Just saying Hi!

[mlytle0]

 I have been creating my own math formulas for 'some ' fractal events.   They have applications in markets, average world temperatures, and a few other things.  I feel the people here can give me some ideas..

[youhn]

Hi and welcome to the forums.

I'm curious about what you've created. Please show (some details of) those formulas. Any specific example for (special, funny, new, original) applications?

[mlytle0]

Well, I actually published a 60+ page book on this with all of the formulas.  I spent a lot of time making it clear and readable, and would upload to this site if there is interest..

Basically, markets are considered to be similar to 'brown noise', referencing the info in Manfred Schroeder's book on 'Fractals, Chaos, Power Laws'.  That is under the assumption that these type of fractals are 'innovation processes'.  I never entirely believed that, and found from careful statistical analysis, that the (not particularly reliable) 'Elliot Wave' theory, could be 'amended' by means of equations that contain both Fibonacci and logarithmic elements to something more useful.   What you end up with in a 'Time Transform', with 'Nodes' that represent higher probability periods for trend change.  In that transformed space, new trendlines appear, some always defined, and some empirically defined.

I had a blog for a while called 'marketmathematics.blogspot.com', where I featured my 'special charts' along with more conventional ones, and was a guest on Michael Yorba's site for a while. & did a couple of radio interviews for him, on this method.  Bailed out of it, when I decided the markets were becoming far too manipulated for any method to be totally trusted.  I have found the equations seem to work for other things, though, but I haven't investigated this thoroughly...

The posts on Yorba's site are here: http://yorbatv.ning.com/profile/MarkLytle

[mlytle0]

I have attached a fairly recent example here of the Standard and Poors index, $SPX.  Market turned at -1.5  and -1.0 node, which references the 2002 and 2009 market bottoms, respectively.  Note the declines stopped at the dashed brown trendline.  The market is much higher now, since I haven't updated this spreadsheet for a while, but you can see what this math does..

[youhn]

I probably don't have enough knowledge about economics and markets to read your book. But fractals are linked to markets since the word fractal is mentioned. I know it can describe aspects of the markets, mainly the roughness of price indexes. The main predictions are "outliers" or black swans. But as far as I know, the fractal math on markets can only predict that they will be there. Unfortunately not WHEN they will be there. This view of mine could be totally out of date, since pricing and markets cannot really grab my attention or interest for longer periods of time. But I somehow hope or expect progress the last half a century. Are the predictions getting any better?

And besides the markets, what are your expectations for the method to work?

By the way, I wish I could understand that graph. Price on the vertical axis, but what is the horizontal axis?

[mlytle0]

The horizontal axis is time, but instead of 'calendar dates', the fractal math generates an index.   The horizontal axis can be thought of as log(time) where the log function is not based on 'e' or base 10, but a seed related to measured characteristics of the previous market action.   It has worked spectacularly well, as long as it's paired with other indicators.   Right now, the financial industry is distorted by 'hot money' handed out to the top 5 'TooBigToFail' banks, so I would question whether any method, even mine, is real safe right now.  Technical indicators are based on the assumption of hundreds of thousands of 'random players' and that is what the current markets most assuredly are not, at the moment.

The current accepted fractal math of markets has had limited use because it assumes the market patterns are just 'brown noise', which it does resemble.  That is why they have no time predictive capability.  The assumption taken there, that markets are just random noise, is wrong.

Here I have attached a 'standard' chart showing the calendar year along the bottom.  You can compare this one to mine and see how the time axis has been 'compressed'. Should make more intuitive sense when seen this way.  I have found that not all people have the same visual knack for these graphs, but by comparing the two, you might see what has changed from one to the other.

As a reference, the first big peak on the chart here, which you can see was around the year 2000, is the same one in the earlier chart I provided that sits at just past -2.0, on the horizontal axis.