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)
The single most important chart in your setup, and why
As I noted earlier, John Ehlers discusses Laplace distribution versus Gaussian distribution in his book Rocket Science for Traders. I'm sure others have done the same analysis. Volume can be 300 or 3m, but price distribution just does not fall neatly in the bell curve.
Ehlers did a couple of webinars with Big Mike, which were very informative. You can watch them over on the FIO YouTube channel. Some of the stuff went over my head (advanced/applied mathematics), but a year later I have a better understanding.
FYI, I am not a statistician, nor do I play one on the Internet.
well... you just have to look... and understand the chart
the Gaussian distribution is wider...
meaning that there are more values or samples away from the center
the Laplace shows that these values drop off fast...
meaning the farther you get from zero, the drastically fewer values or samples exist
remember... these are charts of distributions..
in both cases, there are just as many up moves as down moves
there are a lot more 10 cent moves in the normal/Gaussian distribution than in the Laplace
The Laplace (or double exponential) distribution, like the normal, has a distinguished history in statistics. It has applications in image and speech recognition, ocean engineering, hydrology, and finance.
What does it look like?
The plot in Figure 1a depicts the Laplace probability density function centered at zero and, for comparison, the normal one.
To explain what these fatter tails mean, consider that the cumulative probability of the standard normal at 3 is approximately 99.9%, which justifies the common assertion that events 3 standard deviations or more above (or below) the mean are unlikely. However, the corresponding cumulative probability of the Laplace with the same median and variance is about 99.3%. Therefore the occurrence of an “extreme” event in a Laplace population is more than five times as likely as in a normal population
Thanks for the explanation @Artfldgr and @vmodus I'll try to find that webinar and find out a bit more about the laplace theory although it will be more for curiosity than anything. When you talk about 99.9 vs 99.3 percent.'extreme event ' probabilities are you not getting caught up in irrelevant minutae? It does not matter that one is more likely then the other, what matters is that when an extreme event occurs, both models are capable of identifying it. and in any case you can define your own hypotheses rejection zones. Just because a textbook says it is 3% or 5% doesn't make it a rigid rule that you can't bend.
for us its more about the likelyhood of the extreme event..
if the stocks were gaussian, there would be more of them moving 10 cents..
if the stocks are laplace, fewer of them will...
if the stocks were gaussian, fewer would move zero
if the stocks are laplace, many more of them stay at zero movement over time
its harder to make money in a laplace financial world..
i think you might get it easier if you think of lots of stock moves of different amounts being plotted...
how many go to zero, how many are 5 cents, how many are 10 cents, etc..
if there is this spike in the middle, then a lot more go to zero or 2 or 3 cents..
a whole lot less go to 5 cents or 10 cents...
THATS what the charts are about
my plot in the post shows the price movement from teusday close to thursday...
by far.. most of it amounted to zero... very few were 20 cent moves compared to gaussian
the bottom line about it is that its harder to make a profit..
and if you find the webinar, post it... i would like to see how to better explain it!!!
sucks to know your not doing a good job of it...
Found and watched the webinar (in the elite section @Artfldgr). All I can say is that if you discovered an edge that printed money you would be using that edge to print money. Not writing books and selling indicators.
Also, its one thing when something works in theory and you can show how marvellously predictive an indicator is on a perfectly presented sinewave (with arrows and all) but another thing altogether when you throw it at a real chart and need to explain the losses with disclaimers like 'ah but you know thats the market you cant win them all', and others like 'it wouldnt have worked in this particular situation because of XYZ..'
This has NOTHING to do with what comes next it is simply information modelled on the past. In keeping wth the theme of the thread, here is how I use this stuff in my daily trading ritual. I am currently using the past 14 days because we are in a volatility period and I want information about whats happening now and not watered down with anything prior.
I extract RTH data for the ES and calculate the daily range. From that I determine the middle, edges, st dev, and range.
Once RTH opens, I work out what price levels are likely based on the session open.
I plot those on the graph right at the start of the session, and can be relatively comfortable that my expectations are synchronized with the current market environment.
Now the key thing to remember is that this model is simply one perspective. The best it can do is highlight what the probabilities are. It is NOT predictive in any way. If you are an inside out trader (breakout trader from value to a target) then this is a great system to determine what targets are likely to be achieved. If you are outside in (reversal trader from extremes back into value) then same thing in reverse.
So if you note that the market on Friday did not push beyond the 3rd -deviation and did not quite make the push to complete the 3rd +deviation although it got close and acheived 2nd +deviation. This is totally expected behaviour and within the rules that the market has been playing by over the past 14 days.
I did get some of that final move on Friday (target to +1SD). Not all but enough to close the day out positive.
Nicely done. It's good to be able to read something and know if you think it makes sense (I think it does ), and if it shows you something in an enlightening way (ditto.)
Let me make two comments, not criticisms, more like calling attention to terminology or phrasing:
1. Of course "information modeled on the past" can tell you something about what comes next. You just did that, in a probabilistic way.
2. Of course it is "predictive," in the only sense that anything can be predictive: there is an assignable probability to it. If you can't assign an exact probability for some reason, but you can still distinguish between more probable and less probable scenarios, then that is also "predictive," but less so. If by "predictive" you mean something besides "predictive with at least some sort of assignable probability," then you don't have any ability to predict anything.
For example, I accept the weather forecast as being reasonably likely to happen, and often plan what I do based partially on it. I know it is often right, but not a certainty.
I hope I didn't state this so strongly that it comes out that I disagree with what you wrote. Tell me if so, because I think what you wrote is very good, and I didn't mean to say otherwise.
An edge after all may be only a slight leaning of probability in one direction, which is quite enough over time if you can manage other factors right (losses, asset weighting, etc.). I doubt that anyone has anything much more than that.
Bob.
When one door closes, another opens.
-- Cervantes, Don Quixote
