Value at risk with informed traders, herding, and the optimal structure of trading divisions

We scrutinize the use of value at risk as traders' limit in banks. Thereby, we compare a bank with uninformed traders dealing on a perfect capital market, with a bank in which traders receive a noisy signal about the future price of the stock they are dealing in. Additionally, they are able to deduce some information about the market trend from the observation of the behavior of other traders. In the imperfect market setting, informed traders tend to herd in informational cascades, which increases the probability of extreme results and value at risk. Thus, banks should either avoid or optimize information flow between traders. We discuss different optimization approaches to maximize a value at risk-based RORAC through an efficient information policy. Likewise, we compare our results with quot;neoclassicalquot; value at risk both from an ex ante and ex post-perspective and identify systemic risks from neoclassical negligence of informational herding