In this chapter we consider a modeling approach that uses a set of historical data, such as bond prices, share prices, claim sizes, or exchange rates, to model the value at a future time T > 0 of portfolios whose values depend on a given set of assets and possibly also liabilities. Here we want the data to speak for themselves in the sense that the model for the future values should only be based on information available in the given historical data samples. The assumption we make is therefore that the information in the samples is representative of future values and that no additional probability beliefs of the modeler are relevant.