Well, I guess that might work...But from what I know and have experienced, you cannot create solutions to a problem, until you have actually quantified the problem to begin with...Most people do not realize where or what they lack in terms of position management skills until they sit at their computer and say, "s###, the market dropped x points, I entered at y and I still did not make more than a dime"...THis is the crux of the problem...How does one max out exposure with minimizing drawdown...I am certain that many mechanical systems could be used to get the signals, but it is still such a small part of the whole equation...And really, how does a trader without experience differentiate between a terrible signal or a market that is so news driven as to make all signals basically irrelevant...How does one do all this, with alot of experience...
Vulture: Here your discussing technique, which is uniquely personal and is guided by what appears to work best for you. Don't misunderstand me, I think there is alot to be learned by discussing technique (I define technique as a more personalized strategy), and I'm certainly interested in discussing yours more fully (just not tonight). When trying to improve my technique (or lack thereof), I work on it using simulators (one for position management; another for daytrading). Yes, I know the limitations of simulators (nothing's perfect), but it does help to explore and improve technique. Or maybe I'm just blowing smoke.
Yes, I agree with you... Perhaps, I might have come across about too forthright in my opinions...you actually might be able to help with me with a question I have... Although I have heard rumors about this, are there simulators that will actually play back a specific days worth of price action real time(only it would be at a later date) kinda like a VCR...So that instead of just going over static charts to recreate a given scenario, one could spend a Saturday or Sunday going through different scenarios in "simulated time" and using a quasi workstation to enter, adjust and manage trades? Let me know if you have any knowledge of this?...
Vulture: That's my point. I don't think mechanical systems should be used for signaling entry points. In fact, I don't think entry points are as critical as position management (you'll recall that some Market Wizards comment on it's better to have a poor stock selection strategy with good position management, then the reverse). It would be of value for learning traders to sit down and determine what the risk is for any trade they plan to enter; that is, what are they willing to risk (in total loss), what the expected gain is (the reward), and how volatile the stock is to determine the proper initial position size. Obviously, there has to be an awareness by the trader that these are issues to be addressed. Understanding that there is much to be learned separates the wanabe traders from those who truly understand the commitment to be successful.
Vulture: Try Cybertrader.com. You can download their simulator (the Pro version) and it plays back real data (even on weekends). It's pretty good, even in trying to simulate fills (although, it's still just a simulator).
I'm still sitting on the fence with regard to the automation vs discression trading, but recently I've been trying my hand at automation. It's been going "OK," but I can't say as I've given it 100% fair shake yet, since several times I've over-ridden my own systems! :eek: So here's my question: Assuming that an "edge" can be programmed, how much rocket science does it really need to be? For example (I haven't tried this!), if you take a four bar candle pattern, program it, and test it on 100 stocks, and throw in some money management, you are bound to find some winners. This doesn't require an advanced maths degree and by using stocks with positive results, you are assuming an edge. Anyone have any thoughts? It stands to reason that simple stuff like this has been done before, so why don't more people do it?
I decided that I should try to answer my own question! I tried writing some simple entry criteria to simulate some candlestick patterns. The first thing I discovered is that it's not as easy as it looks. Some of the descriptions are a bit vague (the 1st bar must be long) and being new at this, some of the programming was a bit tough for me. Having said that, I did a reasonable job of simulating a morning star pattern. It worked alright on some stocks, but didn't appear with enough frequency to be of much use. I suppose if you scanned a ton of stocks it might be worth while, but I don't think I'll be automating it any time soon. Does anyone have some candlestick code for TradeStation (6)?
Lightningsmurf: There are two points that I emphasize (it's ok if you disagree) when discussing Stock Selection and Mechanical Systems... 1) Stock Selection is best learned by the trader (a discretionary approach). Why? Because the Market and the Stock have their own unique behavior which changes over time. It is difficult to program a dynamic behavior, and this is why a system performs poorly overtime (curve fitting results in terrific backtest results, but they rarely hold up moving forward). 2) Mechanical Systems are best applied to areas which are rule based and static in nature (can be applied to any environment). Position Management, which is far more important to the success of a trader then Stock Selection (this is where most will disagree), is a perfect candidate for a mechanical approach (whether we adhere to these rules internally, or rely on a mechanical system). Of course, Stock Selection is clearly an important component in trading (you can't trade without picking a stock), but it offers the least reward for the effort (because of the randomness of the price fluctuations and the Market). Position Management is the least understood, and more complex to master (requires discipline and control over your emotions). Most traders do NOT apply any meaningful Position Management policies.
And I agree with KISS. BTW, a test was conducted several years (or more) ago in which stocks were randomly picked to test the value of Position Management. The results (I don't recall the details) were impressive (that is, strongly positive). If one could also improve the Stock Selection to be better than random (50%), you could become another Nicolas Darvas.