Modelling of stock price changes: a real analysis approach (Q5926471)

In continuous-time financial mathematics the solution to the Doleans-Dade stochastic differential equation is often used as a model for stock price changes. The semimartingale driving this equation is called the return. Since many conclusions on the price behavior depend on the return, it plays an important role in mathematics of finance. On the other hand, the returns in econometrics of financial markets are sometimes modelled by stochastic processes which are not semimartingales. To provide a theoretical justification for such cases, it is introduced in the paper a Doleans-Dade type equation with the stochastic integral replaced by a different integral. One may ask then whether it is possible to build up a model of stock price changes which is independent of a particular integration theory? The present paper addresses this question and provides a new insight into the relation between theoretical and applied financial mathematics. A stock price and its return are defined in a duality to each other provided there exists suitable limits along a sequence of nested partitions of a time interval, mimicking sum and product integrals. It extends the class of stochastic processes susceptible to theoretical analysis. Also, it is shown that extended classical calculus is applicable to market analysis whenever the local \(2\)-variation of sample functions of the return is zero, or is determined by jumps if the process is discontinuous. In particular, an extended Riemann-Stieltjes integral is used in that case to prove several properties of trading strategies.