For support and other queries please contact us at service@chaosmonitor.com

Frequently Asked Questions

  • What is ChaosMonitor?

    ChaosMonitor is a web application and the underlying quantitative model that analyzes nonlinear properties of financial time-series and issues signals when these properties reach extreme values. These extremes indicate moments when markets are in an unstable state and are predictable in the short term.

    << back
  • Can a market be predicted in principle?

    Traditional portfolio theory based on the ‘efficient market hypothesis’ believes that returns from market prices follow a random walk and that the statistical mean and variance represent the best available estimate of future behaviour. These views are still widespread among the investment industry and mainstream academia and have not changed much since the 1950s.

    In recent decades a significant progress has been made in the study and understanding of a mathematical structure called ‘complex dynamical system’. The weather, avalanches, earthquakes and financial markets are examples of such systems. Their overall behavior is a result of interaction and mutual feedback between many small parts. The mutual feedback and resulting complexity makes the overall behavior of such systems look completely random most of the time.

    Even though a dynamical system (such as the market) appears random — there are a number of important differences between it and a stochastic system. Market returns do not follow a standard distribution. Instead they obey power laws producing a significant number of extreme events (fat tails). Secondly, since the market is a deterministic system, there are moments when its short-term behavior can be predicted rather accurately. ‘Short-term’ here means ‘a few units of measurement ahead’ (several minutes ahead if we are using minute scale, a few months ahead if we are using monthly scale).

    Long-term behavior of the market is still impossible to predict because a complex dynamical system is ‘sensitive to initial conditions’ and a small amount of short-term noise can make its potential future trajectory differ widely.

    << back
  • What does ChaosMonitor measure?

    The model focuses on several so called scale-invariant properties of the market. They describe state of underlying system regardless of chosen time periods and scale.

    The first property is fractal dimension, also called statistical self-similarity. It indicates to what extend the current system behavior is similar to its previous behavior under a scaling transform. One way to estimate fractal dimension is through the Hurst exponent (H) calculation (there are several others).

    The second property is log-periodic scaling behavior of system tension buildup and relaxation. The concept of self-organized criticality (SOC) was introduced by Bak et al. in 1987 (see selected bibliography on the front page) using the example of a sandpile with grains of sand being added. It has been observed that the building up and release of small avalanches follows a predictable pattern and can serve as the basis for critical event forecasting. This is currently an extremely active research field with applications in earthquake prediction, prevention of epileptic seizures, and now in financial market risk analysis.

    Pasted Graphic Pasted Graphic 1
    From Algorithms for random fractals, by Dietmar Saupe, 1988S.Kris Kaufman, Parallax Financial Research, 2008.

    << back
  • Is it a 'black box' model?

    It is common to define a model as a ‘black box’ when it is not based on ‘first principles’ and could not be explained using rigorous scientific concepts. For example, a neural net trained on a specific data-set may acquire some predictive power but would be considered a ‘black box’.

    ChaosMonitor is most categorically not a black box model in this respect. It is based on cutting edge fundamental research in the field of nonlinear complex systems and self-organizing criticality and takes advantage of natural phenomena that has only recently been discovered. Certain aspects of the underlying algorithms remain proprietary, but the scientific approach behind it has been discussed in recent peer-reviewed publications.

    The model does contain selected variables that have been derived through experimentation and nonlinear regression, during which neural networks have been used to find an optimal fitting.

    << back
  • Is it technical analysis?

    Technical analysis is normally defined as an opposite of “fundamental analysis”, i.e. a study of market prices versus an analysis of economic/political trends and company financial reports.

    Interestingly, nobody calls Markowitz portfolio optimization an exercise in technical analysis, even though so called “Modern portfolio theory” relies on mean and (co)variance of past returns to forecast the future.

    ChaosMonitor analyses fundamental properties of a complex non-linear system, namely its propensity to develop extreme behaviour as a result of interaction between its individual parts ( grains in a sand pile, tectonic structures in case of earthquake forecasting and orders from market participants in our case). We believe it is an essentially fundamental approach but are not too keen on terminology debates.

    Traditional technical analysis, on the other hand, can be described as a naive application of chaos theory. Its practitioners have empirically established that the markets are not random but most of them lack effective mathematical apparatus. When rigorous nonlinear analysis methods become widespread - then the distinction between ‘technical’ and ‘fundamental’ will be even more blurred.

    << back
  • What are the magenta 'extension' points?
    Extension signals are moments when the market displays extremes of both self-similar behavior and expected log-periodic oscillations. They show that strongly trending behaviour of the market is becoming unsustainable and is expected to stop.

    Very often after an extension, the current trend is followed by a dramatic reversal. However, this reversal is not part of the model forecast. The prudent approach would be to expect the prices to establish a sideways trading pattern after the signal.

    << back
  • What are the blue 'compression' points?
    Compressions are the opposites of 'extension' points. They show that the market is extremely directionless. They mark periods of extreme negative feedback and lack of trend, and this lack of trend is becoming unsustainable. The market is expected to move sharply from this point, to develop self-reinforcing feedback and to form a trend.

    The model does not predict in which direction the prices will move. There are several ways to estimate the direction of the emerging trend:

    • use nonlinear dynamics analysis of the larger time-scale on the same instrument
    • apply macro-economic and structural analysis (to identify misbalances that need to be corrected)
    • wait for the trend to start, it is very likely to keep the direction it has acquired initially

    The last option is fundamentally sound as it recognizes that an unstable equilibrium (trendlessness) has been broken and a new system state is taking its place.

    << back
  • How to base an investment/trading/risk management strategy on ChaosMonitor signals?
    A complete investment strategy always includes element of risk budgeting / money management. ChaosMonitor does not provide risk budgeting tools at this point. We are working on separate investment strategy product that would incorporate both.

    We advise to look at ChaosMonitor as an advanced market risk indicator and to integrate it in your own investment decision framework. This model's unparalleled accuracy and scope is very likely to give you a powerful competitive edge.

    A few thoughts on the potential integration.

    Compression signals are valuable indicator of coming spike in market volatility. Risk- and portfolio managers may think of buying portfolio protection and scaling down their active market risk at this point. Trend-followers are provided with a very accurate indication of imminent trading channel break-out. ChaosMonitor will help them to reduce costs associated with false break-outs. Volatility traders may want to purchase a long straddle here.

    Extension signals provide reassuring and accurate guidance during times of 'fast markets' and can trigger exits from successful positions or indicate point of position entry when a larger-scale prevailing trend is identified. For example, during the bull stock market of the past decade ChaosMonitor bottom-extension signals provided for a very profitable 'buy on the dips' strategy.

    << back