Inferring trader behavior from transaction data: A trade count model
- 130 Downloads
We present a model that uses trade counts to infer the arrival of private news and the probability of informed trading (PIN). Although similar in approach, our model avoids problems with factor-driven biases and standard errors associated with estimating the buy-sell model of Easley et al. (1996). In particular, tests using the probability of informed trading may suffer from spurious correlations between the Easley et al. (1996) PIN and firm or market characteristics. Results for our model suggest that trade counts, independent of trade direction, are able to capture important features of a firm’s information environment. (JEL C51, D82)
Unable to display preview. Download preview PDF.