Why it is possible to predict price movement?

predicting random events, sounds like a vegas strategy, altho vegas isn't random if you're on the right side of the aisle.

it's all about grasping the probabilities

Perhaps this is what you're after; all systems with a positive expectancy work because they are based on repeating patterns ie when xyz happens, abc usually happens too, that allows someone to buy and close at a higher level or sell and close at a lower level with profit that covers all losses too.

These patterns what ever they may be will vary from system to system but they are all based off observations of price movement, and price movement itself is based on the actions of the largest traders in the market, when they decide to buy or sell a pair (for whatever reason) that demand or supply will move the market. The price levels where they put their orders will determine what happens to price at those levels, ie if a support level holds because there is a huge amount of buy orders triggered.

So the question now becomes who are the biggest traders and how exactly do they trade? Well as of 2011 77% of volume on currency trading went through DB, Barclays, UBS, Citi, JPM, HSBC, RBS, CS, GS, MS (wikipedia). These are all banks/investment funds. At the top of my head from glancing at the screens of institutional traders from docos and shows and reading interviews of ex traders and risk analysts and what not, they use the same tools that average guys use, like moving averages, macd, volatility measures, fibs and pivots, as well as SR lines, candles and price patterns (on top of fundamentals).

So perhaps the reason why these things work...is because the largest market participants use them, creating a self reinforcing usage, or a self fulfilling prophecy. Take for instance the start of candle theory in Japanese rice trading, special candle formations during trends would represent a shift in buying or selling sentiment, obviously these candles simply represent the records of price, but the more speculators that knew of these certain signals/patterns would then trade on them greatly increasing their effect. Another example is the birth of pivots in the pits of the CME, some traders realized that price usually hovers at certain levels proportional to the previous day, and formalized these relationships in the formulas of the pivots we know today, word spreads, and now many use them. And remember that people will usually tend to focus on the times where they work rather than don't work. So now consider fibs, after every large market move there is usually a retrace of a certain %, now applying fibs of 23, 38, 50, 62, 78...that's 5 lines drawn spaced out...price will definitely retrace to one of them if they "did or didn't work", but traders will focus on the times they work, word will get out, more will use them until so many are using them it represents something self fulfilling.

My 2c, in the old pic below these things are everywhere.

some would say that technical analysis is useless, and that one must understand macro economics and anticipate what causes devaluation and appreciation of an asset.

There are studies indicating however that Tech analysis is valid to a point, conducted by the university of Pittsburgh " The predictive power of candles"

another research by the University of Columbia, among others.

what i consistently see is that people over estimate the order of the markets, and people see patterns where there are none.

60% of price movement is completely random
while 40% is actually information based.... I believe Berkeley and George town research papers show this in "the market micro structure" and the other paper escapes me at the moment.

Random walk theories and Brownian motion, stochastic process, wiener's motion... are actually beneficial theories to look into.

inter market correlations and calculating markets beta are also pretty useful.

What I personally find in trading as well as most of the replies is basically a calculated gamble with probability and positive expectation over a large sample space is the best we can hope for.

Yes tendency is what I meant to say just as stated.

to say trend first we have to define what a trend is, I cannot seem to find two traders to agree on what a trend constitutes.

[quote=kk007;5625929]This sounds a simple question, but it is not indeed. There are a lot of trading methods claim working. Nevertheless, I wonder in general what logic these methods based on in predicting price movement.

Can anyone give me a good answer?

"Sometimes" the market "tends" to respect: support resistance, fibos, ovebought or sold, Elliot ..etc......and the hundreds of different "future telling" toys we all use...
Between "tends" and "sometimes",you play the game, most important: position size to stop loss risk..add to that stop hunting, spread and spikes , BINGO it´s "almost" impossible to make anything better than BE...
STILL, the very skillfull, "can" create an edge, mostly tiny...play it enough times...youré rich...
Imagine a big gorilla at the entrance of richness,successfully punching every one away, and having fun in the process...

It seems that our approaches are conceptually similar.

That's a very broad question, at the most basic level I suppose because people look for optimal levels to buy and sell.

I'm not sure anyone really understands your question, are you asking from a technical perspective, a psychological one, a mathematical one, a logistical one......?

Agreed. It is all about orderflow. Now, where are the majority of buy and sell orders located? Without market depth info, we can only come back to price: support and resistance. Because S/R has become so pervasive in its use that it creates a kind of self-fulfilling prophecy. Specifically, confluences of different S/R types (recent candle highs/lows, session/daily highs/lows, trendlines, pivot levels, round numbers, fibo retracements) all within a few pips of each other — that is where price almost inevitably stalls, and occasionally reverses.

On the thread topic: does this mean that I can predict price movement? Absolutely not. But if (for example) I'm selling on rallies into close-knit resistance barriers during a strong long term downtrend, then I have both directional bias (the downtrend) and behavioral bias (S/R creating a barrier for low risk entry that makes a high RR trade possible) improving the probabilities in my favor. I don't need to be able to predict price movement, so long as I can use these biases to create a method of entries and exits that, over time, wins more money than it loses (aka positive expectancy).

But of course that is just one approach that happens to work for me; there are many others, too.

Take a look at these charts below... Notice any patterns...? Really...?









These charts have been generated using excel's rand() function (generates random number). My point being, if price action was random (which I don't believe it is), these same patterns of support/resistance, trend lines, fibonacci levels, etc. would still occur, though perhaps not as often as they do occur.

Interesting! Could you tell me how to make those charts?

Here, press ctrl+= (cmd+= on mac) to recalculate

Looks similar to what I have posted before.
http://www.forexfactory.com/showthread.php?t=227918

Reminds me of another thread started by kk007.

Do you start these threads just for the fun of it or do you actually want to know something?

Not for fun. Serious thread here.

Wait, here's a question kk007, you've asked why it is possible to predict price movement, and you've received a bunch of responses which revolve around repeating patterns, where big traders places their orders and none of it seems to be satisfying your query...how about we flip the tables and ask; "What would price movement have to look like to you, for it to be impossible to predict?"

hello jeanlouie,

I must tell that most of your assumptions are wrong, I work on a FX desk in a bank as a trader and I know how this stuff with orders works and it is not in a way as you described,

orders on futures are not so important because they are relative small and they are not affecting the spot price that much, the thing that really matters is orders on spot FX market which is decentralized OTC market so that you dont see the orders and transactions.....only thing you see if you work as a spot trader in institution are your own orders from clients in bank's orderbook....also you can see actual bid/offer sizes for current price in Reuters and also market depth/liquidity information on EBS, but it still is not the stuff to help you at all....if you are retail trader you see actually nothing

about the trading style of market movers - as you noticed that they look on the same stuff as another classic trader - mostly wrong, I know few spot fx traders from GS, DB, Citi who have the custy flows which are moving the market ..... yes true they look sometimes on some technical S/R or TL but its only for "playing" with really very small ammounts which in no way move the market - the situations when they are playing big and when the most alpha is generated is after some big custy flow from HFs or corps and they only ride the actual flow nothing more, all you can see on the screens is 1-2 screens maximum with line charts, 2 screens for orderbook, position view, internal systems and other 2 for some BS stuff and messenger

i agree that there are big HF players who take into consideration technical analysis when making decisions - but if they do you will definitly see the fakie and maybe get stopped, when buying 500mio u just cant afford to buy it on the support level where is no liquidity at all

i hope this helps,

best regards

I posted a chart before the fact indicating a price and time zone I was looking for a turn in the EURUSD. On cycles it is about the only time I use fib relationships, this turn was at 127% of the last swing, but the price and time I calculated before was just 39 pips out. If you had waited for the 127% swing you would have nailed the turn.

So when people here say price is random you can see why I don't agree with them, if this were so I could not predict where the EURUSD was going. So for my next trick I think I have to predict its next move, I already have thought about this and what I think is that we take out the high and get to 1.33266 by the 5-6th May. Let see.

Mike Haran

I think the question is partly answered. I am now thinking how to reply again here.

Just fund your account then any prediction is covered...