If we could represent forex in equations. | Forex Factory

Options

Permalinks
Print Thread

Similar Threads

Simple math equations and their applications on charts 3 replies

Differential Equations With Forex! 6 replies

What do Candlestick Wicks Represent? 14 replies

Use MAs to Represent Higher T/F Trend on M1 3 replies

How can I represent motion around fibs with continuous variables? 0 replies

Attachments: If we could represent forex in equations.
Exit Attachments

If we could represent forex in equations.

  • #1
  • First Post: Jun 7, 2007 1:08pm Jun 7, 2007 1:08pm
I heard that every graph and line can be written as a function. The equation could be extremely complication but what if we can find it anyways? If there was a super computer that can calculate all the historical graphs and charts, we could be able to find some rules or something.
When I buy it, it moves down. When I sell it, it moves up.
  • #2
  • Jun 7, 2007 2:15pm Jun 7, 2007 2:15pm
Even if you could, the past cannot be used to predict the future.

andersenws
"Youth is the trustee of prosperity." - Benjamin Disraeli
  • #3
  • Jun 7, 2007 9:13pm Jun 7, 2007 9:13pm
Quoting andersenws
Even if you could, the past cannot be used to predict the future.

andersenws
Yes it can, history repeats, behaviour repeats, patterns repeat, perhaps not 100% reliable, but certainly enough to provide an edge
  • #4
  • Jun 7, 2007 10:12pm Jun 7, 2007 10:12pm
I agree and do use patterns in my trading, but the whole notion of using supercomputers and deriving equations for price is ludicrous, and that is what I was referring to.

andersenws
"Youth is the trustee of prosperity." - Benjamin Disraeli
  • #5
  • Jun 7, 2007 11:19pm Jun 7, 2007 11:19pm
This is absolutely true...

If I gave you the following facts:
f(1) = 30
f(2) = 35
f(3) = 67890

You could easily represent f(x) as a function...the issue is that the function would be an extrapolation of the 3 points, but it would be exact. SO...to answer your question, if you took a price chart and plotted 128,000 data points, you could represent that series as a function of price...however, it would be completely dependent on all 128,000 price points (in other words, a function that I REALLY wouldn't want to try and write down). Where the function breaks down will simply be the 128,001st data point that doesn't exist yet.

The only way you can prove that an algorithm is true is to be able to prove that for k + 1 data points, the value of f(k + 1), which wouldn't necessarily be true...which automatically invalidates the function...

So, while computer models could make decent representative graphs that would come close to 100% accurate, but it still wouldn't be 100% accurate...that doesn't mean that it wouldn't be effective enough to be profitable...that's a different issue...BUT, you wouldn't be able to describe the future points on a forex graph with a function accurately.

Quoting pujolshomerun
I heard that every graph and line can be written as a function. The equation could be extremely complication but what if we can find it anyways? If there was a super computer that can calculate all the historical graphs and charts, we could be able to find some rules or something.
  • #6
  • Jun 7, 2007 11:25pm Jun 7, 2007 11:25pm
Quoting pujolshomerun
I heard that every graph and line can be written as a function. The equation could be extremely complication but what if we can find it anyways? If there was a super computer that can calculate all the historical graphs and charts, we could be able to find some rules or something.
It's curve fitting. Patterns exist in stocks, forex and other financial instruments. It means general patterns/trend can give you assistance in your decision making. But to have a supercomputer that take account of every point/price and formulate a function for future prediction, it won't work.
  • #7
  • Nov 1, 2007 10:31pm Nov 1, 2007 10:31pm
Quoting pujolshomerun
I heard that every graph and line can be written as a function. The equation could be extremely complication but what if we can find it anyways? If there was a super computer that can calculate all the historical graphs and charts, we could be able to find some rules or something.

The question is why would you need a super computer to perform an estimation for the equation of a price chart. I think the computer that are available to us today can be used to estimate the equation. The good reason for this, is that the solution can be used to keep graphing the price, which converts to prediction, that is how i see it. Is that wrong? I just got done with taking a differential class, in which ordinary differential equations are used to describe electrical/mechanical systems. A price chart can be seen as a mechanical system. I suppose the only way to esitmate the equation of a price is to use matlab.
  • #8
  • Nov 1, 2007 10:48pm Nov 1, 2007 10:48pm
I have thought abt this before. The idea was to find frequency of highs and lows...(i did not do it in thte end coz its all too time consuming and i am happy with my current approach)

I wanted to make and intraday trade and wanted to see at what time of the day will the CABLE usually make a High and Low.

I collated a bunch of data and planned to make a histogram to come up with a frequency table. The plan was

- to look at the overall data(1year) and see which period over 24 hours has a strong tendency to create a high (good strategy to approach when u are trading the trend, you you are literally selling rallies / buying dips)

- to look at the quarterly data and see which periods usually dominate the highs and lows.

I did not plan for this to turn into a system but instead a rough guage for when i can expect a turn in tides and look for reversal signals.

There are also a few other factors to note, i'd note that if the london hour made a price high today, i'd expect a different session to make the high the next day, you know somethin along those lines, to come up with reasons for probabilities

I did not get abt to finish the study coz i was (plainly said) too lazy, ahahaha. well maybe one day. If not would anyone be interested in doing this?

regards
jest
  • #9
  • Nov 1, 2007 11:42pm Nov 1, 2007 11:42pm
HERE,

this is what matlab can do, i plotted the points and did curve fitting with 4th degree polynomial, i can go as high as 10th degree on this view, but internally i can go as high as i want. How can this information be used?

You can see from the chart, that y=equation of the EURUSD price chart!
Attached Image (click to enlarge)
Click to Enlarge Name: EURUSD EQUATION ESTIMATION.JPG Size: 44 KB
  • #10
  • Nov 1, 2007 11:51pm Nov 1, 2007 11:51pm
Quoting c0d3
HERE,

this is what matlab can do, i plotted the points and did curve fitting with 4th degree polynomial, i can go as high as 10th degree on this view, but internally i can go as high as i want. How can this information be used?

You can see from the chart, that y=equation of the EURUSD price chart!
Arghhhh....
THE Absolute Biggest Culprit in costing Freshmen Traders Money!
I understand you "Fitted" the Results but still, the words make my blood curdle.
  • #11
  • Nov 2, 2007 12:54am Nov 2, 2007 12:54am
Quoting Bemac
Arghhhh....
THE Absolute Biggest Culprit in costing Freshmen Traders Money!
I understand you "Fitted" the Results but still, the words make my blood curdle.

noone is using this to trade with real cash, just a discussion!
  • #12
  • Nov 2, 2007 12:18pm Nov 2, 2007 12:18pm
Quoting pujolshomerun
I heard that every graph and line can be written as a function. The equation could be extremely complication but what if we can find it anyways? If there was a super computer that can calculate all the historical graphs and charts, we could be able to find some rules or something.
I am of the belief that the markets are a chaotic entity, which exhibits fractal behaviour. Take a look a fern, the small bits are similar in appearance to the large bits, or a tree for that matter. Isn't a trunk just a big branch?A branch just a big twig?
Now look at the daily chart of any high volume, liquid market.Now look at the 4 hour chart. Now look at the hourly.Now look at the 15 min etc.
The most obvious thing is that the market displays this approximate "self-similarity" on any time frame.

This simple observation precludes the use of Euclidian devices to quantify the market. This is why (I believe) a trend line really means nothing, not withstanding the self-fulfillment argument.
There's a seminar video by Bill Williams floating round on the web which talks about this stuff. He even cites research that states markets display the same fractal properties as rivers.(Interesting stuff)

I'm personally interested in such natural phenomena. Ever asked yourself why the ratio of a circles cicumference to it's diameter is Pi?Why can the decay of radioactive particles be described by the exponential? There are those who cite God/Allah/Father Christmas etc as the reason for these, but they are not correct. When we look a these numbers we are looking at the very fabric of nature itself,distilled from the moment the Cosmos appeared from the "ether". In looking at these numbers you are witnessing God (if you choose to believe, I don't) these things do not describe creation, they are creation!

Heres the problem though, what's that all got to do with the price of fish?
Even if markets do display similar characteristics to natural occurrences does that mean that we can predict? I think not. This would require rigid definitions, like in Classical physics. X bumps into Y, Y does this, then moves away at Z speed, that sort of stuff. In fact Newton himself rubbished the idea of predicting the markets during the South Sea bubble.

Regards,

Ed.
  • #13
  • Nov 2, 2007 7:17pm Nov 2, 2007 7:17pm
Quoting c0d3
A price chart can be seen as a mechanical system.
It is more customary to model price charts with some kind of Brownian motion. Brownian motion is ultimately random and unpredictable. Here's an applet that shows various "price charts" produced by Brownian motion: http://www.stat.umn.edu/~charlie/Stoch/brown.html.
Domo Arigato, Mr. Roboto!
  • #14
  • Nov 2, 2007 9:16pm Nov 2, 2007 9:16pm
Quoting Ted1983
In fact Newton himself rubbished the idea of predicting the markets during the South Sea bubble.
Oh well, then that settles that.
  • #15
  • Nov 3, 2007 3:15am Nov 3, 2007 3:15am
Quoting SunTrader
Oh well, then that settles that.
Yes, newton was not familiar with correlation of two instruments, haha. Today, you can cancel volatility.
  • #16
  • Nov 3, 2007 4:52pm Nov 3, 2007 4:52pm
Quoting c0d3
Yes, newton was not familiar with correlation of two instruments, haha. Today, you can cancel volatility.
What? Not familiar with correlated instruments.
Pure hogwash!
  • #17
  • Nov 3, 2007 6:35pm Nov 3, 2007 6:35pm
Quoting Ted1983
What? Not familiar with correlated instruments.
Pure hogwash!

lol, way too serious!
  • #18
  • Nov 3, 2007 6:47pm Nov 3, 2007 6:47pm
Quoting c0d3
lol, way too serious!
  • #19
  • Last Post: Nov 7, 2007 9:49pm Nov 7, 2007 9:49pm
Quoting SunTrader
:
0 traders viewing now