Probabilistic aspects of finance (Q373529)

The paper is devoted to the analysis of the modern mathematical aspects of financial markets. A brief historical review is presented. It is emphasized that there is a broad interdisciplinary consensus across mathematics, finance, and economics that the discounted price fluctuation of a liquid financial asset should be viewed as a stochastic process even if the central limit theorem does not work. So, the standard setting in mathematical finance is probabilistic, and it involves two types of probability measures. In view of this, market efficiency is related to the existence of a martingale measure. Derivatives together with the paradigm of perfect hedging are discussed. A great attention is paid to the notion of Knightian uncertainty that distinguishes between ``risk'' and ``uncertainty'' in the context of economic decision theory. Price formation, market microstructure, and the emergence of algorithmic trading are studied. The reason to consider algorithmic trading is an increasing need to complement the classical microeconomic picture of noise traders and information traders by taking into account a variety of trading algorithms which are actually used on the financial market. This may make the analysis of the resulting price dynamics more tractable, since the structure of trading algorithms is more transparent and easier to model than the behavioral characteristics of individual agents. The key to understanding algorithmic trading and its potential benefits and risks is the phenomenon of price impact which is studied in the concluding section.