Better to Predict...or React?

[Walther]

Quote from Cheese:

Neither.

Accurate prediction modelling as the most successful form of immediate trading is not a favorite among independent traders if indeed used at all. Many or most claim either they don't need to or that reading the market is better.

What is the reason(s) why predictive modeling is not used by many ?
What do you basing your models on ?
Thanks

 

[reno4nook]

all speculation anticipates (predicts) a price move in your favor.

So whats the difference between predicting and reacting??

Lets say XYZ company is about to announce earnings.

This is what I call predicting - Buy XYZ before the announcement because I think earnings will be good.

Here's what I call reacting - After XYZ announces good earnings, I notice it's moving up on good volume. The overall market internals are also good. I buy XYZ.

I react to the market instead of predicting it.

 

[Cheese]

Quote from reno4nook:

all speculation anticipates (predicts) a price move in your favor.

So whats the difference between predicting and reacting??

Lets say XYZ company is about to announce earnings.

This is what I call predicting - Buy XYZ before the announcement because I think earnings will be good.

Here's what I call reacting - After XYZ announces good earnings, I notice it's moving up on good volume. The overall market internals are also good. I buy XYZ.

I react to the market instead of predicting it.

Yes, you are talking about reacting.

Predicting as a serious market approach means having the prediction worked out beforehand.

 

[Cheese]

Quote from Walther:

What is the reason(s) why predictive modeling is not used by many ?
What do you basing your models on ?
Thanks

I will give short answers.

What is the reason(s) why predictive modeling is not used by many?
Independent traders want to get started banging in some 'buys' and 'sells' and then carry on from there trying to get better.
Institutional approach: it will it turns out take, say, a couple of years full time dedicated work to build and test a particular trading model, fine, then you go in and clean up the market. But don't think thats complexity beyond one individual .. its just means completing the route of problem-solving and fine tuning.

What do you base your model on?
Data and parameters. Not chart based. No indicators needed.

 

[agtrader777]

I scanned through the posts and didn't see one person say "try to be an accurate observer of the market." It would seem to me that 'prediction' and 'accurate observation' of the market go hand in hand.

To take a position in any market, you have to believe that your 'prediction' of market direction is going to be correct. This is true of equities, ag spreads, options, EFT's anything. But, to react quickly and effectively, you have to accurately observe whatever the market is telling you.

Good Trading!

 

[zmostatabi]

Trading behavior is reactive, not predictive.
As a wise trader once said, “Do what the market is doing, not what you think it is going to do.”

 

[stfreak]

react !!! prediction is for analysts..

trading is about modelling a set of simple rules with few parameters.

when u sit in front of the screen trading, u should know the rules already. just concentrate on execution only, means reaction to the setups your rules provide you with. (if you try to react to the market itself, you mostly will be to late and be the bagholder)

when you trade its your job to execute the rules like a robot, so there is no time to work on your strategy during trading hours. do that before or after trading.

if you don´t have a set of rules, you have to predict the next move. predicting a move means modelling an opinion and opinion will blow up your account in the end.

to reply to the first posting in this thread: if you had a set of rules for your entry, you might have had an opinion about TOL going down, but you would have been long on this newstrade however.

I personally enter the market according to my rules and not according to myopinion and I have to say, that there are times when I think about market prepares for a rally. However when there is a short signal, I have to execute it, and guess what: no rally, but a selloff follows

it´s best to hammer it into your brain, that market doesn´t give a damn about your opinion and your prediction so you should learn not to care of your opinion either (sounds schizo but helps )

 

[Cheese]

Quote from stfreak:

react !!! prediction is for analysts..

trading is about modelling a set of simple rules with few parameters.

when u sit in front of the screen trading, u should know the rules already. just concentrate on execution only, means reaction to the setups your rules provide you with. (if you try to react to the market itself, you mostly will be to late and be the bagholder)

when you trade its your job to execute the rules like a robot, so there is no time to work on your strategy during trading hours. do that before or after trading.

if you don´t have a set of rules, you have to predict the next move. predicting a move means modelling an opinion and opinion will blow up your account in the end.

to reply to the first posting in this thread: if you had a set of rules for your entry, you might have had an opinion about TOL going down, but you would have been long on this newstrade however.

I personally enter the market according to my rules and not according to myopinion and I have to say, that there are times when I think about market prepares for a rally. However when there is a short signal, I have to execute it, and guess what: no rally, but a selloff follows

it´s best to hammer it into your brain, that market doesn´t give a damn about your opinion and your prediction so you should learn not to care of your opinion either (sounds schizo but helps )

Your rules can't be called reaction can they? Would your set of rules be your standard method of prediction then?

 

[stfreak]

no...those rules ain´t reaction, in that case you are right. but those rules predict the direction by reaction to certain behavior of the market.

clearly written: I see some patterns I spotted before and i know those patterns are mostly followed by a profitable move.

when working on my rules, its my job to find out what patterns ( i like the word market situations...) I should trade.
when trading, it´s my job to react to the patterns when they show up.

by reading through this thread, u can easily find out, that some guy´s are sitting in front of the screen and try to find arguments for the market going up and arguments for the market is going down. based on that arguments, they make their trades, means:
"uhm, housing looks quite strong today but index is in a downtrend. index hits support now, so I´ll enter long for...lets say 500 shares of XYZ"in other words: sophisticated guessing...and that blows u up in the end.

U exactly have to know what to do in every single situation you might get into when entering a trade. better:
"sector X is strong today and the sector leader Z is showing the beginning of my entry setup. I´ll prepare for going long of exactly 379 shares ( because I only want to risk 0.75% of my account for every trade) at exactly this price. index hits support now, so I pop in (reaction to my setup) knowing my final exit level is at that price, but I will exit before that level, when index slides throug support.

key here is, to enter by setup, no matter what u think about the next move. a setup is something static in that dynamic environment, something you can get used to. by using it over and over again, you get to know, what should happen after the entry, it´s home sweet home.

no pattern means nothing you can practice, means that every entry is a whole new situation to the trader. that doesn´t matter, when the trade shows a profit, but when it is a looser, emotions kick because you feel like a baby left alone in a shopping mall: u don´t know the situation, don´t know, what to do, so guess what: u start to cry

think about that: predicting is very rewarding for the ego. u where right and it´s the kick for u trade by trade.
reaction is so boring, because you repeat something very well known. it doesn´t pump adrenaline into your brain...but there are less emotions...and it´s more profitable

 

[Cheese]

Ok, I note what you say. It is interesting.

I would definitely say your rules are a basic predictive model. But, no, I wouldn't argue with you on that point because that would just be semantics.

I could rename my prediction parameters as 'a set of rules' and since the model operates both before the Open and during the market I would have to say that is further detailed guidance.

I can say it is not reaction.

However I would say on the notion of prediction that it is not or should not be about ego. I'm not pitching against the 'unknown' to get a buzz if I make a big win. I'm simply exploiting the 'known', the pattern or cycles of gyrations intraday (eg. of the Dow index).