14th February 2012: Chaos Theory

 

Chaos theory and the complexity sciences have come up in several recent discussions. In one exchange, I found Peter J. Carroll‘s writings to be ill-defined and unclear. In another exchange, different underlying epistemologies were discussed. Richard Metzger and Jason Louv were influenced by contemporary chaos magic: to do rituals and to create sigils in the tradition of Austin Osman Spare. I went in a different direction: studying chaos theory (James Gleick; Edward Lorenz; Benoit Mandelbrot; and others); then reading about how such models can affect your life (e.g. Steven Strogatz’s Sync or Albert-László Barabási’s Linked); and more recently, looking at dynamical systems, mathematics and simulation modelling in the context of financial markets. For me, contemporary chaos magic is but a shadow of the scientific vistas of chaos theory and the complexity sciences.

Trading Chaos

Williams, Bill & Justine Gregory-Williams.  Trading Chaos: Maximise Profits With Proven Technical Techniques (2nd ed.), John Wiley & Sons, New York, 2004.

The father-daughter authors summarise a personal methodology based primarily on: (1) the technical analysis of oscillations in market securities; and (2) the opportunities for day traders and swing traders to appropriate value from institutional funds through ‘countertrend’ signals which occur in commodities futures and currency/foreign exchange (forex) markets.  The first (1995) and second (2004) editions coincided with the IT and subprime bubbles which created day trading subcultures and market volatility, so it would be interesting to see how the authors have fared during the 2007-08 global financial crisis.

The book’s first half synthesises various ideas on formulating a trading plan and the psychology of market trading.  The ideas include a social constructionist view of money as a holder of value (John Searle); crowd psychology and rational herds in markets (Gustave Le Bon, Charles Mackay); the new paradigm of chaos theory in markets and how fractals and self-similarity create new trading perceptions about pricing and signals (Benoit Mandelbrot), and the popularity of Eastern belief systems amongst traders as models for skills acquisition and stress management (notably Western popularisations of Zen and Taoism).  Thus an awareness of broader intellectual trends can be useful to unpack the building blocks of a system and for comparative analysis with other theorists and models.

Ben Williams’ original contribution is to explain how his background as a psychologist informs his trading approach.  Chapter 7 outlines a generic model of skills acquisition — novice, advance beginner, competent, proficient and expert — that was explored in the book’s first edition, and can be integrated with Agile, CMMI and other frameworks for integrating operations and strategy.  Williams summarises exercises from autogenic training for stress control in the face of market volatility, symbolic interactionist approaches to align the trader’s individual psyche with the market, and cognitive psychology techniques such as cognitive chaining for surfacing deeply held beliefs which lead to self-sabotage and trying to trade out of a losing position without stop losses.  The cognitive psychology approach reminds me of physician John Lilly‘s mid-career work on meta-beliefs and it also parallels recent work in behavioural finance.  However, some descriptions — such as a section on Taoism, Zen and visualising the market as a river which follows the path of least resistance — seem to be closer to New Age beliefs about zero point fields which integrate consciousness and matter than the original metaphysical systems.  I agree these systems can be applied to training however they need far more grounding than detailed here.

From the earlier material on trading approaches, the book’s second half develops a trading system to anticipate the price movements in market securities through fractals and self-similarity which occur in volatility.  It’s always interesting to see how traders justify their approaches and the example trades given.  I’m closer to the adaptive markets, event arbitrage and behavioural finance schools of investing and remain to be convinced about the validity of technical analysis that the Williams propose, beyond the obvious role of pattern recognition.  Actually perceiving nonlinear dynamics and turbulence can be very different to the language and paradigmatic thought that makes chaos theory a popular explanation.

I did experience some perception changes after reading Trading Chaos: (1) charts might be interpreted in a different psychological frame using fractal, self-similarity and volatility metaphors; (2) viewing charts at different timescales (e.g. 1 hour, 1 day, 1 week) might develop the cognition skills to quickly scan signals in a real-time environment; and (3) the juxtaposition of lead and lag signals for trading decisions and triggers has potential, particularly if combined with game theoretic modelling of the market and volatility effects from institutional investors, monetary policy and rational herds.  It remains to be seen if these perceptions are sustainable and verifiable in trading conditions, and not just subjective reactions based on past research about chaos theory models.

That said, the trading system may also have several criticisms and weaknesses. Finding signals in oscillations and nonlinear dynamics may be difficult in a volatile market.  Analysts can be subjective particularly if de-leveraging and other actions by institutional investors are not factored in.  Swing traders may be exposed to market sensitivities (aka the Greeks): Gamma (the rate of change in the underlying security’s price), Vega (sensitivity to volatility), Theta (time-decay) and Rho (time-decay of interest rates).  Finally, modelling turbulence and uncertainty in a grey or white box system remains a major challenge for financial engineers in new market environments.

Threaded throughout Trading Chaos are the mix of useful insights and shibboleths in day trading subcultures.  CNBC, investment experts, and the plethora of courses and newsletters thrive on investor insecurity yet create noise (pp. 34, 42, 56).  Trading decisions, trading volume, and speed and type of momentum may be lead indicators of price volatility (p. 126).  Broad market knowledge purports to trump expert/specialist understanding (p. 135).  Market facts must be distinguished from opinions and beliefs (pp. 8-11).  Trader personalities shape risk tolerance, time horizon, the asset allocation process and types of controls (pp. 92, 155), a factor relevant to human resources consultants and the ‘transition in’ process for trading desks in investment banks.  Analysis risk involves emotions and perceptions of a signal (pp. 52-53).  The interest in Fibonacci numbers and Golden ratios are partly because they are iterative, geometric structures applicable to price movement forecasting (pp. 22-23).  Grey and white box systems with transparent, programmable rules are preferable to expensive, high-end black box systems which use artificial intelligence and neural net algorithms (pp. 53, 56).  A useful bibliography highlights the Santa Fe Institute‘s influence on chaos theory applications in finance and macroeconomics.  It suggests this area needs far more research to verify the claims and provisional findings in this book, to separate the gold from the dross.

Perhaps the most pivotal insight of Trading Chaos is buried in the text.  “We all trade our belief systems.  When some of you think about this, it produces a crisis,” the authors assert.
  Now that could be the basis for a ‘contrarian’ trading system — probably the one that hedge funds with a short/event arbitrage approach use to scalp day traders in currency/forex and commodities futures markets.