Systemized profit-taking?

..sorry, perhaps my Southern self-deprecatory humor got in the way here. To get the context straight, I consider an intraday trade in NQ to be mediocre if it makes $15 or less per contract after commish and slippage (trading with market orders). I am thrilled with a system that makes $30 or more per trade intraday (not all day, for which I would be impressed with $60). That is the context in which I find that five rules (earliest entry time, entry price, stop price, take profit price, and latest exit time) work. Having low aspirations, I am happy with a system which on average makes 1.5 NQ points. If it is not proprietary, may I ask what your context is?

Also, perhaps you have posted elsewhere, but I do not understand what you mean by failure points. I am visualizing breaking previous day highs/lows, floor pivots, strike prices on the related options, same day opens, etc. Is that what you mean? I do not mean to pry beyond the philosophical into the specifics. After all, we have paid handsomely with time and money to get the specifics! Best regards. - Mike
 
This is the thing. I'm preprared to accept scratch trades and losers, - I just consider them as the ones that didn't make it, but with entry points happening in good places to pick up anything from 5-20+ points on NQ, to me it is crazy to leave that much on the table when you are already in the right place to collect it.

The failure points are the points at which the market either runs out of steam or reverses itself. (and I'm not talking about small retracements) I'm talking about underlying shifts in sentiment that reflect in price behaviour and action.

They are easy enough to spot with the naked eye, but awful to try and code.

Therefor I have extensively tested against mechanical exits and more subjective exits (both forward and backwards testing in order to obtain relaibale data with backtest data being adjusted to the both lowest denominator and the mean performance of the forward tests).

So it is possible to obtain reliable data over time (but it is very time consuming to do so), without actually mechanically coding. At some point I will discover how to mechanically identify reliably such exits points as I routinely use in live trading, but this has proved time and again to be one of the hardest parts of system design, recognising the difference between a retracement within a trend and the trend actually being broken in a relaible mechanical manner.

Kind regards

Natalie
 
...thank you for explaining what you mean by failure points. With the exception of trying to be on the right side of the market for the 10:00 AM Eastern news announcements, I don't try to catch the big ones. I sympathise with the problem you have described. I am a terrible intuitive trader, so much so that I have a note above my screen which says "If you think your intuition is so much better than the system, code it up and test it." With your "intuition" of where the failure points are, may I ask how many NQ points per day on average you are able to extract? I ask because you seem to have the courage (which I lack) to hang in there for a substantial part of the day. For amusement I draw trend lines and identify breaks during the day, but given the complete unpredictability of the start time of the big moves, I would never consider trying to trade them. My approach is that if there is no time predictabilty, don't trade. Best regards. - Mike
 
I never use mechanically generated exits to take profits. Most all of my exits are discretionary from viewing the price action. Yes there are many times when I bail early in a trade and leave profit on the table, but there are also times when I hit 5 to 9 point moves during the day. I just like trading my own style, and I would not do very good trading a mechanical exit strategy style with my personality type. I don't use the "autopilot" very much when I fly either. I'm here to live not watch! :D
 
Within the context of this conversation I'm not talking about discretionary entries at all - that is a different subject. I'm only talking about my findings with relation to system generated entries. I do however use both.

Kind rgeards

Natalie
 
Girlpower, may I ask a few questions? As you know I do what I do but I am trying to add to my skills and knowledge.

You say system trading is "mediocre with fixed targets" but that the "vexing question (you are) grappling with right now is how to identify/quantify the failure points."

Just so I get the terminology correct, are failure points where upswings and downswings exhaust themselves?

If you find it difficult to "identify/quantify the failure points", does that not take you back to fixed exit targets? Can you not layer fixed exits (ie y, y+1, etc)? Admittedly you wont get the max juice short of intervening and continuing your position if you need to, right?

Where a definite repeating pattern is identified and it satisfies backtesting, why shouldn't one use a fixed target exit system to exploit this?
 
Yes- that is pretty much what I'm talking about with failure points.

There is no reason why you couldn't if you are happy to take a smaller fraction of the good moves while still collecting the same failure rate for the trades.

refer back to my post on the previous page...

Best

Natalie
 
I found myself mentally nodding at just about every post here. I suppose my biggest quandry is that I am pretty rigorous when it comes to entries -- they are not mechanical by any means but I do have a very strict set of conditions to warrant putting on a position. However, like some of you, I am pretty much "intuitive" when it comes to profitable exits, and I prefer to let things ride rather than take the small but surer profit. Mostly it comes down to deciding whether or not to sit through the first retracement.

The reason I stopped entering trades based purely on intuition was that I did not like the mental baggage associated with being wrong based on just a hunch; I wasn't confident enough, and so it led to alot of second-guessing and frustration in hindsight. Placing strict parameters based on many hours of research allowed me to trade, take losses, miss big moves etc without much psychological wear and tear.

I believe these issues still exist for me for exiting trades since they remain intuitively executed; I suppose I tolerate leaving these issues unresolved by the simple fact that they are profitable trades, but there must be a better way. :)
 
Girlpower, may I add a question?

Backtesting system trading .. when you do it, what is a sufficient period? Currently I use the last 7 months.

Is there a best sampling period? Can it be too short or too long?
So what is better do you think .. 5 yr, 2 yr, 1yr, 6 months, 3 months?
 
You must log in or register to reply here.
Back
Top