This was the first question that came to mind after reading VT. First of all, your argument is logically flawed: The fact that people don't do it does not mean that it doesn't work (you are using a negative premise). You could of course argue that it is common sense, and not even worth considering but I'm not so sure about that. If people always did the smart thing, there would be no business for casinos for example, and they are doing pretty well as far as I know.
Can you be profitable with random entry?
- Thread starter Tremaine
- Start date
This was the first question that came to mind after reading VT. First of all, your argument is logically flawed: The fact that people don't do it does not mean that it doesn't work (you are using a negative premise). You could of course argue that it is common sense, and not even worth considering but I'm not so sure about that. If people always did the smart thing, there would be no business for casinos for example, and they are doing pretty well as far as I know.
The results for using the same exit system no matter what your entry is (random) just cannot be reliable over the long run esp in scalping, given the spread, commissions, etc. Sure, one person's specific set of random entries could happen to be quite profitable, but how can you expect that to translate or continue for another?
Look at it this way, you could draw up an exit methodology that would likely be profitable for strongly trending markets no matter what the entry but that would fail miserably when thing turn choppy and rangebound. When you start to think about how you would figure out when to shift gears between the two, you start getting into the things that keep 95% of people out of being profitable -- it goes far beyond any simplistic exit methodology. So my point is, if you're unprofitable to begin with picking your own entries and exits, it's just illogical that "simplifying" the variables down (whether via randomization or otherwise) will somehow get you across that line.
Look at it this way, you could draw up an exit methodology that would likely be profitable for strongly trending markets no matter what the entry but that would fail miserably when thing turn choppy and rangebound. When you start to think about how you would figure out when to shift gears between the two, you start getting into the things that keep 95% of people out of being profitable -- it goes far beyond any simplistic exit methodology. So my point is, if you're unprofitable to begin with picking your own entries and exits, it's just illogical that "simplifying" the variables down (whether via randomization or otherwise) will somehow get you across that line.
Good thoughts!
Let me comment: to make random entry work, you should shift the problem to "detecting strongly trending markets". I agree. This is the same as saying that it's not going to work at all. Huge loser crowds can't even detect trending or choppy conditions - other than by hindsight of course.
VT's blurbs amuse a lot of idiots but don't foster any practical trading methodology.
It seems that most people agree that it is easy to make money in a trending market. It would seem that the smart thing to do would be wait for a trending environment. I think the biggest problem people have is sitting on their hands so I am kind of a fan of Woodies approach to just trading the indicator, up to a point. It will keep you from overtrading. Why not pick a trending indicator, macd, adx or whatever appeals to you and not watching anything but that indicator until it tells you to start watching price? I do not agree with Woodies choice as I dont think 6 bars constitute a trend but the concept makes sense to me.
If I knew the answer I would'nt be trolling on this message board.
The answer is ..... *all* entries are random.
Based on prior history of your setup, and the fractal you're trading, the *probability* of the price moving in the direction of your entry could be high enough to give you a profit.
Tremaine you are quite astute in your observations of the value of random entries for simulations. I agree with you and Bolter, there are many applications to this. The individual who has done the most work published on ET would have to be Acrary so you should read his posts on 'Edge Test' with your ideas in mind.
But the uses of randomisation in system developement go beyond that as well. I use randomisation to evaluate indicators for example. So that in a given system I know that a particular indicator may be in the 95th percentile of random results (ie, results based on an indicator that randomly filters out the same number of trades as my indicator).
Another test I'm using looks at results in systems (and portfolios of systems) if there are frequently too many overlapping trades. The problem would be to decide which trades might have been filled and which were passed over(once our account was fully utilised). Given that the Best Case Equity Curve may be fantasic, and the Worst Case is terrible, which do we choose as representative? So I generate hundreds of equity curves using a random priority to choose from the overlapping trades (in their original form and timeframe). Then using distributions of the equity curves helps determine the probability of the various outcomes.
It's the basis of scientific testing to index results against a control. I cant imagine a medical researcher submitting data to the FDA that simply says drugX is good because we injected 500 patients with it and more than half of them got better. The real question is how many got better using the random control vs. those that got better on drugX (not to mention the question of what heppened to those that didn't get better
)
But the uses of randomisation in system developement go beyond that as well. I use randomisation to evaluate indicators for example. So that in a given system I know that a particular indicator may be in the 95th percentile of random results (ie, results based on an indicator that randomly filters out the same number of trades as my indicator).
Another test I'm using looks at results in systems (and portfolios of systems) if there are frequently too many overlapping trades. The problem would be to decide which trades might have been filled and which were passed over(once our account was fully utilised). Given that the Best Case Equity Curve may be fantasic, and the Worst Case is terrible, which do we choose as representative? So I generate hundreds of equity curves using a random priority to choose from the overlapping trades (in their original form and timeframe). Then using distributions of the equity curves helps determine the probability of the various outcomes.
It's the basis of scientific testing to index results against a control. I cant imagine a medical researcher submitting data to the FDA that simply says drugX is good because we injected 500 patients with it and more than half of them got better. The real question is how many got better using the random control vs. those that got better on drugX (not to mention the question of what heppened to those that didn't get better
)this aint a stupid thread and the questions the op is ponderin upon aint stupid either; maybe addin' to your winners would make your system work better, still, i think that we may be coin flippin' already and what makes the diff between profitable traders and not profitable is strict money management. maybe there are traders that have better instincts or a better system to make the odds of winnin' skew decidedly in their favor but i also agree with mhashe that all our entries can be viewed as random anyways.
I need more information - specifically about the prior three "inflection" points, where they price breaks, did they hold S/R levels, how strong was that recent low?... Without that knowledge/info, and without knowing the markets volume mood that day, I would pass on this trade.
If I had to make an impulsive decision, I would go long at a tick above the low of the 11:30 candle because:
1. The anticipation of many others looking at the price point would be another sell-off from the mini-consolidation between 11:30 and 11:45. I think there will be more stops positioned above the 12:15 candle, all the people that profited off the downmove as well as new "missed the boat" short entires.
2. The highs of the last two candles were higher than those of the mini-consolidation mentioned in point 1.
3. I can set a stop under that 11:50 low and if the market continued to go up I potentially get a good R/R on the trade, possibly 1-3 or better.
This is a good long setup IMO but I need more info. Thanks for putting it up. I think this chart illustrates the principle of the market doing the thing that the greater of number of people do not want to happen.
Mike
P.S. I'm bored on a Sunday looking at charts and trades for last 2 hours
