My I suggest taking your thinking up another notch - and start thinking in uncertainty
It the overriding force / law / influence / #1 ruler of - trading..., and the mkt
RN
Understanding Probability
- Thread starter Tavurth
- Start date
My I suggest taking your thinking up another notch - and start thinking in uncertainty
It the overriding force / law / influence / #1 ruler of - trading..., and the mkt
RN
Thank you, I was unsure of whether to introduce Heisenberg into the discussion as his principles are a subject which has been on my table for several years.
My problem is that his math is impenetrable to me, and so I can but slowly work my way through logic and experience.
I guess you need to explain the graph to us first.
Normal distribution is Not quite applicable to trading reality in terms of risk management. Perhaps the most important thing practically related to probability distribution would be its tail risk that would kill many traders' accounts.
http://eztrade.com/probpart1.htm
If you can properly manage tail risk Economically, you have got the holy grail, sort of, imo. I recall that HarryTrader on ET talked about that a lot in the past.
http://www.elitetrader.com/et/index...stically-significant.19873/page-5#post-393760
http://www.elitetrader.com/et/index.php?threads/normal-distributions-are-not-the-norm.25943/
Last edited:
Thank you for the links!
The graph shows simple probability of a two dimensional system, in trading it could be percentage increase of a security vs profit from buying and holding at that point in time.
I agree wholeheartedly on managing tail risk, for even simple refinements (ATR, Standard Deviation) can clip the black wing of the swan effectively.
One large problem with effective optimisation of the clipping process I've outlined in the diagram below:
As we can see, clipping {A} to remove the tail risk of Black Swan events, will in most cases, remove the probability functions of {B},{C} and {D} at the same time.
This leaves us with the following:
In this case, we may still be left with a profitable distribution. However, even and especially if we overfit this Black Swan avoidance system, we will be at risk of market change, which can turn a profitable system into one a consistently losing one.
I believe the trick is in how we define these events, and by skewing our probability so as to negate the clipping effect.
Last edited:
there are two types of guessers in the game, some who base their bets on some kind of Russian probability law which states, "Two Aces are Better Than Seven Duece Offsuit."
and others who will play almost any two cards and "See What Developes."
In the end, they both just bet on probablility, but there are levels.
It may not be that obvious on the first two cards.
and others who will play almost any two cards and "See What Developes."
In the end, they both just bet on probablility, but there are levels.
It may not be that obvious on the first two cards.
I think I am a bit slow in learning the graphs.
1. I think you should label each axis on the graphs for easy understanding and reference. Maybe your focus of this thread about Probability is for something else completely, such as forecasting directions, position management, and others. But labelling axis is still preferred.
2. I would assume the horizontal axis is for Annual/Expected return, vertical for Probability distribution. Left-hand side should be usually/conventionally for Loss. Now your clipping effect seems trying to limit the Profit on Right-hand side, while allowing the maximum Loss on the Left-hand side.
http://www.mathwave.com/articles/extreme-value-distributions.html
3. All the above may be not as important as How a trader is able to carry out the Clipping of tail risk effectively with an Economical approach. I guess every educator could produce a proper graph to represent what traders want to attain, but How to attain the expected result!
1. I think you should label each axis on the graphs for easy understanding and reference. Maybe your focus of this thread about Probability is for something else completely, such as forecasting directions, position management, and others. But labelling axis is still preferred.
2. I would assume the horizontal axis is for Annual/Expected return, vertical for Probability distribution. Left-hand side should be usually/conventionally for Loss. Now your clipping effect seems trying to limit the Profit on Right-hand side, while allowing the maximum Loss on the Left-hand side.
http://www.mathwave.com/articles/extreme-value-distributions.html
3. All the above may be not as important as How a trader is able to carry out the Clipping of tail risk effectively with an Economical approach. I guess every educator could produce a proper graph to represent what traders want to attain, but How to attain the expected result!
Last edited:
A good point, thank you for bringing it up, I'll make the modification in the future.
A few years ago, I first turned a Market Profile chart on its side, causing the left to be high, and the right to be low.
Eventually I became habituated to this model and forgot that other people may work differently.
Exactly, and I will restrain from planning obsolescence of my posts. I'm sure what is right and profitable for me today, will be laughable, come 10 years the road further.
so the job of the trader is to tie what people are guessing on to something physical, like how much cash they are sitting on, and how much crude we have, etc. And then it's just as easy as rolling dice. You always know the most likely outcome is a seven.