Hey! I'm also following "Son of Something Very Simplistic" at the same time so I have a bit of an excuse.
Yep, it would have worked out well...but of course I wasn't in it, heh. More proof someone up there hates me: in each of the days I have been doing NQ, I missed 2 or 3 of the first trades while waiting for the first to conclude. In each day, all the ones I missed were winners. (I suppose the case could be made its my own fault for trading while underfunded, but it feels better to blame some mysterious being) zorak
That's right, Quah. But I don't think you made a mistake because Db had the same data. I think it was just the difference in data providers. SSVS. Selfishly, I hope you share it with us. If you never posted again I certainly would understand that too. Oh boy oh boy! I'm taking the kids to grandma's house so I can go to my office this afternoon. See you guys in a few. Maaan, zorak, we just talked about that! That's why I say, when I trade it for real, I will let you guys know so you can be sure to trade the other contract for winners all day.
This shouldn't matter. The first consideration is whether or not you can trust whatever strategy and tactics you're employing. Whether you take each and every trade is not critical, though if you don't take a big chunk of them, you may be using entirely too much discretion. As for waiting for the first to conclude, at least consider offsetting the early trades so that you don't have to take on extra contracts. --Db
I disagree that not taking every trade is not critical. In system trading, every trade is critical, because you do not know which ones will be winners or losers. There is a lot written on this very topic. A common mindset is for people to stop trading a system after a few losers and wait until the system starts winning again before jumping back in. The problem is that many times it turns out they stop just as the system begins to win again, and get back in just before it begins to lose again. Thus they have absorbed all the losses and none of the gains. So when asked, they can and do say, "That system doesn't work, all I did was lose money." Yes, because they did not take every trade. If in the application of a new set of rules some of the orginal method's trades are skipped, then a new system has been developed. But every signalled trade must be taken. That goes for not being able to add one if called for. Each trade in this method is a separate entity. A minimum of 11 separate and unrelated trades occur each day. Each must be taken.
Here's a simple idea that I just thought of: Look at the direction of the previous 2 minute bar or even the last 3 minute bar as a signal to go long or short. DraXon
Taking every trade is critical only if one is trying to match someone else's results. Even Quah doesn't take every trade, so why should anybody else be compelled to do so? As for each trade being a separate entity, that is true as far as data collection and paper-trading. In practice, however, you are limiting the "system" to those who can afford to trade 6-8 contracts simultaneously. I suspect that such a condition would eliminate most of the people reading this thread. --Db
Actually, somebody already thought of that, but I don't know if anybody has followed up on it. Would you be willing to test it? --Db
The system produces its data and thus its profits based on each trade. The goal of any system trader is to match the system. Not taking every trade is already well documented as disaster for any system. I don't know where you're getting the 6-8 contracts number though. Unless for example the market opens and trends higher continually thru the first 6-8 trades which takes you all the way to 1059 without any pullback. I suppose it could happen, but on a 1 minute chart? It hasn't happened yet, and realistically, doesn't seem likely to happen.
Everybody is trying to follow Quah exactly, yet Quah doesn't always follow his own system. I see some problems there. Simple. If you're trading both the ES and NQ, which is what the "system" does, there're two. If the 0933 trade doesn't resolve itself by 0935, there're another two. If neither of those have resolved themselves by 0938, there're another two. Don't be so quick to etch all this in concrete. The data collection is less than three weeks old. --Db