Feb

22

It is interesting to consider whether the levels of volume of the major markets are normally distributed. Related to this is the question of whether a spell of low volumes is predictive of a undue likelihood of a high volume in the future. The question is related to vix predictivity also, but volume may be useful.

One hypothesizes that volume must be shot upwards from time to time for the public to have an adequate chance to lose more than their fair share, so THEY can take their emoluments to live well and cover their overhead.

One notes that volume has been inordinately in the 1 to 2 million area with only 4 days of 2 million volume so far this year for SP futures, and they have not been clustered. What is the hazard curve for a 2 million volume day? Does the lack of high volume days this year relate to all the people who were gipped by the MF bankruptcy or does it merely reflect the normal tendency of scholarly chairs to preclude big declines in an election year so that the stuck out nails in the public will remain small?

Are such queries relating to volume useful?


Comments

Name

Email

Website

Speak your mind

5 Comments so far

  1. douglas roberts dimick on February 23, 2012 2:06 am

      Volume is as Volume Does

    During the first three years of working on whether one could defend the Theory of Quantitative Relativity (or QR), the issue of volume initially commanded a focus parallel to defining interval correlations of price action. Similar to defining market psychology for purposes of quantification, an interesting consilience appeared between the Vix and mapping (high-low polarity) of price action: paired sets of data input and indicator outputs appeared statistically independent.

    How could that result be possible given known (indicator) price-volume symmetries between the two variables? They are not mutually independent – aren’t they?

    For instance, consider the Volatility Standard Deviation Indicator…

    Market Synopsis
    The Volatility Std Dev indicator calculates and plots the volatility of a market using prices from the number of bars specified by the input Length. The value returned is 1 annual standard deviation in price expressed as a percent of the market price. Options traders will recognize the Volatility StdDev indicator is akin to statistical volatility.

    While you can apply this indicator to any time frame, it is designed for use with daily bars because the values are annualized based on a daily timeframe. Using another time frame causes the values not to be accurate because the annualization is not calculated properly.

    Plot Information… When applied to a chart, this indicator displays one plot in a separate subgraph from the price data. When applied to a grid, this indicator displays one value in a column of cells.

    (Excerpt from Definitions at Tradestation.com [Help] Definitions)

    Now consider the All OptsROC Indicator…

    Description
    Volume - All OptsROC is an options-related indicator that plots the rate of change in total option volume that is calculated as the sum of Put Volume + Call Volume over a user-specified number of bars..

    When you note the relative position of the 2 averages: the short moving average above the long moving average would yield a positive Oscillator value; the short moving average below the long moving average would yield a negative Oscillator value.
    Calculating the difference between the two averages and plotting this as an oscillator makes extreme positive and negative values stand out as possible buying or selling opportunities.

    Plot Information… When applied to a chart, this indicator displays two plots in a separate subgraph from the price data.

    (Again, excerpt from Definitions at Tradestation.com [Help] Definitions)

    In that VoltSDev plots velocity of price as a percent, OptsROC quantifies acceleration volume. It appears that the two indicators are absolutely independent.

    Yet one could claim that they are tautologically bound sets. Consider how the conditional probability of one indicator here may be the same as the unconditional (or marginal) probability of the other indicator. Are there symmetrical roles operating here (within the systematics of electronic exchange), whereby the issue of independence is not applicable?

    While studying Victor’s work these past three years, I was taken by how often he returns us to that baseline question of whether markets exhibit nonrandom sequencing or may nonrandom patterns be discerned from quantitative impression or randomness merely abounds as an ecological norm. Why would he do so with such frequency?

    As in racquet sports, is he merely forcing himself back to center baseline after each shot at answering the mother of all questions concerning the financial markets universe?

    Also consistent with this issue of independence relative to probability of price/volume distribution is Victor’s observation about lack of clustering in the SP Futures. He is referring here to cumulative distribution functions and probability densities that are or are not independent?

    The Chair’s query about volume may prove to be a trick question. If you are a fund manager with $1b and an aggressive strategy, then, sure, such queries relating to volume are useful as is the likelihood that the flexions are preparing to skin ma and pa for a little more than usual during our 2012 patriotic obligator with its en mass crescendo this November.

    Actually, one-trick pony comes to mind…

    http://en.wiktionary.org/wiki/one-trick_pony

    For volume “appears linear” by definition, providing another (high-low) polarity that, by stochastic definition and as a market indicator of probability distribution, can alter what may be considered a rules-based construct for determining if, where, and when there is nonrandom sequencing occurring within the market. Is that true?

    The are better and bigger minds than moi here at daspec to discern the truth of it.

    I shall conclude only by noting that when I beta tested a QR constructed indicator-set of functions mapping the Vix relative to price action, efficiency of varying copulas failed in terms of any multivariate normal distribution. To that end, volume is like a box of chocolates; you do not know what you’re gonna get… and that is the problem.

    dr

  2. Edward Lam on February 24, 2012 8:35 am

    On flexionism (unrelated I know, but a topic of interest to speclisters I know):
    http://reason.com/archives/2012/02/09/warren-buffett-baptist-and-bootlegger/singlepage

  3. douglas roberts dimick on February 25, 2012 2:46 am

      EL,

    Actually, although the Chair’s query here primarily references a different subset of players, your cited article as to how Buffy moves political cache on the financial gameboard is more analogous to the volume issue than perhaps most of us would first consider.

    That author’s conclusion that ole WB is a political as well as financial genius points directly to my questioning observation to V on volume…

    “For volume “appears linear” by definition, providing another (high-low) polarity that, by stochastic definition and as a market indicator of probability distribution, can alter what may be considered a rules-based construct for determining if, where, and when there is nonrandom sequencing occurring within the market. Is that true?”

    In fact, the politicians both during and after public office determine the rules of the game. It is a point that I find many a financial and market maven choke on when admitting or requisitioning in discussing market dynamics; they prefer to continue the allusion (or delusion) that external and internal dynamics of economics and markets are the key factors for governing price action to include volume(s).

    Buff-meister consistently and repeatedly proves wrong all those slaves of the mistress of the markets — I sense many of them despise him for it.

    Regardless, your article is on point and relevant to the conversation here, perhaps central. Until more of that curtain is raised (e.g., industry and actor transparency as well as small-investor franchising), the wizards (being what daspec’rs refer to as flexions)shall continue to run circles around the linear egocentricity of the rest of us.

    dr

  4. dvdw on February 27, 2012 11:47 am

    Sum values of volume are really just numbers without characterization of the types and categories. Was often defeated by counting volumes under the terms of free float analysis, without proper fuzzy characterization of the observed volume….you can not discern intent…intent is afterall the proper characterization of a human system in action.

  5. JACOB on March 5, 2012 4:36 pm

    RE:..Alex Epstein;.JUNTO SOCIETY..3-1-2012.."INDUSTRIAL PHILOSOPHY" …FOR THE FUTURE……..

    "Quite a Fascinating Youngster"………
    …..Seemingly The Same 90 Seconds Of Content For 95 Minutes Or So..!!……………"..Really Drive His Point Home !! "..?
    ………………………………………………………..
    .."That Greedy Capitalism is…Really "MUCK-LUCKING" The Waters(Environment)…"………………."..& That Elusive 'PRIVATE PROPERTY'..Keeps.Rearing Its Head"…..ETC………………….
    ………………………………………………………..
    Perhaps: " A CLEANER INDUSTRIAL PHILOSOPHY  FOR THE FUTURE …??."

Archives

Resources & Links

Search