|
|
|
|
 |
Daily Speculations The Web Site of Victor Niederhoffer & Laurel Kenner Dedicated to the scientific method, free markets, deflating ballyhoo, creating value, and laughter; a forum for us to use our meager abilities to make the world of specinvestments a better place. |
Write to us at:
(address is not clickable)
4/25/2004
Weather Prediction and
Markets, by James Sogi:
The techniques of weather and surf prediction may be helpful to speculators seeking to predict the market. The surf forecaster Pat Caldwell is well known among surfers for his accurate surf predictions and works with the National Weather Service. He analyses satellite data, the observations of ships at sea, computer models, and applies them to his knowledge of local surf conditions. He predicts the time and size of the surf.
http://www.prh.noaa.gov/hnl/Products/SRF.php
He analyses his accuracy rate also.
The theory is that storms with high sustained winds in one
direction aimed at my surf spot creates waves there a week
later. A critical factor in getting good surf is the
periodicity of the wave swell from peak to peak. 7-9 second
period is considered choppy. Buoys at sea measure the wave
period and size, so we know in advance what size swell is
coming. 20 second period results in big long fast smooth
waves. He looks for weather systems on far sides of the
globe and determines the
intensity, the direction, duration and path through the
island chains that may block the wave swell. Depending on
the swell and local wind conditions, the surfer will plan
which side of the island to surf, and the size board to
ride, the time of day to go out. Waimea bay calls for a 9
foot "gun", small beach break calls for small wide "fish"
board. Time of day is important as the tide creates certain
predictable conditions.
The market sends out similar signals. When the Fed starts a major windstorm on the far side of the globe, it sends out strong and turbulent winds. Like the fed jawboning about interest rate increases.
The press generates a lot of wind. Depending on the storm, its intensity, its duration, its direction, it affects the periodicity of the market in the amplitude and period of the sine wave. The Desert Storm kicked off a heck of a swell during March '03 and if one looks at the charts, the waves were sine-like from then to now. If one were to be able to determine in advance the period of the market waves by counting the period of the swings it would help in timing ones trades.
Ehler's new book has a sine wave predictor and takes only 1 ŸŸ swings to determine periodicity and can be used to avoid getting caught on a trend break out after range trading. There might be a way to see what kind of waves are generated from certain conditions. There are some pretty standard rules of thumb, equal a b c moves for example. On Fed day, one can see in the past that there may have been splashes, x point swings x minutes apart. That is usually good to know to stay away, or where to take a stab. Or on employment day, what are the expected amplitude of the swings and their periodicity. Or on average, what are the daily waves periods.
I think that is what the Chair's recent post computed the daily range, or wave size, but how about the periodicity. Time of day, like the tide, has some predictable characteristics. Perhaps certain vix levels can be correlated with periodicity or amplitude. If we could count the time between the peaks or valleys of the waves, we could do some wave forecasting. I'm not talking Elliot Wave here or Fib, but counting the timing.
Here's my count of $spx wave top periods since last
March.
"3/21/2003"
4/7/2003 17
4/23/2003 16
5/6/2003 13
5/16/2003 10
6/6/2003 21
6/17/2003 11
7/14/2003 27
7/31/2003 17
8/22/2003 22
9/8/2003 16
9/18/2003 10
10/15/2003 27
11/3/2003 18
12/3/2003 30
1/8/2004 35
2/17/2004 30
3/5/2004 17
4/5/2004 30
4/22/2004 17
Average wave is 20 days. Ave dev is 6.349.