These models can be very complex for most people, but a trader does not need to control the types of financial engineering to understand the situation of his trading house and decide. For our trader, even if chance cannot be predicted or controlled, it is no less evocative. Mathematical models can be brilliant, but the more they are, the more they may tend to model the world as we would like it to be. In fact, the problem is not as mathematical models, but how we use it.

A short review of quantitative finance in France under the lens of the Next Finance website: Profile of quants, origins and outlook of the industry.

In terms of uses and we crossed the yellow line with the quants, which appeared on the front of the stage after the sudden financial crash of May 6, 2010 (Wall Street plunged 9% for 20 minutes). Simply put, the quants are mathematicians and computer programmers gathered in the trading rooms of major investment banks of the global financial system. The Quants are at the heart of another revolution in finance: trade in financial assets at the speed of light via mathematicians robots.

The objective of quants is to avoid future financial crises by "quantifying" human behavior in economics and leaving aside the hazard and randomness. And yes, with Quants human behavior is deterministic in the markets and the market is not random!

You knew the reign of financial mathematical models based on random data, you liked the Monte Carlo simulation. Now all this according to the Quants, belongs to the conventional academic environment! Today, the real world are the quants and their robots.

But the problem of quants is not only a problem of robots. We believe, indeed, that their work almost always lead to systematic underestimation of rare events or designated as extreme.

The statistical assumption of normal (log-normal) distribution of stock returns (prices) is not that strong and tail events’ occurrence is largely undervalued. Nevertheless, this modeling framework has been widely used for strong (...)

Indeed, in financial markets, mathematical modeling based on the statistical assumption of normality of prices and yields distributions has emerged in three main areas: risk quantification and their supposed right hedging, the pricing of options; asset management. Certainly, more sophisticated modeling can be used to describe the behavior and risks of complex financial portfolios, in which case we move from a parametric method based on the normal distribution with a method called Monte Carlo simulation based on different scenarios that may follow a shape similar to scenarios experienced in the past but are not limited by history (in this case the market variables are not considered as random variables following a normal distribution)

But we can say generally that it is still often assumed that the statistical and probabilistic Gaussian environment (referring to the normal Laplace-Gauss) is adorned with all virtues in the world of modern finance.

banks and investment banking (CIB) were supposed to perfectly control their market risks with a "magic" tool called Value at Risk. We see instead that in times of stress no one have control over anything;

asset management subsidiaries of these banks were supposed to optimally manage the savings surplus of individual and institutional investors;

Finally, economic agents could hedge their financial risks from the CIB through the use of derivatives perfectly priced.

From a technical standpoint, the Black and Scholes formula to evaluate the price of some of these products could give scientific legitimacy to the new environment. And for top management, simplicity and speed of calculation are the arguments against which no one, or almost, can fight: after all it was it is just to know how to integrate a differential equation (sorry for non-mathematicians); efficient computing capacity and therefore not weighing too heavily on bank operating ratios.

In his latest book, Philippe Herlin, in light of the crisis and following the work of Mandelbrot and Taleb, presents the agenda the question of the reliability of the classical models used in finance.

Fortunately we have often heard of internationally renowned scientists warn us about the dangers of the methodologies used for modeling financial risk management and asset pricing? One naturally thinks the late mathematician Benoit Mandelbrot disappeared in October 2010. But he was too alone to face the political and economic leaders and his work remain largely underutilized for at least three major reasons:

1/ To the world leader, this unanimist type of modeling is ideal since mathematics (at least some use of them) allow to self-persuade that major disasters are almost impossible

2/ Then minimizing extreme risks enable to continue to fund waste and excess debt that led to bubbles. There is behind this two types of findings: often the greed of some institutions highly paid to sell financial products whose risk is clearly undervalued, but also the need to justify the waste of our bloated and incompetent bureaucracies. All this to ensure continued funding of public deficits, which leads to universally consider that the overvaluation of governments debt under scrutiny is not considered risky (which is false)

3/ The minimization of risks by the quants also legitimate a model of development:

That of abnormally high profitability standards in terms of economic fundamentals (again greed and profit inefficiently reused)

There is also the financial engineering of banks aiming to transfer risk on some private economic agents to save capital and improve again and again the return on equity employed.

**What should we do?**

First, we must continue to invest in research to improve mathematical modeling (which will always include deficiencies by construction).

While complex finance should be preferred to improve risk modeling, it continues, in times of market stress, to help designing unmanageable and useless complex structured products

But of course, do not persist. We must above all try to better understand the markets. It is often said that being a good professional in market research, trading and structuring is also being at once a good mathematician, a physicist and a good computer scientist(in the service of what you say?) . If you can be all these at once would be perfect but if you not interested also in the field of behavioral finance, all this impressive curriculum is useless.

One thing is certain: it will never be possible to model fear, mimicry and even less the impact of regulatory, prudential and accounting on investor behavior. But it is the understanding of these phenomena that can help managing and apprehending "modern" finance crises.

In total for us, the quants symbolize greed, fear and the randomness and obscure Wall Street. Certainly they were called to order by the securities and exchange commission, but do you seriously believe that things are different today?

There is no shame to accept uncertainty and randomness of this world. We believe that the simple recognition that uncertainty is inevitable is already changing things in a fundamental way. However, this recognition must be genuine by overcoming our aversion to uncertainty. We must learn to take the uncertainty and ambiguity in trade and remember that it is not a science but an art.

We are for intelligent uses of mathematics in finance and not by setting equation of human behavior. Because the human charm is to surprise its environment and for that he has no need of robots transmitting price instructions, at the speed of light, for example. Because it will be very difficult or impossible to control such systems. Experience it tonight in your car, change the instructions of your air conditioner, not at the speed of light, but let’s say every 3 or 4 seconds (select 17 °, then 3 seconds after 23 °, then after 3 seconds 19 ° and 27 °, ...) and tell us what is the temperature inside your vehicle after 10 minutes (which is "relatively a small play" in relation to a day in a trading room with robots that transmit financial instructions at the speed of light)