beat the dealer" firm engages in? Isn't that a stat arb operation on an intraday timeframe. I understand he has been amazingly profitable. best, J-Law
Pairs trading is all I really do anymore. I'm willing to acknowledge there's other strategies out there that work, but the thread is about pair trading in US equities, and the advice I've given only applies in that regard. I also have never heard of Ed Thorp. The only real way to beat a dealer is through flash trading.
Thorp was an M.I.T mathematics professor whom devised a system to beat the blackjack dealer back in the 60's. He later went on to Wall street where he made billions in Convertible & stat arb ......& pairs trading. http://en.wikipedia.org/wiki/Ed_Thorp Please look at the Reference section under Stat Arb page for articles on his trading. As a trader, but more so as a pairs trader you will enjoy the read. http://en.wikipedia.org/wiki/Stat_arb
At least 34 ways to skin a cat, venerable Brother Bowlinsky and you may not have tried all the different slicing and dicing implements out there Regarding trading pairs of US equities, just because you may not know anyone that is making money scalping or swinging intraday, doesn't mean they don't exist. IMO, IVV:SPY would be difficult intraday due to IVV's relatively low liquidity as you are going for pennies...slippage, etc... Not my game, but something like CMCSA:CMCSK or any stock and its rare uber-liquid preferred may work better. Check out some pairs like APC:OXY, HCP:HCN or even USB:WFC and see if you don't see intraday opportunities (as well as multiday swings, of course). I do both intraday and hold some long term, if need be but know people that are out of everything at the end of the day every day and are consistently profitable, month in and month out. If you think it is better to stick with only multi-day plays, entirely up to you but know that it is a potentially profitable technique. On Ed Thorp, besides the articles cited above, check out the book "Fortune's Formula," a good read I am thinking you would get into. Angus
That dude's fucking rich. Quote: "In May 1998 Thorp reported that his personal investments yielded an annualized 20 percent rate of return averaged over 28.5 years. [16] " Well, I'm about at 18.5% annualized, so I'm sure I'll be able to keep up. Every 10k is about 1.5 mil, so at the very least, he's a millionaire, but more likely to be a multimillionaire.