Sep

28

 What comes to be known as being unknowable may or may not be consistent with randomness, rather than, what is unknown. Let unknowns be distinct from the random lest the spirit of inquiry, scientific or even otherwise, be stifled.

Within the unknowables, there are many types of problems (for one example, those suffering data insufficiency) that are not worthy of achieving the classification of belonging to the random. By the way, why does the classification consistent with randomness carry that less worthy connotation?

It is the consistency with randomness that helps insurance companies buy risk, shop floor foremen decide that maintenance shutdown is yet not required, two surfaces are allowed to produce sufficient friction to make the value of work to be non-zero, etc. etc.

A generalized extension is that all cognitive systems including human traders and non traders are able to undertake risk and achieve when they recognize that their willingness to assume the ascertained level of risk is going to produce a draw down, a negative incursion, an unprofitable outcome etc., over a course of several such ventures consistent with randomness. Whenever there is a risk expectation inconsistent with randomness the system would stop and re-evaluate if that is rather an opportunity playing the negative system.

To be consistent with randomness is not useless but useful in specific decisions. It is more valuable than what we do not know or will not know or will not be able to know.

The negation of a negative expected outcome is the way to capture positive outcomes.

If a regression of the results of the trading activity of various participants on an information curve ranked by their seniority would produce an R square that would explain how much of that information seniority explains the trading outcome and thus help explain to those existing at the bottom of the curve that there was no randomness in the impact of information and trading.

Thus, consistent with random is a classification within the larger subset of the knowns rather than unknowns.

Therefore I would surmise that for trading as much as any other human endeavor the idea is to filter the potential impact of consistent with randomness fate and luck away, and focus on effort. Maybe this is what is implied by the saying "fortune favors the brave."

Adi Schnytzer responds:

I'm sorry, Sushil, but I don't buy it. Imagine that you knew every order that was going to be placed when the market opened today and also knew all the financial details of all the brokers and traders and market-makers. And that you knew all of the news of the world and inside info. And, further, that you had a serious computer at your disposal. I suspect that little in the way of randomality would remain. In other words, in a market, I would argue that the "error term" is basically missing variables, many of which won't ever be known to any single trader even if all are known to the aggregate of all analysts. Bottom line: get the info you can and learn really well how to analyze it.

Kim Zussman adds:

The big Wall Street firms, hedge funds, etc., have the most accurate and up-to-date research, and it's hard to imagine we can beat them at that game. But what makes it still a game is that the reaction can be quite hard to predict — even if you could know the future.

For example, Ben's 50bp cut — only now in hindsight do we get to weigh his put (about 50 S&P points). However it was also possible the market could have taken these cuts to mean the risk of unpreventable recession was higher than expected, and they sold. That the prior week was up also threw a false signal — that the market had already priced a big cut.

Perhaps panning for fear in its many disguises, even getting it right only 52% of the time, is the best place to prospect.


Comments

Name

Email

Website

Speak your mind

2 Comments so far

  1. Gary Rogan on September 28, 2007 8:57 pm

    A human-based system in a middle of a frenzy, positive or negative, is not easy to model. The extremely high loop gains involved and self-modifying coefficients don’t lend themselves to straightforward analysis. Yet as of now the trajectories are still likely to overshoot the target, so fear is still good to pan for.

  2. Robert O'Neil on October 12, 2007 8:30 pm

    Can you please tell me about the artwork at the top of this posting?

    Thank you.

Archives

Resources & Links

Search