If the markets followed something like a Riemann distribution (yes, I know there's no such thing, pedants, relent for a second) then mathematicians would say they were random, beginners and mathematically 'educated' traders would say they were random, but traders would notice they form (unexplained) spirals and they would trade them.
The truth is probably that markets are 'boundedly random', some type of Brownian motion with Levy flights, in multifractal time, or something like that, but if I could prove it I'd have a Fields medal and be too busy lecturing to trade.
Also, if the markets were truly random they would be a cinch to trade since everything would revolve around clever money management. In fact, lots of gurus claim that's exactly the case and maybe they're right.
I know you can make chart patterns that look like real charts using random data, but has anyone tried making chart patterns using non-random data like weather patterns? Choice theory data? With the right setup I'm sure they'd look similar, the only difference is we have an explanation for their cyclical elements. Actually we have an explanation for the cycles in finance too, but that doesn't make it forecast-able, necessarily. You can't see the important differences from a single chart. The problem is so big it has confounded economists like Cambell and Lo, mathematicians like Mandelbrot, engineers like Shannon (probably) and physicists, though I can't think of one. So either there is nothing to discover or it's very difficult to discover. Either way, only an arrogant fool would choose to lock horns with it which is why there's no point in debating it with them. Al Brooks thinks there is no randomness in the market at all, and for obvious reasons there are no gurus flogging their random walk theories (that I know about - anyone?) but I'm sure traders at the extreme ends of that spectrum can make some kind of go at it even if it's just as survivors of randomness.
The truth is probably that markets are 'boundedly random', some type of Brownian motion with Levy flights, in multifractal time, or something like that, but if I could prove it I'd have a Fields medal and be too busy lecturing to trade.
Also, if the markets were truly random they would be a cinch to trade since everything would revolve around clever money management. In fact, lots of gurus claim that's exactly the case and maybe they're right.
I know you can make chart patterns that look like real charts using random data, but has anyone tried making chart patterns using non-random data like weather patterns? Choice theory data? With the right setup I'm sure they'd look similar, the only difference is we have an explanation for their cyclical elements. Actually we have an explanation for the cycles in finance too, but that doesn't make it forecast-able, necessarily. You can't see the important differences from a single chart. The problem is so big it has confounded economists like Cambell and Lo, mathematicians like Mandelbrot, engineers like Shannon (probably) and physicists, though I can't think of one. So either there is nothing to discover or it's very difficult to discover. Either way, only an arrogant fool would choose to lock horns with it which is why there's no point in debating it with them. Al Brooks thinks there is no randomness in the market at all, and for obvious reasons there are no gurus flogging their random walk theories (that I know about - anyone?) but I'm sure traders at the extreme ends of that spectrum can make some kind of go at it even if it's just as survivors of randomness.
1