Daily Speculations The Web Site of Victor Niederhoffer and Laurel Kenner


The Chairman
Victor Niederhoffer


About Victor Niederhoffer

Write to us at:(address is not clickable)

The Four States of Matter, by the Chair and Doc Castaldo

We took a walk to discuss volatility and the artful statistical properties of the elements that make up such a thing as the market. We thought it might be good to extend the ideas about the movements, components and temperature of the market volatility situation by working with models, constituent elements and stocks.

There are four states of matter: solid, liquid, gas, and plasma. These four states can be characterized according to the kinetic particle model based on the attraction between, and the speed of, the particles. Change from one state to another requires considerable energy, with much potential energy relative to kinetic energy and much greater movement of the particles as we move from solid to liquid to gas. This simple model enables us to predict and explain many of the properties of matter, ranging from where and when something is going to melt or boil to how much heat and movement will be generated or lost as we move from state to state.

It would seem that we need a similar understanding and benchmark for the market's states, and as a first step towards this we have calculated the movement and degree of closeness of the components of the market as of the end of the month, for all months from year-end 2003 to August 31, 2005. To measure the degree of proximity of the individual members we used the inter-quartile range, and the divergence between the 90th and 10th percentile, and to measure the speed of movement of the market, we used the monthly change. All data are for the S&P 500 companies as they exist today. Adjustment for adds/deletes would have added 1000-fold to the time for the study, and a man cannot have breakfast all day.

        25th%ile        75th%ile          IQR  S&P Return
        --------------  ----------------  ----  ----
Jan04    JPM        5.9%  ING RAND   -2.0   7.9   1.7
Feb04    APPLE      6.1   COOPERTYRE -1.6   7.7   1.2
Mar04    OFFICEMAX  2.8   AUTOZONE   -4.2   7.0  -1.6
Apr04    ACE LTD    2.8   PERKELMER  -7.0   9.8  -1.7
May04    JONESAPPL  5.1   FNMA       -1.5   6.6   1.2
Jun04    3M         6.5   ST JUDE    -0.8   7.3   1.8
Jul04    NORFLKSO   0.6   MERRLYNCH  -7.9   8.5  -3.4
Aug04    LINCNATL   3.7   XPRESSCRPT -3.7   7.4   0.2
Sep04    WYRHAUSR   6.4   SOUTHERN   -1.2   7.6   0.9
Oct04    CSCO       6.1   AFFIL CO   -2.0   8.1   1.4
Nov04    FORTUNEB   7.8   DOMINION    1.8   6.0   3.9
Dec04    XPRESSCRP  6.2   WATERS      0.3   5.9   3.2
Jan05    PPG        0.9   YAHOO      -6.6   7.5  -2.5
Feb05    KROGER     5.2   JPMORGAN   -2.1   7.3   1.9
Mar05    MERCK      2.1   BLK&DECKR  -4.7   6.8  -1.9
Apr05    FPL        1.7   HOMEDEPOT  -7.5   9.2  -2.0
May05    EMC        7.2   MELLON      0.2   7.0   3.0
Jun05    PG&E       5.0   KELLOGG    -2.3   7.3  -0.0
Jul05    MAYTAG     7.8   SOUTHERN    0.9   6.9   3.6
Aug05    HMA        2.2   I.F.F.     -4.8   7.0  -1.1

Months with a large rise have a small IRQ, (such as Nov'04 and Dec'04), and months with a large fall have a large IRQ, (such as Apr04, Jul04). There seems to be a negative contemporaneous relationship between IRQ and S&P return, (-0.56 correlation). The down months are perhaps the ones that correspond to the gaseous state, the up month to the solid. Also there seems to be a positive relation between IRQ and next months return, (+0.30 correlation), but there may be a more direct way to get at that. All in all no breakthrough findings here, but that is often how science proceeds - in small steps.

Jim Sogi adds

The S&P's recently, regularly occupy states within 15 point ranges. There is a barrier on either side of that range for about a week, when it breaks through and changes to a different state where it occupies a different 15 point state for a week, then breaks up or down with the addition of energy or increase of entropy. The time periods are regular. The ranges are regular.

It is like a hotel in New York. You have your lower story rooms, 1200-1215, 8/24-31. The lower price rooms tend to be more noisy. You have your deluxe rooms floor 1215-1230, 8/16-24 and 8/31-9/2. You have the super deluxe rooms on the upper floors 1230-1245 from end of July to the beginning of August and now through - ?. The prices go up and down the hotel elevator like maids going up and down the same floors until the rooms are clean and they move on to the next level. One bar breaks through to a new range in a rush after three or five bounces to start a new cycle.

Thermal states provide parallels and insights.

Water crystals have characteristic snowflake patterns, three points on each side. The endless, intricate, delicate and varied patterns seemingly carefully crafted by a master craftsman never cease to astonish in their complexity and beauty and bilateral symmetry. Photos of snowflakes. In a short time, the bump sticks out even farther than it did before, and so it grows even faster. We call this a branching instability -- small bumps develop into large branches, and bumps on the branches become side branches. Complexity is born. This instability is a major player in producing the complex shapes of snow crystals. Markets form branches as traders who are too impatient to wait pile on the closest branch.

When we put cool tap water into a beaker and place this on the stove and turn on the heat we may see ripples in the water as the water heats. These ripples are called Schlieren patterns and arise from light being diffracted off of the regions of water with different temperatures and hence different densities. This is exactly the same optical effect that gives rise to mirages."

Many of the patterns and movements in the markets are mirages. Don't be fooled by the appearance of water in the desert, or the little mini ripples.

More writings by Victor Niederhoffer