bdog,
nice post.
been working for that 65pct too,lol. been doing alot of system testing intraday. most cases i've seen so far and by no means am i done, just getting started actually, but most one indicator systems are losers, if yer taking buy and sells on the same indicator. i dunno if that's how things were backtested or not, nor do i care to enter pissing matches. i am also curious as to how to quantify the entire system into a backtestable approach. there are more uses than what can be tested for i think, its a little more dynamic when it comes to trendline breaks on indicators vs buy this level, sell that level. there are thresholds and pattern situations to consider as well. so yer not necessarily always gonna get an entry based on indicator values that can be slogged out by a machine. all in all, i think that doug's eod system and insights into technical analysis are useful. sometimes stocks don't abide by good systems. it happens. but its behavior analysis that provides the opportunity no matter what type of system you use, whether time or indicator based or both. i'd throw in some other questions regarding time measure as well, mainly, isn't that what the period settings are for? i understand where cube is coming from and i've been studying what i believe he's talking about which is fib time and fib price, pesavanto type stuff mainly i've been looking at, its new to me, so i in no way would challenge those types of trading techniques. although i have seen failures there already as well. i'm not looking at it as one better than the next, just another possible utility. i'm guessing its some sort of cycle/gann/fib type of thing. my personal style isn't really a buy and hold anyway, but i am enjoying the studies and they do help when a stock is "out there without a net".but then again, i'm not much of a bottom fisher either. so i guess when i get there, i'll hafta check if a 55 or 89 setting on an indicator acts like a cycle, or if i can determine the cycle and adjust the sto by half the cycle length as i thought some techniques indicate can be done. i have seen a system lately however that for the backtest data i have where even losing records were profitable. in this particular system, the win size is almost nearly double the loss size. when its had a winning record, its out performing buy and hold. i'm not trading it yet, but if the losses are kept in line, maybe its a start. maybe the tops can be better sold under a multiple sell/exit system. its currently not set up to do that and perhaps the longs can be gradually exited. so as far as the 50/50 scenario, if yer wins are 10-15pct profits, and yer stops loss is 5pct built into a system,well, a losing record would still be profitable.mathematically wouldn't that provide for 2x many losers b4 you'd be even? i stopped trading for several weeks becuz the system i was using wasn't working as well as it used to. mkts change, and i apparently didn't have anything on the burner to adjust to, as this was the first time its happened to me, so i lose some confidence in what i'm doing,miss a lot of nice trades in the process,even take a loss only to watch things recover, so its driven me back into the lab. for the time being, i trade light, if at all, until i can find a new system i'm confident in that meets my trading techniques. i don't have problems with the eod system, i'm still happy with it. maybe its just a matter of paying too close attention to every trade during the intraday time frame that has me stumped. others have told me "you used to hold things longer". so perhaps part of my technique was lost somewhere in wider intraday swings, where exits became too close. seems what you want out of a system is a simple entry point and then assess whats happening after. after the trade is entered, the tricky part starts, buying is easy, selling is another story. that's where the plan/money mgmt have to be established before the trade. and that has to be part of the system.consider you are wrong when you take the trade and determine your exit and stick to it. consider you are right, then how much should you take profits on? will one indicator, or perhaps even one set of indicators turning against you result in taking full profit, or holding onto half the shares. its a function of who you are in style as to how to determine all this. the fact is, even if i did come up with 65pct win/loss ratio, chances are you wouldn't trade it with the same results as i might. lots of factors can come into play to skew those results, as you know. one that has not been mentioned yet, say i give you this great system and you take 2 quick losses on it. you then get a third signal and hesitate while the losses weigh on yer mind, guess what, you don't take the trade, and thats the one that flies. we know over the last 2 yrs 30 m25's made it to the ndx. its a scan i came to heavily rely on. its a scan i've had many requests for. but in the downtrending mkt, most of those scan hits set tops. so one system is not gonna carry you thru your lifetime of trading. we also know that the m25s will normally kick out anywhere from 15-25 stocks on an avg nite. and i'll say this again, cuz i've done it here before. suppose we load all the scan hits. who cares if 10 fail and 10 succeed on the open. one of the purposes of the technical study is to provide yerself with opportunities every day. you now have 10 chances b4 the day is out to find a good trade or two out of the ones that are working. psychology in mkts can't be backtested for either.the closest we've come is the comkey.we know its fly or die. indicators say one thing, mkt perception maybe exaggerates them at times, ie a paste situation. the point being under that type of scan scenario, we can still assess properly our risk mgmt. does the risk allow for chasing? can you protect it? personally,i'm not buying every scan hit on the open in hopes of my system kicking out 65pct consistently...just show me the ones that did work, and i'll try and find my way into them, and that usually occurs early enough in the day to make proper choices. in all the time i've been doing this, i've never been short on opportunities. i may add more later, but i think your ideas are pretty much along these lines.
milesov |
