Measurement of Market Risk – the Value at Risk Method
PThe problem of measuring and controlling risk emerged with the advent of new trends in global financial markets in the mid-1970s. At that time, the volatility of financial parameters increased dramatically, which directed the interest of traders towards methods designed to mitigate newly emerging risks. There was then a natural increase in demand for futures contracts, which, in response to specific investor needs, stimulated the diversification of these instruments [Jajuga, Kuziak, Markowski, 1998]. The new financial products, although intended to control risk, were themselves a new source of risk. In addition, the phenomenon of globalisation emerged at this time, signifying the emergence of strong linkages between markets, which is associated with an increase in the correlation of financial variables. All these factors interacted with each other, further compounding each other. In this situation, risk management, whose overriding objective is to protect institutions from unacceptable levels of loss, has undoubtedly become much more difficult [Grabowska, 2000, p. 34]. Recently, the financial world has been inclined towards the view that possible adverse deviations from expected values should be taken into account when measuring market risk. This leads to the concept of risk measures, of which Value at Risk methods have become the standard [Jajuga, 2000, p. 112; Jajuga, 1999, p. 64]. A user wishing to apply the Value at Risk method has a choice of several possible ways of calculating it, each of which has at least one parameter whose magnitude can be flexibly selected [Bałamut, 2002, p. 93]. The models constructed may differ, for example, in the assumptions made about the distributions of risk factor returns, the methods of aggregating different risk categories, the treatment of options, etc. The use of different Value at Risk estimation methods often leads to inconsistencies in risk assessment, so it is particularly important to know the assumptions and their consequences.
