Close this search box.
Close this search box.

Modeling Stock Markets

Why is it not easier to use an algorithim to model the stock market?

Modeling Stock Markets The one critical difference between markets and physical systems is that markets are not mechanistic in the way that physical systems are.

Human intentionality is the main driver! As a result, there is not a robotic or consistent pattern in the markets that one can constantly observe and study. Essentially, all of the patterns that occur in the markets, finance and economics are constantly changing and never stay the same pattern.

Indeed, every type of market phenomena. (For example: momentum, volatility clustering, mean-reversion) is driven by human behavior.

Furthermore, while some rough analogies can be made between physical systems and markets.

Moreover, the mechanistic approach to economics (driven historically in large part by ‘physics envy’) has done more harm than good to the practical applicability of economic and financial theories. In fact, psychology (see e.g., behavioral finance) and game theory are arguably more useful to finance than physics.

But should an investor engineer start with the news?

Not only they are unpredictable, investor reaction is unpredictable, too, as there are contrarians (sometimes many of them) besides plain ‘rational agents’. Remember ‘buy on gossips, sell on news’?

Then there is a hypothesis of order-driven inelastic markets that emphasizes importance of market impact of trading (regardless of news). In conclusion, the result is that finance is full of analysis of past market patterns. In addition, various (sometimes ingenious) models attempting to describe them. Lastly, but there is no unified theory that is truly useful for both academic and corporate financial engineering practitioners.

Michael Littman on Deep Learning VS Reinforcement Learning, Machine Learning & Markov
F-16 Fighter Pilot & Squadron Leader Captain Jeff Orr – Rebellion Research

Trading and Investing – Rebellion Research

Modeling Stock Markets : Modeling Stock Markets