​

 

Primum Tempus intends to provide innovative financial market tools. In the contemporary environment of Artificial Intelligence, innovation generated by human beings seems to be difficult. Yet, AI is not to this day able to deal with counterfactuals, which are the basis on which forecasts can be made. In Artificial Intelligence, intelligence can be understood as a gathering of a large sum of information more than as an  ability to create innovative content. Addtionally, be it for humans or machines, dealing with counterfactuals on the markets remains difficult since the elements that are used to set the hypothesis are not even well defined concepts.  

​

Despite or because of the sophisticated mathematical engineering needed by an industry, the definition of main concepts of asset management has been based upon a need more than a reality. They are artificial reifications designed to create the conditions needed to use the mathematics of uncertainty. However, no one can compare a quality with a quantity, and doing so can only lead to unjustified assertions, as any reader of Kant would know. In this aspect, how can it be sound to compare a return and a risk, be it through its measure ? How this already strange comparizon can become a factual bijection, as it is in the CAPM- on which, in essence, most concepts of financial engineering is based upon ? How is measured a quality ? What is the relationship between the unit of measure and the object that is measured ? What means the concept of efficiency ? Is there really an implication between informational efficiency and price reflecting the value of an asset ? 

​

All these are legitimate questions. Yet, everybody uses CAPM, be it in its plain vanilla version or in all its subtle variations and/or in all its methodological consequences. Sharpe ratio, asset allocation matrixes, allocation vectors, variance-covariance factors, risk measurement etc. It goes all the way to the collection of data that cannot be compared over time, but which are still considered as homogeneous datasets. This is how the industry says that this fund/manager is safer than this one, or that this one is more "aggressive" than that one... These are more opinions than measurements. 

 

The elementary concepts used in financial engineering are not independent from each other. Price, value and information form a complex system. In an article published in 2022, titled "Unaccounted forms of complexity : A path away from the efficient market hypothesis paradigm", we have provided the first theoretical demonstration that complex system computational irreducibility de-substantiates the notions, concepts and consequences of the Efficient Market Hypothesis. Price is not a neutral measure of value, and price can behave in an automous way This simple fact is the radical proof that most if not all concepts defined by the school of Chicago are, to say the least, vague* or to be more correct, unproven, unjustifed reifications or concepts. However, these tools and methods, taught in business schools - which rankings are based on the hiring of their students by an industry that heavily rely on these tools and methods- are widely used. 

​

There is no ceteris paribus on the markets. It is an environment where nature is not neutral. Repeating an "experience" is essentially not possible. The possibility of forecasting has therefore to be considered under a double aspect : demystifying the pseudo-scientific wordage and knowledge of technical analysis (in its two forms, statistical or pattern recognition) on one hand and, on the other, understanding how unscientific the current "science" of financial engineering is, from the Sharpe ratio to asset allocation theory and risk control. Any CAPM used, even modified, enhanced, turbocharged, as it or in its consequence, is wrong. In another article (to be soon translated), we explain how these wrong modelizations aggravate or cause, by propagation, the reasons for the overshooting during in severe market downturns and overpricing in market bubbles.  

​

All these elements suggest that the way price action is considered has to change.  

​

​

​​

* https://doi.org/10.1016/j.ssaho.2021.100244

​

​

​

​

​