how do you predict the future?

[Joules MM1]

......the general assumption is correct but the question ...remains......

My general answer is that you don't have to predict.....

anyone who enters a transaction, in the auction process, has an opinion or a prediction regardless
of how they reach their conclusion(s) no matter the speed, technical vehicle or philosophical approach......
.....a trader transacts in the present to receive a result in the future.....how that is not predicting is a complete mystery to me......

if a trader is placing a bet against the house without a prediction then what is the trader actually doing?

this almost invalidates the whole "hft are bad for the market" debate......everyone has
an auction opinion/prediction about things work and how theyre gunna work whether it's the next
micro second or the next quarterly.....

i predict the next Game of Thrones epi is a cracker

 

[counter_violent]

Sometimes I find it difficult to determine when a trend will continue and when it will stop and reverse. What if I enter a "pullback" but it turns out not to be a pullback but the top end of the distribution cycle where price ultimtaely stalls and starts to reverse the trend? How do you determine this? And where does the confidence come from? Is it just subjective experience again?

This makes no sense at all.

What do you mean when you say, "sometimes I find it difficult to determine when a trend will continue and when it will stop and reverse" ?
All this says to me is you don't know what is going to happen any of the time. You may just as well flip a coin with this kind of muddled thinking !

Then you say, "what if I enter a pullback but it turns out to be a pullback but the top end of the distribution cycle where price ultimately stalls and starts to reverse the trend".

Let me try explaining, see if it lands. When you enter a position in a new trade, you will need to have pre-identified level where you are wrong about the trade and this will be your stop out. Price does not need to hit this stop, you may close it early if you know that your assessment is wrong and that the price action is not conforming to your expectation of it.
Additionally you will need to know in advance what your expectation is of the trade. So you enter the trade with your expectation of the favoured outcome and the price action will need to continuously deliver in line with your pre-determined expectation.

Everything else is irrelevant.

 

[DionysusToast]

How cheap and sh*t is your dictionary?

Apparently, it gets better for $210 a year.

My impression is that some on here think that all predictions come true - otherwise they wouldn't have been predictions.

Perhaps another word would suit people better:

Prophecy
Forecast
Divination
Guess
Anticipation
Hunch
Prognostication

In any case, unless you are hedging, you enter a trade because you expect a certain outcome in the future. You expect it because you weighed up the current situation and decided that outcome was likely enough to put money on.

Like Joules, I can't for the life of me see why prediction is such a naughty word.

Trading is about making money when you guesstimate the future correctly.

 

[wackypete2]

Apparently, it gets better for $210 a year.

Peter

 

[DionysusToast]

This makes no sense at all.

What do you mean when you say, "sometimes I find it difficult to determine when a trend will continue and when it will stop and reverse" ?
All this says to me is you don't know what is going to happen any of the time. You may just as well flip a coin with this kind of muddled thinking !
.

C_V - maybe you can spell out how you know a trend will end or continue. It's a common problem to not know if a move is a pullback or a reversal. It's hardly muddled thinking for someone to not have a method of distinguishing one from the other. In fact, we both know there is no sure fire way.

For me, I personally look for the following to end a trend:

- News
- A rapid, extended move in the direction of the trend
- Participation - should be low on the pullbacks, so if you see what looks like a pullback with a LOT of participation, don't trust it

Now - if you are a daytrader looking for 4-8 ticks on the ES, the 'trend' on the daily or hourly chart isn't going to help you at all and would be best ignored. The word "Trend" itself means different things to different people. If I catch 8 points on the ES, I consider myself to have traded an intraday trend. It is these trends I try to capitalize on.

 

[DionysusToast]

Peter

gotta love teh interweb...

 

[bbmac]

'...So it seems you are giving an example, similar to the first few responses, of finding a trend and riding with it....'

There are other hi-probability trading opportunities that are not necessarilly ' with trend ' ie buying or selling at pre-identified potential support/resiatnce for eg and of course conversely buying at resistance / selling at support when the said resistance or support is potentially diminishing in strength and more liable than not to an upside/downside break respectively as may be shown by PA with HL's/LH's respectively leading up to the multiple tests of said resistance/support.


'...Sometimes I find it difficult to determine when a trend will continue and when it will stop and reverse. What if I enter a "pullback" but it turns out not to be a pullback but the top end of the distribution cycle where price ultimtaely stalls and starts to reverse the trend? How do you determine this? And where does the confidence come from? Is it just subjective experience again?...'

This is the age old question - is it a a. pullback, b. an extended pullback, c. consolidation, or d. an outright reversal in the trend in question. First of all making the working assumption that it it is a pullback in a trend will see you being wrong only once - ie when it isn't and the trend ends in favour of the last 2 of the 4 outcomes detailed above - so in a trend and particularly when the trend is strong (ie extends thru the t/f's above your 'longest' t/f make the working assumption that the trend will continue after an a. pullback or b. extended pullback.

Next thing is to find where the most likely places are that buy the dip/sell the rally participants will act again and join the trend at a discount in anticipation of it continuing. This I explained in my post above.

When looking to enter the trend after a pullback you need to look whether the intermediate and longest t/f's are showing signs of weakness/strength in their PA at the pullback lo/hi you are intersted in - ie for example let's assume there is an uptrend on your intermediate and longest t/f's and above and you drop down to the shortest t/f looking to buy a pullback to potential resistance becomes support/fibs/trend line on the longest t/f...but at the point price reaches the area you may be interested in on your trigger t/f you have a LL and on the longest t/f a L, - both these t/f's are showing signs of weakness and this may put you off getting involved in the trend again for now - ie either c. or d. of the outcomes listed above is now more probable than a with trend impulsive move to new highs...so you stay flat and/or await further confirmation re PA. Ie in entering an uptrend for example, after a pullback you may wish to see price at a potential HL on the t/f above which first exhibits the trend, - and this co-existing with a HL or at least a L on the t/f below that you have dropped down to time entry after the pullback into it's trend.

What may also help (although I would advise caution on this unless you ahve a lot of experience) is looking at the last trend hi or lo whether there are conditions on momentum oscillators that suggest an 'unwinding' or loss of momentum at that last trend hi or lo may lead to a pullback or more.

I think in any trading your edge - you need to have a sensible rule set based on historical precedent of what is highest in probability - and stick with it... this will help you to develop the confidence to act when it sets-up.

G/L

 

[bbmac]

Look at Gbpusd, a strong move up from point A to B (strong uptrend on 4hr and below) then followed a period of consolidation from B - C an upside break to point E and a false downside break to F that saw another break to point X - a pullback to G and another trend Hi at point H - pullback to point I a move to K and a testof I at point J and L before the next upside leg to the Friday Hi of 6279. Point X -L was effectively consolidation but look at the 3 candles I have highlighed and their importance ion showing the greater probability was for more upside. Of course point L - the 6279 Hi was again a strong uptrend on 4hr and below and dropping to 4hr you can see where price re-tested a previous Hi before going to a new trend Hi after the pullback. (circled.)

My point is that in the period A-B and after K when it was clear that the greater proabbility was for a continuation in the uptrend (ie strong bullish daily candles and in the period above K with similar - you make the working assumption that trend will continue - until it doesn't and just try and develop a methodology for getting involved with that trend whether that be a pullback methodology or some kind of momentum methodology or both.

G/L

 

[bbmac]

So, If I were looking to get back involved in this general opa uptrend on the Daily t/f, I would fib the moves from point 1, 2 3 and the average of points 4,5, and 6 (5803, 5807 and 5817) and along with the previous swing hi/lo zones, trend lines and topsides of breeched descending trend lines (wide dotted) I would work out the most likely area that buyers may re-emerge should there be a pullback...and then time my entry on 4hr (possibly even dropping down to 1hr)

If I was looking to get back involved in the co-existant 4hr trend in the same way I would drop down to 1hr (possibly even 15min) etc etc..

G/L

 

[Joules MM1]

...you..need to know in advance what your expectation is of the trade.

hmmm.....mebe expectation of direction and extent, not necessarily the whole trade, allowing for the unfurling of price rather than stamping a whole idea upon the unknown.....rather than just stamping "ABC should happen or i'm wrong"

Everything else is irrelevant.

again, mebe, this is telling unknown price and unknown activity of price how it should conform to your own ideas or experience

of course, i am speaking from my own perspective.....limited as it is......

 

[bbmac]

anyone who enters a transaction, in the auction process, has an opinion or a prediction regardless
of how they reach their conclusion(s) no matter the speed, technical vehicle or philosophical approach......
.....a trader transacts in the present to receive a result in the future.....how that is not predicting is a complete mystery to me......

When I enter a trade I have no opinion or prediction _ I am merly playing my edge...I do however have an expectation based on the up to date historical strike rate of the set-up I am playing but this is different from a prediction - a prediction would be entering the trade and saying that it going to do x or y. The expectation is that x % of these set-ups have produced a winning outcome in the past till that point but I have no idea whether the current one will be one of those successful outcomes or a losing outcomel as I cannot predict the future and have no idea the order in which they take place in the total forward sample. Let's say the strike rate is 70% (ie winning outcomes as a % of total outcomes of that set-up) then my ' expectation ' of a succesful outcome is 70% - but this is still not a prediction.

Anyways whether this is semantics or an issue of definition, the only reason I am labouring the point is really for the benefit of newbies who shouldn't go away from this or any thread with the impression that the future can be predicted. I accept though that folk do try to predict the future in many walks of life as DT points out but in trading it is not neccessary.

Note to self: leave it alone now - you ain't gonna change any minds Lol !

G/L

 

[Joules MM1]

When I enter a trade I have no opinion or prediction _ I am merly playing my edge...I do however have an expectation based on the up to date historical strike rate of the set-up I am playing but this is different from a prediction - a prediction would be entering the trade and saying that it going to do x or y. The expectation is that x % of these set-ups have produced a winning outcome in the past till that point but I have no idea whether the current one will be one of those successful outcomes or a losing outcomel as I cannot predict the future and have no idea the order in which they take place in the total forward sample. Let's say the strike rate is 70% (ie winning outcomes as a % of total outcomes of that set-up) then my ' expectation ' of a succesful outcome is 70% - but this is still not a prediction.

Anyways whether this is semantics or an issue of definition, the only reason I am labouring the point is really for the benefit of newbies who shouldn't go away from this or any thread with the impression that the future can be predicted. I accept though that folk do try to predict the future in many walks of life as DT points out but in trading it is not neccessary.

Note to self: leave it alone now - you ain't gonna change any minds Lol !

G/L

BB, not sure what made me post on this thread other than sunday night boredom......anyways; Positive Expectation!

....are you really saying that you came to the auction game with a neutral expectation? I doubt that. I think you came with a positive expectation as all good business people do. Period. You anticipated (then) you have the right plan. You anticipate (now) you are running the right plan. You anticpate youre outcomes are going to, on average, exceed your drawdowns. You implement and enter trades, based on forward positive expectation, whether it's based on statistical knowledge or a style that you know has a high(er) win rate.........


No individual can come to the auction process without first predicting that there is something in the game for them .......and they'll reach it....whatever that is.......

have a nice sunday

everyone predicts because no one knows

 

[The Leopard]

I will indulge one last time, I am not here to argue semantics yet thats all you want to do.



You think that the statistics 101 concept of variance is profound? Obviously you do, because you go on and try to justify your post with some trivial winrate/roi comparison. With 1:1 risk reward your winrate needs to be > 50%. with 1:2 risk reward your winrate needs to be > 33%. This is 9th grade algebra that you are trying to pass off as non-trivial.

Wait maybe you dont think its all that profound, because then at the end you admit that it is 'basic' after all.

Earlier you wrote this brilliance:



If you are expecting to lose, then you need to refine your system. Expecting loss and Accepting loss are two very different things. If you 1:2 r/r system wins at 40% obviously you are +ev. So you will accept losing trades 60% of the time. But if you go into the trade expecting to lose, then you need to re-evaluate your trigger.

But the thing is, I know you know this already. I'm not telling you anything new. You're just sloppy with your language because you want to have an argument to show your knowledge instead of a discussion of the underlying topic.



Stupid, perhaps yes, and thats probably why this thread quickly deteriorated. Only stupid because it riled up the trolls and prevented real discussion. I would rather describe it as 'sensationalized' since at its core the topic could be easily reworded as "how can you be sure of your edge?" But if I used that I probably wouldn't get many replies

Whether you want to argue over the word "prediction" or not, whatever your system is, you are saying that given your risk/reward parameters, you think that X will happen more often than not to give the system positive EV. Betting your money that something will happen more often than not, is a prediction. You are predicting that your model will provide success more than the 50%, 33%, whatever winrate required.

if you are using automated mechanical systems, presumable you've backtested these systems which gives you the confidence in your predictions. In this case, your answer to the topic could be as simple as "statistical historical analysis"

I can't be bothered to respond to this idiotic post in detail, and in any case you don't deserve a serious reply.

But I will just point out the part highlighted in red. If I take a type of entry that wins less than half the time, then yes I go into the trade expecting to lose. I am perfectly happy to do so, understanding that this is not about predicting (the future ). Why would I expect the trade to win when it does so less than half the time?

 

[DionysusToast]

Why all the insults, here?

People talk about the vendors killing the forum but I think the fact people aren't willing to debate like adults is much more likely to kill it off.

sigh....