Project CEMAPRE internal
| Title | Value-at-Risk and Expected Shortfall with cryptocurrency portfolios |
| Participants | Rui Louro, Nuno Sobreira (Principal Investigator) |
| Summary | In a previous paper, we conducted an extensive study applying a substantial amount of risk measurement forecasting procedures to a set of stocks representative of the PSI20 index. A comprehensive set of backtests were then applied to the estimates provided by the procedures to judge their adequacy. This resulted in a number of procedures being deemed as providing acceptable results, thus requiring the use of cost functions to select the best ones for each stock. Several interesting results were obtained. In particular, the methods based on Extreme Value Theory (EVT) more often provided the best results. This research has been published in Finance Research Letters. We expect to produce more research in this field but applied to Cryptocurrencies, a virtual currency based on computerized ledgers. Hence, in a second chapter, we would like to analyze which risk measurement forecasting procedures provide better results when applied to various cryptocurrencies. I plan to apply a wider range of forecasting methods and strategies, backtests and cost functions. Furthermore, a detailed study about the properties of this specific type of time series is planned. This research is motivated by the fact that cryptocurrencies are an emerging category of risky assets throughout the financial industry. Even though once dismissed as being no more than irrelevant technology mainly connected to illegal activities, nowadays cryptocurrencies are increasingly seen as a legitimate investment even by well established investment banks. These have been developing financial products based on cryptocurrencies. Their potential is also motivated by their ability to take advantage of an increasingly digital financial world, while providing a platform that limits most actions detrimental to a currency, such as excessive issuance. We also believe that this wider appeal of cryptocurrencies will make risk measurement forecasting an even more important field of research. |