Of course-- but considering the audience, I'm not so sure many will be able to distinguish differences, even when explained. I do hope people will give you a chance to speak without trying to thrash the thread; you are one of the few people who have added profound insights to my own understanding of markets as a system, and I'm exceedingly glad I was able to shut up (just enough) with my counter arguments to actually listen to some of the wisdom you wanted to share.
Do you see patterns in Random Walks?
- Thread starter atlTrader666
- Start date
Of course-- but considering the audience, I'm not so sure many will be able to distinguish differences, even when explained. I do hope people will give you a chance to speak without trying to thrash the thread; you are one of the few people who have added profound insights to my own understanding of markets as a system, and I'm exceedingly glad I was able to shut up (just enough) with my counter arguments to actually listen to some of the wisdom you wanted to share.
Markets, like nature, are opportunistic. If you plant at the right time, and harvest at the right time, you get good wine, the final outcome is affected by "little things" like sunshine, rain, etc, etc, etc. But one thing you can be certain: If you plant on harvesting season and harvest on planting season, your results will be crap. Like walking into a sorority when all the friends of your girlfriend have their period: A disaster.
You are too kind, as usual.
Here is the first shot at the definition.
RANDOMNESS is a universal property of matter that manifests it self through the sequence of changes in its observed characteristics that could not be described through any given set of definitions of limited complexity.
what about the realizations of N(0,1)? It can be perfectly described with a finite (and small) set of properties, but not predicted.
Yes, but not all of its outcomes. That is the point. There are some systems that could be well defined through their initial conditions, however, if they present infinite complexity of outcomes there is no set of descriptors with finite complexity that could adequately present them. I hope I was clear, if not, I apologize and will try again.
But I think that's exactly what I was trying to point out in your framing as to lead to a nihilistic state. If we consider a ~N(0,1) generator that takes an initial seed, then we can describe all subsequent outcomes with a finite starting state. Fine - you say that's not true randomness. Okay, but then that makes your definition of randomness a bit of a No-Real-Scotsman, no?
That is, any generator of realizations that can be described by a finite state vector is not randomness - only random generates that cannot be described by such a vector is; - so your definition requires your conclusion which requires your definition.
Finally, such a descriptor of random is a bit unless in the sense that it doesn't apply to anything that we actually do.
That is, any generator of realizations that can be described by a finite state vector is not randomness - only random generates that cannot be described by such a vector is; - so your definition requires your conclusion which requires your definition.
Finally, such a descriptor of random is a bit unless in the sense that it doesn't apply to anything that we actually do.
I see your point. You probably have in mind a time sequence where I was referring to a property that is determined by the infinite number of outcomes of infinite number of simultaneous (parallel) processes.
I probably need to work a bit more on the definition as it clearly allows for veriaty of interpretations. Damn, I though I had it!
In terms of utilization of such definition I had in mind an illustration of why and when its applicable in a very practical and close to this forum application.
But, 10 years ago when a fund went down to the floor on the NYSE they did not dump a few million shares at once. They would hire a so called penny broker... who would go over to the post and ask the specialist where he could move 50,000 or 100000 shares. The specialist would tell the broker where he could make the trade.. Once done the specialist would ask the broker if he had any more behind it.
with many stocks the crowd and specialist would frequently follow their recent patterns once learning the info.
Generally off the floor we would wounder if the large print was a clean up print.... or something which might indicate a big order.
One long standing pattern was back to back larger double prints 5 to 15 trades later. You would see two 5000 or 10,000 share prints hitting the tape back to back. This was an indication the brokers in the crowd or the specialist and a broker were on the same side of a trade. After that given proper conditions and time you could expect the trend to continue with many specialists. Sometimes with teach stocks it was free money for the afternoon.
It was odd because there were people arguing that prices were random, yet a dozen of people in my brokerage made six figures every year off those patterns.
Other patterns I loved had to do with VWAP and larger trades. Patterns so obvious we could train traders to make a living for about 3 years and take half their profits.
--
Patterns still exist... they are just not as obvious.