Member |
Leon,
As you can probably see from my post count, I very rarely post to the forum. I have been trading for long enough now that, like many in the same position, I limit my posting to the few occasions where I can potentially add some value. For some reason, your post struck me as one of those occasions - I see someone who is on the right track, but just can't seem to get everything to 'click'. While I truly believe that most forums like this are largely comprised of the blind audaciously leading the blind, there are the few people like yourself who are humble enough to put it out there in a reasonable way - only after having explored extensively themselves. Perhaps I can be of some help.
Firstly, I understand how you have grown tired of the platitudes: using sensible risk management, cutting loses, mitigating emotion, naked trading, trendlines, etc, are thrown around casually, but you can't just bang them all together and hope to have something that is remotely profitable.
So what do I truly believe you are lacking, Leon? Quite simply, you are lacking a market inefficiency that stands to be your 'edge' (to use another of those terms). Once you have that, and you figure out a way to consistently capitalize on it, many of the tools that you have already learned will serve to keep the machine running - but, unfortunately, they alone can't build it without a proper foundation.
Before I start, I will disclose that I am extremely biased towards mechanical trading. In my view, the factors that makeup every 'A-trade' can be mechanized if one endeavors to take the time to meticulously define every aspect that made it so (at least if you are using technical inputs). There are certainly some great discretionary TA traders who I believe have completely internalized the specific criteria that comprise their edge, but I find that having it clearly defined makes the entire process much less stressful, and far more consistent (if perhaps also less fun and exciting).
Anyway, you can't even begin to do any of that until you have a strong foundation - and this is precisely where many get it wrong. They want a mechanized model, so they do a 'backtest' of various rules that they have composed. However, what is the logical basis for those rules, aside from the fact that they appear to historically work? Finding that you make x% over x years based on rules that are not grounded on a specific understanding of market dynamics is likely to be no more than a mere coincidence. To borrow from Taleb's monkey example in 'Fooled by randomness', if you tinker with enough variables you are going to find something that fits. This of course is the process of curve fitting - something that is easy to inadvertently do, mainly because it is so perversely satisfying and so seemingly reasonable.
The only way to really avoid this is to first truly understand what your chosen inefficiency is before developing a system to exploit it. And I don't mean understanding exactly why x bank placed x order that moved the market x amount on xyz date; rather, the overriding concept that what you're designing will attempt to capture.
I know this all may seem somewhat esoteric, but figuring out an edge can be somewhat of an esoteric (and exhausting) process. Doing all of what you have done; learning all that you have already learned; falling into all of the traps that you have likely already fallen into, are generally all prerequisites to the process of understanding, in a broad sense, why certain things tend to happen more often than not. It is understanding why, given certain identifiable situations, people 'tend' to act in a particular way with at least enough reliability to exploit.
Once you have that, you can than take that concept and build yourself a great system. You recognize a system that you have designed in this way because watching it in action will be a satisfying experience (a somewhat slow, unexciting pleasure), and something that you have a very large degree of confidence in. And all of this will ultimately be constructed with many of the tools that you already know.
I have put a link to another post of mine where I go into a reasonable degree of detail about my development process. There is no need to rehash that here.
To give you a few concrete things to take away:
- if you are going to use trendlines, support/resistance, etc, make sure that you have a very consistent way to define them. Simply going by 'seeing it' hardly provides much of a basis for an edge, even if one's judgments often prove to be right.
- Make sure that your system dynamically allows for differing volatility levels. Something that is shown to make x pips in 2004 means very little in today's market.
- Take every single trade that meets your rules. If you are trading on an hourly basis, that means either looking at the charts every single hour or potentially developing an EA sophisticated enough to encapsulate the entire process. It is for this reason that I don't trade on anything less than a daily basis.
- Making sure that you will know what you are going to do in every possible situation. Be it placing an order, not placing an order, taking out an order, adjusting a trade's risk, exiting a trade - none of this should be left to whim, mood or chance. No matter what happens, you should have an unflinching response - one that would have been exactly the same if everything happened all over again.
- Finally, have a fail-safe. No edge is guaranteed to continue to work. Some are more robust than others, but nothing is certain. You should have a measure for live trading whereby you know when your edge has ceased to work (or potentially never actually worked in the first place). The easiest is statistical - having an understanding that when you lose x trades consecutively, or go into x drawdown, then you should probably take a closer look at either your edge, the way that your system attempts to capture it, or possibly both.
Anyway, I hope this helped Leon. From something about your various posts, I think that you have the combination of open-mindedness and many of the necessary tools to do this. I wish I could just tell you (and really, I probably could if I so choose), but that would likely hinder you developing something organically yourself, which could potentially lead to a better equity curve than mine.
http://www.forexfactory.com/showthread.php?t=67217
Немає коментарів:
Дописати коментар