Over the last years, business concepts of new cyber intermediaries in the market of intellectual property have quickly transformed towards aggregating extra large portfolios of intellectual assets. Namely, in our previous work we analyzed aggregator models for digital images, video, audio and literary works (texts). Main focus was solely put on general concepts of managing risks related to large portfolios of copyright assets. In this paper, at first, we study specific nature of a market mechanism involved in sale of such assets. Sales statistics over a long time for a certain digital image is presented graphically in structured price brackets. Secondly, VaR (EaR) methodology, well known in financial engineering, is adapted and applied for the first time for quantitative evaluation of risks related to real intellectual assets (digital images) portfolio. Based on empirical distribution of revenue positions across a portfolio, a nonparametric method of historical simulation is applied to solve the task. Next, divergence of results of different approaches was evaluated including a simple parametric model. Our work proves that VaR (EaR) model can be successfully used for the portfolios of such assets class.