Probability, Statistics & Trading

[Perrington]

RUDEBOY

It looks like probability is not for me, hopefully I continue in the fashion I am presently, and I wont need to investigate further. Iam so entrenched within intermarket analysis, that I couldnt ever see me taking out a position based on probability in one specific market, as to me everything is relative to bonds and commodities, through the medium of correlation.

 

[RUDEBOY]

Absolutely nothing wrong in that, if it works for you, don't try and fix it! Hope we chat soon, Good trading! RB.

 

[frugi]

Casinos and bookies enjoy the luxury of fixed probabilties, thus they can ensure that their expectancy is positive, given a sufficient number of bets. There is no roulette wheel, deck of cards or overround in trading, and traders have to learn to accept this. Quants may try, employing every sort of model in their quest, but at the end of the day there are simply too many factors involved for them to predict the future with consistent, watertight reliabilty. One day the Black Swan will appear and destroy them, despite this appearance being a so-called "infitesimal" probability event. Given time and humans, regression to the mean is inevitable, at which point strategy has to change if one is to survive. Bramble's right: ditch the theory and find something that works, until it doesn't anymore, because the anomaly has been ironed out and it's time to find another one. Gaps are always filled, as it were, but new ones will be created in their place and we have to find them. ?

 

[Perrington]

Fortunately I already have something that works. Modest, but pays the way, and is likely to remain constant until free energy comes along.

 

[donaldduke]

Most people trade worse than random. In fact they normally do exactly the opposite to what a better
than random winning trader would do:

The point at which a winner will enter with a sell is probably a point at which a loser will enter with a buy.
The point at which a winner takes his losses will be the same place the loser takes profits.
And the point at which a winner takes his profits is the point at which a loser will take his losses.


A losing trader will probably only ever trade randomly if they use a random number generator for entries
and exits.

Human nature is not random.

 

[TheBramble]

until free energy comes along.

You might new a new thread for this one...(Tesla et al)

 

[Tuffty]

Perrington,

You said "It looks like probability is not for me, HOPEFULLY I continue in the fashion I am presently"

"Fortunately I already have something that works. Modest, but pays the way, and is LIKELY to remain constant until free energy comes along."

The words "hopefully" and "likely" are both all about probabilites even though you say probability is not for you.

 

[Perrington]

Indeed, but those comments dont set about designing a strategy for trading markets based upon probabilities, they merely reflect the unknown future.

 

[Perrington]

Lets look at the facts here.

Market flux, how would I set about defining market flux, bearing in mind this is essentially what kills off probability based strategies. If it was not for flux, probability would be perfect, but as we all know, its far from it.

Lets list a few scenario's that are potential killer's for any probability based set-up at the drop of a hat.

1) Insider trading (blurb)

2) Terrorist attack (not that this threat actually exists back in reality)

3) Economic data ( a very wide range of events that fall into this one)

4) A big player dumping or buying (happens every day, market manipulation)

5) 6) 7) Fill in the gaps time !

So, given that this is infact the case, how many of you guys are placing faith in this kind of strategy to make your living ? I personally dont like it, but then again, we all have differing idea's.

Having said all this, one area of probability that may be "interesting" is a strict mathematical set-up based on inverse probabilities. Even taking into consideration the above. Possibly the contrarian's solution.

 

[RUDEBOY]

Perrington, i thought you were on the right lines until you mentioned 'strict mathematical' set ups. Now youv'e cocked it up! Mmm!

 

[Perrington]

Ok, strict is possibly a little specific, however Its possible you are failing to percieve what I mean by this, as I in no way expand on how this could be possible. Consider retracements as direction.


However, please do show me where Iam going wrong !

If your using probaility set-ups your trading by numbers are you not ? And if you are trading by number's you are open to all the above I posted, which is why I personally avoid it.

 

[Tuffty]

Perrington,

I don't really understand what you're saying in all of this thread. Surely everyone has a 'probability' approach to trading as the future in unknown.

What specifically do you mean when you refer to a probability based strategy?

 

[Perrington]

Well, maybe Iam looking for something that doesnt exist ?

What I mean by a probability strategy. Something that is defined as providing an edge due to the numerical frequency of one event being more likely than another event. What I mean by strict, is something that is accurately defined.

 

[RUDEBOY]

Sorry Perrington, wrong end of the stick.

 

[RUDEBOY]

Perrington, whatever you are trying to analyse (wether it exists or not) is a good thing! I'm not going to get all prophetic and philisophical here, but, open analysis of the markets can only be a good thing! Good trading, RB.

 

[Perrington]

Yes I agree, and I think you for your input, after all what is the purpose of the forum ? (making friends aswell, not that I make many lol)

 

[Tuffty]

"an edge due to the numerical frequency of one event being more likely than another event"

The above would be a way of describing any successful participant in the market. I still don't really understand where you're coming from.