NIKKEI Opens Higher

9/17/2001

[Soren]

NIKKEI gapped up about 120. Now at 9762 (+258)

[RightlySo]

Nikkei opens higher...which is bad news, actually. I'm a trader, and have been for 10 years now. I run a small commodity trading pool (I manage other people's money in futures and trade them using a variety of computer based models).

I say bad news because the clear majority of the time, such gaps are filled within days (which means is will have to come back down to fill that "opening gap" from the previous day's close). Almost always within weeks. The markets never travel in a straight line-- for every buy there is a sell, which creates "noise", to use data analyst language.

My take on the Japan market (and the Nikkei is not my forte) is that they are entering a severe recession. Yesterday, BOJ (Bank of Japan) intervened and suppressed a rise in the yen-- they need an export driven recovery and a cheaper yen makes Japanese goods cheaper for us to buy.

The Nikkei will continue to fall beyond today in the coming weeks. This "rally" is likely short covering in my humble opinion.

... but let's hope for better days ahead! :)

[Soren]

What's your take on Japanese banks having to switch to market value rather than book value for stocks for purposes of calculating their reserves beginning 10/1?

[Soren]

such gaps are filled within days (which means is will have to come back down to fill that "opening gap" from the previous day's close).

Could you elaborate on this. What is the driving force that brings the market back down?

[RightlySo]

Well, again, this area is not my forte, but my guess would be that Book Value can be "fudged" a lot more easily than market value. Book Value includes such accounting tricks as the worth of equipment and real estate-- projections for such value can be tinkered with, and are, often.

[RightlySo]

Being a data analyst and researcher who creates computer models of the markets to tell me when to buy and sell, I ask "what is the driving force" occasionally, but more often than not, I ask, "let's test the evidence of this 'as-of-yet-nondescript' driving force", and then assemble a logical hypothesis to describe its behavior. Ensuingly, I'll test that hypothesis even more. In other words, I care less for the driving force than for the possibility of understanding the *consequences* of that force.

Statistically speakly, gaps are filled. In my research efforts, we've found that part of this phenomenon is due to the "Random Walk" effect; a simple mathematical formula that illustrates the notion of random noise in the markets' behavior.

Another potential cause as to why gaps are filled is terribly difficult to prove. It involves human psychology which most good statisticians will desribe as bunk. But from watching the markets for years now, and observing how they trade, I have an untested theory that market participants simply did not have the opportunity to trade at a specified price once a gap has occured. There seems to be a compulsion for the data to close that gap, as if to preach to the participants, "we didn't get a chance to trade through the price gap, so let's retrace the gap with prices and then be off on our way again." Evidence of this in price behavior can be seen when gaps are filled-- the price where the gap is subsequently filled is often the short term peak or trough.

That's my take on it, and there are a million ways to approach the stock market-- a true enigma full of an indescribable number of forces behaving nonlinearly together.

[Soren]

Thanks for the explanation. I'm sure your models are proprietary, but in general, what types of mathematical techniques do you rely on?

[RightlySo]

I rely mostly on AI -- artificial intelligence. That includes, but not limited to, neural networks (nonlinear regression), GA (genetic algorithms for robust optimization), fuzzy logic, and to a lesser extent, SOM (self organizing maps using neural networks), and decision trees.

And strangely enough, sometimes the most simple idea, once tested, works as a model unto itself! :)

[Soren]

That sounds fascinating. Gonna call it a night. Off to read The Creature From Jeckyl Island. Have a good one.