Welcome to NexusFi: the best trading community on the planet, with over 150,000 members Sign Up Now for Free
Genuine reviews from real traders, not fake reviews from stealth vendors
Quality education from leading professional traders
We are a friendly, helpful, and positive community
We do not tolerate rude behavior, trolling, or vendors advertising in posts
We are here to help, just let us know what you need
You'll need to register in order to view the content of the threads and start contributing to our community. It's free for basic access, or support us by becoming an Elite Member -- see if you qualify for a discount below.
-- Big Mike, Site Administrator
(If you already have an account, login at the top of the page)
Disagreement is often good, although I suspect we're talking the same language
I think it depends - if you randomly entered/exited the market as per the above, where would Douglas's definition of an edge fit in?
Of course by decreasing target, increasing stop-loss, your win rate is bound to increase but in itself this does not sound to me like 'an edge' in the way that Douglas described it, i.e. artificially increasing the win rate does not constitute an edge, does it?
My own experience: for a while, I identified a certain exploitable condition on CL. With sufficient data, this condition appeared to take place just shy of 3 times over 4 occurrences (73%).
The screenshot below shows 2 of 3 possible situations occurring. A categorises trades where I'd get filled and reached a fixed target. B (which is not shown) categorised trades where I would not get filled and C categorised trades where I'd get filled and reached a fixed stop loss.
For a certain period of time - this was an edge for me, i.e. clearly A was occurring more often than C, and by exploiting that, you'd get profitable results.
Now if we asked: is A an edge today? My answer would be likely no. But that's not the point. Say that event continued to this day - then, by Mark Douglas's defnition, you would have an edge.
You make some very valid points. I have only quoted a small section of your post though as we are discussing the concept of an edge.
Let's say in the above you have a probability of the market bouncing off a support line and you find that it happens 73% of the time. You could then place a limit buy at support. Now of course, you need to protect yourself so you place a stop-loss two ticks below your entry. Assuming you get filled, you might find that the market still bounces 73% of the time, but you get stopped out 50% of the time of the 73% of the time it bounces. Thus you have a win rate of 36.5%.
This is a silly example, but it is a big part of what happens in real life. Now, assuming you get stopped out frequently, you decide to increase your stop-loss to 100 ticks and this increases your win rate to 100% of the 73% times it bounces. Now your win rate suddenly becomes 73%. Again, this is a silly example, but it shows how modifying one parameter can have an effect on an "edge".
You can do the same with a profit target too. Make it bigger and the win rate will drop, make it smaller and the win rate increases. It goes to show that even with a positive "edge" there is no guarantee of profit. You could also use resistance as your profit target and be subject to the exact same issues with the target as with the entry. Getting the balance between your profit target and stop-loss correct to have a positive expectancy is not always that easy.
It does not mean that your observations is inaccurate or invalid - it merely means that exploiting it is not as simple as a lot of people think. My point being that an edge can not only be "a higher probability of one thing happening over another" - other factors need to be included and they can significantly alter the outcomes of said edge.
Firstly apologies to Rich Independence as I appreciate this is somewhat off-topic.
What you write above confirms - in a way - we are talking about the same concept, only in a slightly different format.
I just want to stress - my previous example was with a fixed target and a fixed stop loss. That fixed target was being hit 73% of the time - meaning win rate was 73%. I will also add: risk/reward ratio was 1:1 (risk x ticks to make x ticks which, in the above example, was profitable)
I agree with you that expectancy needs to be part of the equation and just a win rate alone does not, in itself, give you an edge. In my case expectancy was positive, with the win rate above.
It may look like we went very off-topic, but when you look at the rest of this thread I would say it is definitely on-topic. What we have done is highlight one of the pitfalls that many people overlook and that there is a huge interplay between all aspects of a trading system (or edge if you prefer ).
Regarding your stats, I was just using the probability of a win you gave to highlight how changing parameters can change results. Perhaps instead of using "you" I should have used "a trader" in my post...was only using your stats as a starting point to make my point.
Based on your stats, it also highlights another item that I found especially relevant to shorter-term trading - shorter-term trading systems tend to be very dependent on having a high win rate. Probably a multitude of reasons for this, but it was one of the reasons I chose to move to a longer term approach - as part of my "edge", I don't need to maintain a high win rate. I just need one or two big winners a year and depending on the size of the winners and the frequency of my trades, the win rate becomes almost irrelevant.
I remember Okina and Inletcap debating probabilities where the former would say "x% of time y happens gives you a good starting point for analysis", whereas the latter would say "that's irrelevant because you don't know how far the market goes after y has happened".
It sounds like we're having a similar discussion - So I think it all boils down to the definition of an edge. When you say you disagree with MD's definition of an edge, I now understand why you say that. I don't know that by saying 'a higher probability of something happening over something else' he meant it to equate that with a higher win-rate. But I do know that a high win-rate alone is not the answer. If he meant that, then I think you and I are in agreement.
Once you have established that something happens more often than not, that's where the real work starts, i.e. the edge in itself is not in simply the higher probability of something happening, but the ability in the long term to exploit that probability for a positive expectancy. It all comes down to edge development, I think.
This discussion is only valid to statistical edges.
There are edges that are non statistical and you simply know, doing X will return Y.
It is also important to understand that edge is cumulative.
While it is fairly easy to get to a 50% (tossing a coin should give you that result in theory),
it becomes a bit more difficult to get to 60%, really tough for 70%, and so on...
The higher you go, the more difficult every incremental % becomes
As an example, for the moment i'm kind of stuck to a level of 82%
IMO YOU are the edge.
Trading is a skill your decision will influence your Expectancy.
No matter what is the probability if you can't execute properly and know or feel when to cut your losses it's pointless.
Also trade what you see if the range is 4 tick and you are looking for a pullback then take the 4 tick don't try to shoot for 20 ticks because the market might not give you that.
Know what is the most common rotation of your market aka harmonic rotation on your time frame.
I would suggest you try scalping because you'll get a lot of experience on a shorter time.
Also if you account is small may be the 5 year ZF which is low tick and very liquid can be nice,and rotation are quite small so drawdowns are going to be less volatile and if you become good you can trade more size or switch to the ZN or ZB with a higher tick value.
For example try to trade only withe the momentum/trend and never try to take reversal I think you will be better in the long run.
Trade what is happening right now,trading is about being decisive,not to be right then manage the trade.IMO
I want to add to what you say here. To define an edge and to compile statistics on that edge is something I question the validity of.
Normally in a test environment one has a control sample off which everything else is measured. IMO, Its mostly impossible to replicate the exact circumstances for every trade so there is still discretion involved which leads me to the question of how discretion is measured and quantified into a statistical edge?
Lets say you are an aggressive range trader and you have defined your range as per the image. Your stats say that you have a 70% win rate playing range extremities. My contention with statistical edges is that every situation is different.
Have you taken into account every variable (as per your control sample) that could affect the accuracy of your statistics?
Day of the week?
How volatile is the market?
Is the market leading up to a major news event or pushing away from one?
Has your scenario come off a trend, or a continuation after days of uncertainty?
I agree with you 100% about probability. The mind can see things that are actually not there. If traders used the standard, tried and true, scientific method, they would have a much better chance of developing their edge. This is because the data doesn't lie. Either it works or it doesn't. In the lab, one of the first things you learn is that 90% of what you do will go to "File 13". It's the 10% that you're looking for. And when you find the 10%, you have to whittle it down to the 1%. This takes a lot of time and effort. There's a big problem in this industry with education IMO. If traders were trained from the git-go to think like a scientist, I think their view of what they're trying to do would change dramatically.
The first thing they should do is hang the jpg below on the wall. You may still fail and never find the 1%, but at least you got there the right way.