The Perils of Prediction

[barjon]

DBPhoenix has talked about a fundamentally important aspect of planning in the Where is the DOW heading thread and because it has such wide application I thought it worthy of separate debate.

Here's what he said:

I haven't participated in the thread because of its nature, but if as you say there are a lot of beginners reading it and posting to it, I suggest they scroll back to gmca's post which, to me, is central to the difference between predicting and planning and which also prompts so much of the flaming that goes on in threads of this type.

Predicting involves the ego. I think this or I think that. The X will go up to wherever then down to wherever else before bouncing and doing thus and so. Once one steps onto this particular path, he puts himself into the position of being right or wrong, and that's where an enormous number of beginners' problems begin.

gmca plans. He anticipates. He considers every contingency he can think of. In this way, when the trading day begins, he can -- with practice -- trade objectively and rationally, not by suppressing or repressing emotion but through nullifying the typical emotional triggers. By being available to whatever happens, he can act while others dither and deny and rationalize.

The desire to know what's going to happen next is for the vast majority of people a need, and not just in the trading arena. But in the markets, it's death, whether trading technically or fundamentally. One doesn't learn how to let all of that go in a weekend seminar or correspondence course, but it's essential to long-term success.
_
DB_________________

and here's the post from gmca686 that Db refers to:

Can I pose a question for discussion?

Surely the title of the thread and many of the posts are misleading, because they invite us to think where the market is going to go?
I have read on many occasions the comment "I think the market will do x y z." The market doesn't care what we think, and personally if I find myself thinking what the market might do I give myself a sharp slap on the wrist. This has come about by the aftermath of several painful lessons when the stupid market behaved incorrectly after I thought it would go in a particular direction.

Surely our time would be better placed analysing along the following lines.
If the market does "A", I will do "X"
If the market does "B", I will do "Y"
If the market does "C", I will do "Z"
etc, etc

We need to establish what the market triggers ABC are,
and sort out our personal trading plans in XYZ.
I think this approach could provide for a more beneficial thread.

Now, am I missing something? do contributors feel that this is what the thread does, or do you agree with me that there is too much "I think"
regards, G McA

There, that's set the scene - I'll leave it there and comment in the next post

good trading

jon

 

[barjon]

Wise words indeed Db and gmca. I agree, but I suppose we wouldn't enter a trade at all if we didn't have some expectation that it would move in our favour. So isn't that prediction of a sort?

Maybe we're really talking about the difference between a "prediction" based on the statistical probability of X following Y and one that isn't?

We can never trade the past, only a future which is always unknown.

good trading

jon

 

[DaveJB]

There's a difference, created by how we use the word 'prediction' I would say...
Many people (most?) think of a prediction for a stock, index etc as some sort of fortune telling event based on any number of clever programs/methods etc, whereas gmca's 'prediction' is a sequence prepared for all expected/likely situations where the trader's actions will tend to be robotic, based on (essentially) statistical probabilities. As that probably baffled 2/3 while the other 1/3 are nodding -
Prediction 1 - Using my wonderful software to predict turn points (I'm making this forecast up, by the way, not really using Delta, Reversal Magic, Market Matrix or whatever)..... The DOW is due to continue down until Wednesday, when it will bottom at 10200 and then climb to a top on June 6th of 10630.
- This is what many think of as a prediction.
Prediction 2 - The Dow is currently at 10471.91 on a down bar. It looks like it is consolidating during an uptrend that started in mid April. 10400 is an obvious support/resistance level, the uptrend has faltered as it goes to break the 10500 level (which has not been particularly effective as s/r so far, 10550-10600 has been more evident for that). If the Dow drops below 10460 it will take out an intraday bottom, therefore I will short it as it will then be in an intraday downtrend... 10450, 10435 have provided support intraday so I will watch for reversals on the way down especially in those areas. My exit strategy will be to exit on a stop placed 5 pts above the top of the current bar, moving down as new lower highs are set. If it climbs through 10490 it will have set a new high on the intraday chart showing the intraday uptrend is still in effect, I will enter long and place my stop 5 pts under the low of the current bar (trailing up as before).

We'd then have 'what to do if it gaps up or down' and a few other ideas in there probably.

Prediction 1 is fortune telling, which some swear by etc and I make no value judgement of it, but it's significantly different to that seen in 2 - in '1' we are saying that we know what will happen, and in the worst case we go to trade it even when the chart clearly shows us that the prediction has gone wrong....the method calls for a downturn on a specific bar, and people trade the short even though it's still very obviously still in uptrend.

in '2' we are admitting we don't, but we do know how to react to what occurs and extract profit from it. Frequently the difference is made obvious by the fact that people take positions ahead of predicted turns using method 1, whilst those using method '2' will generally never do that - they'll react to the turn once it's in.... a misconception is that method '2' gets you in 'too' late.

Dave

 

[roguetrader]

the difference between predicting and planning

All trading is prediction to one degree or another, to not accept that is imho a grave mistake. Assuming you can accept that, begs the question Where are we going with this thread?
If of course you disagree of course then we have a discussion.
Dbp refers to...

the difference between predicting and planning

as if they were some how easily confused.

For example I predict that we will have a trend day up tomorrow, with the Dow rising some 100 points.
That is a prediction, probably not a good one, but a prediction none the less. Now surely no-one trading would confuse that with a plan. A plan in this particular instance would be born out of that prediction, with a view to confirming or dismissing that prediction. If there is no prediction then there is no trade. The only way you can avoid predicting, apart from not taking part of course, is to Know what is going to happen. And if you've mastered that little trick then you are indeed very clever, or soon to be investigated for insider trading.

 

[DaveJB]

Okay rogue,
to a point- but only in that the alternative is to accept that charts are purely random and don't exhibit repeat behaviour. I suppose the main difference in my view is that the fortune tellers will trade a position despite what the chart is actually showing at the time - they will go short in an uptrend, one who 'reacts' to what the chart is doing will be trading the current move, not the next one.

Dave

 

[roguetrader]

Okay rogue,
to a point- but only in that the alternative is to accept that charts are purely random and don't exhibit repeat behaviour. I suppose the main difference in my view is that the fortune tellers will trade a position despite what the chart is actually showing at the time - they will go short in an uptrend, one who 'reacts' to what the chart is doing will be trading the current move, not the next one.

Dave

That rather depends on what the fortune teller is using to make his predictions, most "good" fortune tellers merely tell you what you tell them, they ply their trade by being excellent observers. The salient point from what I said before is that unless you know, then you are predicting. When I place an intraday trade, I have a concious prediction in mind of where i think the instrument I am trading is going, supported by all the things that I watch in the market with a plan to manage that position, however since I know that my trade is based on a prediction, however sure I may be of that prediction, I have a plan to extract myself if it all goes Pete Tong. In knowing that you are trading predictions you know and accept that you can be wrong.
The concept of chart patterns is simply that they are "predictive" as such they can be wrong also.

 

[SOCRATES]

Which takes us back to the same discussion/debate/argument that has been going on for thousands of posts.

Forgive me for pulling a Socrates, but I don't know how to make this any more clear, which is my failing. All I can suggest is that those who are interested read what I've written until either they get a headache or the headache they already have goes away.

Note: Women rarely have trouble understanding this since they don't invest their egos in their activities nearly as much as men do (that is, practically never). Athletes may also understand the difference between focusing on winning and focusing on playing well.

But I hadn't planned on instigating an entirely new thread. I just wanted to provide something to think about, not necessarily "discuss".

Hummm....

 

[kriesau]

That rather depends on what the fortune teller is using to make his predictions, most "good" fortune tellers merely tell you what you tell them, they ply their trade by being excellent observers. The salient point from what I said before is that unless you know, then you are predicting. When I place an intraday trade, I have a concious prediction in mind of where i think the instrument I am trading is going, supported by all the things that I watch in the market with a plan to manage that position, however since I know that my trade is based on a prediction, however sure I may be of that prediction, I have a plan to extract myself if it all goes Pete Tong. In knowing that you are trading predictions you know and accept that you can be wrong.
The concept of chart patterns is simply that they are "predictive" as such they can be wrong also.

Without intending to be offensive - isn't this all becoming a bit anal ?

I agree with you RT - ultimately we are all trading on the basis of some form of prediction, however it is arrived at. A plan gives the trader certain parameters in respect of when and how he might enter, manage and exit a trade. Certainly it is desireable to minimize the emotional or egoistic input into that predictive equation but you will never eliminate it on an individual basis. The only way to completely eliminate it is to program some software with a variety of signals and indicators and let it trade for you.

Another point that has not been mentioned is the impact of external events that technical analysis cannot embrace, even comparatively minor events - like a couple of innocuous words left out of a Fed statement which is then amended and issued an hour later - can turn an intraday market almost instantly and move it up or down 50 pts in the space of 5 minutes.

Bottom line is that we are all trying to improve the odds in our favour on both individual trades and in terms of our ratio of successful trades. We employ historical statistics, repetitive yardsticks, fundamental analysis etc etc in an attempt to do this but no infallible system or combination of factors exist.

Lets not get into the realms of another esoteric debate where absurd claims are bandied about in relation to 'higher conscienceness", "pure thought", "supression of the id" and the "eternal search for truth and enligtenment". That might be okay for monks in a monastry seeking the meaning of life or the divine revalation but it has little to do with the mortal reality of trading !

 

[TheBramble]

Note: Women rarely have trouble understanding this since they don't invest their egos in their activities nearly as much as men do (that is, practically never).

Mr. B. I'd be utterly fascinated for you to provide a source or reference for that statement.

If it's just personal opinion - fine, but if so, you and I obviously don't mix with the same sort at all, at all, at all.

While the Alpha Male is a stereotype (quite literally) - he has rarely, in my experience, been any match for the equivalent Alpha Female operating in the same patch/territory/arena/corporate estate etc.

Before the anti-misogynist camp (who aren't always misandrists themselves) saddle up to do battle with me, let me make quite clear my absolute devotion and adoration of the fairer sex and their generally far greater abilities in so many areas that I would welcome even a fraction more ability and skill than I currently posses.

 

[SOCRATES]

Without intending to be offensive - isn't this all becoming a bit anal ?

I agree with you RT - ultimately we are all trading on the basis of some form of prediction, however it is arrived at. A plan gives the trader certain parameters in respect of when and how he might enter, manage and exit a trade. Certainly it is desireable to minimize the emotional or egoistic input into that predictive equation but you will never eliminate it on an individual basis. The only way to completely eliminate it is to program some software with a variety of signals and indicators and let it trade for you.

Another point that has not been mentioned is the impact of external events that technical analysis cannot embrace, even comparatively minor events - like a couple of innocuous words left out of a Fed statement which is then amended and issued an hour later - can turn an intraday market almost instantly and move it up or down 50 pts in the space of 5 minutes.

Bottom line is that we are all trying to improve the odds in our favour on both individual trades and in terms of our ratio of successful trades. We employ historical statistics, repetitive yardsticks, fundamental analysis etc etc in an attempt to do this but no infallible system or combination of factors exist.

Lets not get into the realms of another esoteric debate where absurd claims are bandied about in relation to 'higher conscienceness", "pure thought", "supression of the id" and the "eternal search for truth and enligtenment". That might be okay for monks in a monastry seeking the meaning of life or the divine revalation but it has little to do with the mortal reality of trading !

But you are being offensive to the members who are discussing a topic that interests them, very offensive indeed, if I may say so. This is particularly relevant as your opinions are off topic as well.

I am not interfering, just commenting, because I am only engaged in quiet contemplation of how this discussion develops, as it is already very interesting and promises to become more interesting still as it develops further.

 

[kriesau]

Actually, there's nothing esoteric about it, nor does it have anything to do with "higher consciousness". But, for some reason, that's where these discussions always end up. Which is too bad, because trading without ego would I suspect be relatively easy for most people if they knew how to train themselves to do so.

Well I guess that it depends whether you view it as some kind of exact science or a refined art.
Every human has an ego, some let it run riot (no names and I don't mean you either) and others control it in relative degrees, which will vary from situation to situation. But I really don't buy this "search for the holy grail of eliminating the ego 100%".

a) I don't think that it's possible
b) I don't think that it's necessarily desirable

If that is the "Holy Grail" of trading then I'm sure it can be achieved by correctly programming a computer to trade for you and I would be willing to bet that it's success rate would not beat a talented and successful individual trader.

 

[peterpr]

DBP hits the nail squarely on the head. The central issue posed by Jon's starter really is ego. We can banter back and forth till the cows come home about whether a decision to enter (or exit - or even continue in) a trade constitutes a prediction or not, but the banter would be semantics pure and simple.

Old Charlie Dow summed it up when he said "Pride of opinion has been responsible for the downfall of more men on Wall Street than anything else." - or something like that.

All IMHO of course, rather than IMPO

(We need a smilie for 'tongue-in-cheek' methinks, so a wink will have to suffice.

 

[TheBramble]

If that is the "Holy Grail" of trading then I'm sure it can be achieved by correctly programming a computer to trade for you and I would be willing to bet that it's success rate would not beat a talented and successful individual trader.

That's a really interesting thought.

While not exactly along the lines of a Turing Test, I have an intuition that if it were possible to write code to exactly handle a good trader's style, methods, plan, entry/exit, risk & money mgmt etc. AND the traders ability to modify his plan, systems/methods etc. over time in the light of changing conditions, markets etc. - the whole lot - the automated system would win out in the long run over the human upon which it is modelled.

Why?

Because 'discretion', ego, whatever will creep into any traders fingers now and again. Not necessarily to the point of wiping them out or even denting them too much - but just the occasional 'punt', 'gamble', 'intuition' etc.

Although my first-hand relationship with traders amounts to less than a dozen souls (all of whom have been 'at it' for a while - successfully I might add) - none of them have stuck rigidly to their rules all the time. Myself included.

There may be that one Mega-Meta-Trader who has and does - and there'll be a whole bunch who claim to - but I maintain - a machine if correctly and fully programmed would do a better job.

It would also leave you more time to gaze at your navel....

 

[SOCRATES]

This is getting better by the minute...

 

[kriesau]

DBP hits the nail squarely on the head. The central issue posed by Jon's starter really is ego. We can banter back and forth till the cows come home about whether a decision to enter (or exit - or even continue in) a trade constitutes a prediction or not, but the banter would be semantics pure and simple.

Old Charlie Dow summed it up when he said "Pride of opinion has been responsible for the downfall of more men on Wall Street than anything else." - or something like that.

All IMHO of course, rather than IMPO

(We need a smilie for 'tongue-in-cheek' methinks, so a wink will have to suffice.

So to lets get to the point:

1. What exactly is the desired state of mind that you are seeking to arrive at and how to do you expect to achieve this ?

2. Why not focus on setting up either a computer or a robot to trade for you, since neither has an ego and can be specified with a high level of artificial intelligence, and then you will have reached your nirvana in this particular context !

3. How would you (if you achieved this ego free state) deal with non technical market movers which can happen in an instant and may recquire a fairly instantaneous response ?

 

[kriesau]

That's a really interesting thought.

While not exactly along the lines of a Turing Test, I have an intuition that if it were possible to write code to exactly handle a good trader's style, methods, plan, entry/exit, risk & money mgmt etc. AND the traders ability to modify his plan, systems/methods etc. over time in the light of changing conditions, markets etc. - the whole lot - the automated system would win out in the long run over the human upon which it is modelled.

Why?

Because 'discretion', ego, whatever will creep into any traders fingers now and again. Not necessarily to the point of wiping them out or even denting them too much - but just the occasional 'punt', 'gamble', 'intuition' etc.

Although my first-hand relationship with traders amounts to less than a dozen souls (all of whom have been 'at it' for a while - successfully I might add) - none of them have stuck rigidly to their rules all the time. Myself included.

There may be that one Mega-Meta-Trader who has and does - and there'll be a whole bunch who claim to - but I maintain - a machine if correctly and fully programmed would do a better job.

It would also leave you more time to gaze at your navel....

Well we will never know until someone actually tries it in a controlled experiment.

Many large institutions use programmed trading as we all know but that usually is focused upon managing trades. Human input is still the prime controlling mechanism.

Modern aircraft are flown by computers but we still have pilots (not just as PR to reassure the passengers) but to make and implement decisions where an immediate interpretation of sudden events is required.

It all boils down to optimizing the tools that you work with to try and achieve the best result.

 

[peterpr]

So to lets get to the point:

1. What exactly is the desired state of mind that you are seeking to arrive at and how to do you expect to achieve this ?

I am not seeking it. I have achieved it. Namely an ability (however imperfect) to trade on pure price action aided by both prior and real-time analysis of where the obvious decision points are likely to be - by that I mean knowing who I am trading against and where they are likely to place such decision points

2. Why not focus on setting up either a computer or a robot to trade for you, since neither has an ego and can be specified with a high level of artificial intelligence, and then you will have reached your nirvana in this particular context !

I have spent a small fortune on software over the years, including a ridiculous amount of time focussed on just that; not to mention my own coding. I have come to regard technology simply as a tool. As for 'Nirvana', I am not a Bhuddist and therefore do not accept that it (or rather my understanding of it) has any validity wrt trading..

3. How would you (if you achieved this ego free state) deal with non technical market movers which can happen in an instant and may require a fairly instantaneous response ?

Don't see why that should be an exception.Watching for news related price action is already an integral (and pretty fundamental) part of my trading approach. As is being on top of all the historical news based factors that I judge likely to affect other traders

I am simply concurring that, to allow one's ego to influence the mechanics of trading is to invite a sharp and expensive lesson in just how insignificant you really are.

 

[barjon]

That's the trouble with conversations of this sort - we've got to use words and there are always some of those words that are like red rags to a bull

Personal experience:

1. Became interested in TA and analysed market.
2. Decided where it was going (prediction) and entered.
3. Watched in disbelief (ego) as it didn't.
4. Moved stop to "give it more room" (ego attachment to prediction)
5. Lost (dented ego)
6. Repeated, repeated, repeated.
7. Sought a more mechanical system and drew up specific rules.
8. Tested it to determine probabilities etc (statistical prediction).
9. Set entry trigger orders in advance (non-ego)
10. Set exit trigger entries in advance (non-ego)
11. Won (ego massage)
12. Started second guessing system rules (ego prediction)
13. ..........guess what
14. Saw sense.
15. Bounce between 12 and 14 repeatedly.

So yes, call it what you will - ego, pride, obdurateness, predictive silliness or whatever - "it" got in the way to my trading detriment. For me, I'm wedded to a fairly mechanical system to avoid being influenced by emotion of one sort or another and even then I need to work with advance orders because I can't trust myself to always pull the trigger in real time.

So if ego/emotional attachment is the peril of prediction then I must give myself a sharp gmca slap on the wrist if I succumb to it - but it is fun

good trading

jon

 

[SOCRATES]

Yes, that's right...and now teatime...

 

[kriesau]

Taking 2 first, trading is far more interesting when one can be a detached (not disinterested) observer. If nothing else, the SAR tactic (not to be confused with the indicator) takes on a whole new aspect.

As for 1, I suppose it's a matter of "when the student is ready, the teacher will come". There are books that had me banging my head against the wall from the insights I was gaining that had most other people saying "so what?" And there are books I think are total cr*p that others would grab first if their houses were on fire. If I'd read those book at some other time, would they have sparked such recognition? Probably not. Which is why time has its own role to play independent of experience.

As for 3, it depends on whether or not I'm already in the trade. If I'm in it, the managment doesn't change. If I'm not, I'd elect not to participate because I'm not good at it. Others love trading that stuff and more power to 'em. But I suspect that those who are good at it are successful because they thrust themselves into the flow rather than try to predict the direction of the flow. Or maybe not. I don't circulate among that crowd so I have nothing especially insightful to say about it.

"Trading is far more interesting when one can be a detached observer"

Is it ? Why are you trading in the first place ? Is it purely to make money and if so maybe a robot would be better at it than you or me ! But what about enjoyment, achievement and fulfilment !
Bill Gates has more money than he can ever spend. I'm sure that he enjoys it. He could stick in in the bank or a mutual fund and that would probably produce $2 billion in income every year for no effort. What a hedonistic lifestyle that could bring you. But he doesn't do that - why, because he is human - he thrives on new challenges, personal achievement, risk and helping others. Do you really think that he would find life more interesting if he sat back and just watched some computer or fund manager optimize his yield !

"When the student is ready the teacher will come".

Sounds like some religious revivalist session.

"It depends on whether or not I m in the trade. If I'm in it the management doesn't change, if I'm not I'd elect not to participate"

Ah but what would you computer decide to do ?
It's Oct 2005 - without warning China announces that it is revaluing the Yuan by 20%. You are short on the Dow, the market shoots up - you're stopped out in minutes. Do you do nothing or do you go long and try and take advantage of this sudden development ? Hey this might be exciting ! Oh the computer has not responded and I'm out of the market - okay, but never mind that was quite interesting wasn't it !

Thats an interesting life !!!