Re: From the mouths of retired mathematicians

["Henry C.K. Liu" <hliu@xxxxxxxxxxxxxx>]

But the latest models do have sentiments indexes built into them.  I know of no model on the Street that tries to simulate total macro conditons.  All only try to simulate and "predict" micro convergence or divergence, often in short time span of minutes to squeeze value from market inefficiencies.
They merely try to profit from noise and friction, particularly communication friction and institutional reaction time.  This type of tansaction has one overridig characteristic of danger: because the spread  of advantage is so small and the window of opportuinity so narrow that to profit, trades
have to involve huge volume and fast movement driven by herd instincts.  The speed of electronic trading is such that those who do ot use models simply cannot react fast enough to stay in the game.  Trading curbs and circuit breakers did moederate this danger but not eliminate it.  Structured finance
deals are mostly marked to market automatically and electronically.  A counterparties account can be seen to dwindle on the trading screen second by second.  A hugh fortune can literally be drained empty before the traders eye on the screen if the correct button is not pressed in time.  Two trillion
share trading days are now common place, and that's many times the number of shares outstanding, meaning each share is traded many time every hour.  Technicals have overwhelmed fundamentals a long time ago in the real world.  In the past two years, some $6 trillion of market capitalization have been
wiped out to generate a few billion in profit for the astute few.  That is 60% of the GDP!  No economy has survive such volatility.

Henry C.K. Liu

Ken Gordon wrote:

> While not, perhaps, earth shattering or new, I thought the PKT list might enjoy this from today's Toronto Globe and Mail Report on Business. The whole story is at: <http://www.globeandmail.com/servlet/ArticleNews/PEstory/TGAM/20021012/STMYMO/Business/business/businessNetWorthHeadline_temp/1/1/9/>
>
>      Now that he's left the staff of the University of Waterloo, retired math professor Peter Ponzo has all the time in the world for researching market-beating formulas and strategies. The only problem is that he, er, doesn't have any time for them.
>
>      Ultimately, the 68-year-old says, "I don't think that mathematical stuff does much good.
>
>      "While you can play with past statistics, mathematical formulas don't give you much insight into what will happen in the future because they don't reflect human nature and investor sentiment, and that's what drives so much of the market."