Networks of Heterogeneous Expectations in an Asset Pricing Market

The paper studies the effect of information networks on learning to forecast in an asset pricing market. Financial traders have heterogeneous price expectations, are influenced by friends and seem to be prone to herding. However, in laboratory experiments subjects use contrarian strategies. Theoretical literature on learning in networks is scarce and cannot explain this conundrum (Panchenko et al., 2013). The paper follows Anufriev et al. (2014) and investigates an agent-based model, in which agents forecast price with a simple general heuristic: adaptive and trend extrapolation expectations, with an additional term of (dis-)trust towards their friends' mood. Agents independently use Genetic Algorithms to optimize the parameters of the heuristic. The paper considers friendship networks of symmetric (regular lattice, fully connected) and asymmetric architecture (random, rewired, star). The main finding is that the agents learn contrarian strategies, whcih amplifies market turn-overs and hence price oscillations. Nevertheless, agents learn similar behavior and their forecasts remain well coordinated. The model therefore offers a natural interpretation for the difference between the experimental stylized facts and market surveys.