Roberto Malnati

SIAT Member. Partner of Ten Sigma Sagl

As a Web and Trading Online pioneer in Italy, Roberto Malnati has developed for more than 20 years quantitative analysis models through expert systems and neural networks in the industrial field and in the wholesale market. By refining the above model capacities he has found in the financial analysis field their optimal use.
He has developed and achieved Luxor Trading System which is used by Financial Institutions and marketed by Sole24Ore. Roberto Malnati has also been one of the founding members of and the developer of FinanzaonlineSpa quantitative analysis models.
At present he works in Lugano and hold a partnership in Ten Sigma Sagl, a company whose skills are proven by the market and credit risk management and risk assessment, management consultancy and financial products distribution to Institutional decision-makers.
Roberto Malnati manages the development of quantitative analysis and risk management models for hedge funds and banks.




The Market Master Algorithm: How Technical Analysis and Neural Networks merge through Reverse Engineering

Technical analysis. Back to future with neural nets. The future of neural nets in a financial environment will go at the same pace as the ability properly “reverse engineer”. Many of the decisional processes found by the neural nets are strongly linked to already-known models and institutions, given by technical and quantitative analysis.
The process of reverse engineering deals with a detailed analysis of the mode of operation of a software (usually a neural net), aimed to produce a new software that works in a similar way, but increasing the efficiency.
Why? Because there’s a high probability that a neural net (a “black box”) will point out the purchase of a title that will be object of a takeover in the following days. Therefore, every signal must give a certain model, which, for every given data, will always give the same result.
Basically, every decision must be completely reachable, and this is only possible with reversed engineering.