Many branches of modern economics use methods developed in natural sciences. In the article we briefly recall history of financial mathematics, quantitative finance and econophysics. We give examples of ideas developed in physics which have later turned out to be very useful in quantitative description of economical phenomena. In particular, we discuss in detail an application of random matrix theory, a theory which was originally formulated in quantum physics, to the problem of portfolio selection, being one of the most fundamental problems in quantitative finance.