If you take two random variables that are correlated and create any composition out them the resulting variable will be still as random but with (possibly) lower or higher frequency of oscillation and lower or higher standard deviation. Pair trading is an illusion. Take any stock, future, etc. and play half of your capital you'd normally play with and reserve the other half for a repair strategy, you'll come up with virtually the same results. Just my $0.02. (2 years of in-depth studies and testing) Cheers.
When combining two correlated, but random processes you must have a better information set to make a decision than having just one independent stock. We teach many pair strategies to trade over different time periods, some where you don't actually execute both sides of the pair but use the correlated stock for information. You can make the argument that any educational material in the trading world doesn't work because if it did you would see someone 'go out and borrow millions to trade it themselves'. We do trade these strategies ourselves, but we also understand that no matter how much I trade large cap blue chip stocks I probably wont affect the price of them. This means that I can help educate people with only have a slight change to my personal edge. It is another fashion for us to leverage our edge trading and create multiple income streams. Cheers, Darren
Let's talk about this. Imagine you have 2 stocks that are correlated 100% - its like having one stock twice isn't it? So the spread does not exist. Now, let's have two stocks with the correlation of 90%. Yes, there is 10% more information in this pair but your profit potentials are significantly lower as well. So, having stocks that are correlated too much will reduce your profit potential and does not add more information for your decision making process. Trading stocks that are correlated too little makes no sense because in this case you have no difference between one, two or more stock behaviors (random). One would argue then saying: - "There must be an optimum correlation level in the pair trading that would leave enough spread for the profit taking on one hand and would have a limited risk due to the added information on another hand". The answer is NO! Risk/Reward ratio is always worse in a pair than it is in each individual stock. So, why are people still talking about pair trading? Because of optics, perception and illusion. Because people "teach" pair trading and proclaim it being safe. Because general public is dumb, naive and is not sophisticated. Because someone can sell the idea of pair trading rather than trading it. Which, I agree, is a lot safer way to make money. I'm a scientist. I have been a scientist for over 30 years. And it hearts me when people selling bullshit to uneducated public.
Hello: I am not trading during the lunch hour and this comment caught my attention. I think I understand the idea you are putting forth, but I suggest that there may be fundamental information that is not captured in classic pairs trading strategies. For instance, the "classic" pairs trade is the Royal Dutch Petroleum/Shell Transport trade. Both entities part of the same conglomerate. Actually Shell Transport has a historically higher corellation with BP (British Petroleum) than it does with its own sister company Royal Dutch Petroleum. In the past decade or so, the spread between Royal Dutch and BP has widened significantly about a dozen times, always falling back into a range. The strategy has been to wait for Shell's share price to approach BP, then bet that the spread will increase by going long Shell and shorting BP. Like most anomolies, if there were no value in the spread, speculators wouldn't touch it. I can think of plenty of other examples including my favorite this year which is playing sectors against the larger market. I am not a scientist, but I appreciate the difference between logical thought (theory) and practical application of fundamentals as expressed by market history. Lefty.
Take your example (Shell/BP). Can you plot a synthetic stock (BP - Shell) or (Shell - BP)? What you will see is a stock with less volatility! That is it! Nothing else. But you could find hundreds of stocks with the same volatility that as your synthetic stock. believe me, the result will be the same. Unfortunately, your perception got in your way, but if you read carefully your statement you'll realize that you suggested a directional trade for a stock that has less volatility than Shell or BP, that is why you went for it. No other reason. Just think about it.
Actually it doesn't require much additional thought. You are correct. What I am suggesting is that lower volatility is not a problem. In fact I am looking for it. The volatility exists because the spread widens. It is reduced when the spread closes. I am just suggesting that a tidal process is occuring and the description of it MAY prevent some who think in a traditional way from profiting. Of course there may be something else that I am missing. It wouldn't be the first time. By the way, I have a pretty good history with this spread. Lefty Edit: I hope I appreciate what you are pointing out using a synthetic stock example. I think that I understand it. Unfortunately I have the feeling that I would not have had the same success making a directional bet on one of these entities that I have had betting on the spread itself. It seems to me that the reduced volatility is the "price" one pays for obtaining a higher probability of profit.
Correlation, as often experienced in the market, has a temporal lag to it. Often when you run macro analysis you find things to be highly correlated in the same time period, yet on the micro scale you will find that one stock is the leader in any movement. This temporal lag is in most cases the result of pair traders, and could even be considered a self fulfilling prophecy of pair trading. Either way, it still presents one way to capture micro anomalies in the market. Often the market has been characterized as a random walk with a drift. A rally in the market does not mean that a pull back will follow, nor does a pull back mean a rally will follow. The best predictor of the future price movement in the market is simply the drift. Within a macro view of pairs one can argue the existence of a mean reverting property. This has shown up in the past and is even supported by theory. In a capitalist society the technology and methods of market leaders are often mimicked by other companies. This should cause one stock being the market leader with the other stocks catching up to it as it adopts the methods and technology. On the simplest level, if you are trading a company like MWD, I fail to see how watching and gaining information from MER, GS and LEH is going to be a detriment. More information is always a positive. Cheers, Darren PS MAESTRO... great discussion... nice to think about things!
Lefty, I believe that you had a success with this particular spread. But you would have the same success with any stock that has volatility equal to the volatility of this spread. Except you paid lots of money to get this volatility. You had to hold two stocks with higher volatility to profit from the resulting stock with low volatility. Your capital use is 4 times less efficient as in case of holding only one stock with low volatility. So, your returns could be much better! I would like to make this statement again: THERE IS NO ADVANTAGE IN PAIR TRADING! I'll take anyone on it!
50/50 games with huge volatility don't make traders money... the house always gets its cut... lol. Volatility doesn't matter for someone in my position where I get amazing leverage through a prop trading firm. What matters for me is predictability. This is how I make money. So if my pair has the same volatility (or even less) then the underlying stock, but has more predictability I am fine with that. If I know with certainty that something will move 0.01, I will find a way to get my commission below 0.005 and trade it with as much size as I can. The predictability of the 0.01 allows me to make money. I don't think that the argument of equal volatility discredits pair trading in any way. Cheers, Darren