this is a very intersting post, but I am a bit confused about something. Let's say that I BUY an es ym spread at 40 and it goes against me and becomes 90. If I simply KEEP IT, it will probably come back to 40 or 30. I realize this may take days, but taking a loss, means that you have to then find a spread that works. OR attempt to predict direction of the market by "lifting a leg" and then selling the other side (hopefully) for a profit. OR AM I REALLY, REALLY OFF BASE? Second question: I pick a trade - say a 5 15 ema crossover, instead of selling for a loss, I make it into a spread. Could this help a novice/bad trader who can pick winners, but gets burned/emotional by the bigger loss?
What I mean by that last comment is to turn it into a spread when I have a very SMALL loss. Therefore, I can "regroup" and look to leg out OR have the spread move in my favor and then close it.
I think there are two approaches that can work well.. 1. Always enter both sides of the spread, NEVER hold just one position or the other. 2. Enter one side of the spread, if it moves in your favor sell it for a profit. If it moves against you, take the other position in the spread. If the spread then moves in your favor, sell for a profit. If it goes against you, consider adding to the spread. The idea here is just to maximize your percentage of winning trades. Most important here is low transaction costs (especially for very short term trading), and don't fall into the trap believing that this is "risk-free" trading - that is dangerous.
Have you considered the longer time frame with this spread? It seems you could do just as well with less transactions but with higher profits per trade. Also take a look at using bollinger bands, buy and sell as the rubber band stretches... just a thought.
Yes, although I don't have access to charting that I would like to track this. I previously used Realtick for this, but let my subscription expire recently.
My experience with intermarket spreads (long one market, short another market) is that they often trend better than the underlying contracts... if one is trading the spread trend then this is beneficial, particularly in the context of choppy price action on the underlying contracts... however, if one is trading the intermarket spread under an assumption of mean reversion to some kind of fixed ratio, you can often get hurt more on the intermarket spread than on directionally trading either of the markets in isolation... My experience suggests that intermarket spreads go through phases of both oscillation and trend, in much the same way as the underlying contracts... but there can still be an advantage on intermarket spreads (relative to trading outright) if the incremental commissions costs are offset by significantly lower margin requirements on a spread combination (resulting in greater leverage)... In my humble opinion, the only spreads where there is consistent scope to play the mean reversion game is either on intramarket spreads (often known as calendar spreads) or on inter-exchange spreads (similar markets, different exchanges)... and not spreads between 2 intermarket futures (such as S&P and DJIA)... just my 2 cents, and a highly debatable area...
by no "consistent scope" do you mean on a r/r basis? ie spread between nq and es is likely to revert to mean, but playing on that doesn't make much sense sense from r/r perspective? or do you mean that mean reversion on such a spread isn't likely, or doesn't occur enough? or, perhaps, that such a reversion is only an 'illusion' and that there's no good reason to believe the spread will ever revert? i don't mean to grill you, but i'm just curious, cos i would have thought that the sp, naz and dow would have been ideal candidates for such plays..
I am talking from the perspective of preferring the mean reversion game... so my comments refer to "frequency of occurence"... in the mean reversion spread game, one is ideally looking for most of the spread behavior to occur within fairly clearly demarcated channels, as opposed to longish periods of spread trend combined with longish periods of spread oscillation... again, this is just my 2 cents, based on personal observation... the spread game is full of all sorts of nuances and is a highly debatable area... Of course, you can also use spreads to create trends when the underlying contracts are choppy in a narrow range... this trick is great for trend trading intramarket spreads...
I think the fundamental problem to be solved in trading is determining whether a trend will continue. The trend just being the direction the price has been going in whatever previous time-frame. If it does, simply trade with it (use Moving Avgs). In a choppy market, trade against it (use BBands). The problem with "directional" trading is we never know for sure if the current trend will continue or reverse. In my opinion, the benefit of trading spreads of instruments that are correlated is only that we can bet with better reliability that the trend will not continue, so always fight the trend. If the pair is truly correlated (and this is our assumption to begin with), that trend will certainly not continue, so I don't follow the trend on these or any other spreads such as these, but scale in/out always fighting the trend.