Queue Imbalance as a One-Tick-Ahead Price Predictor in a Limit Order Book

Дата: 
11/12/15
Авторы: 
Martin D. Gould Julius Bonart
Аннотация: 
We investigate whether the bid/ask queue imbalance in a limit order book (LOB) provides signi cant predictive power for the direction of the next mid-price movement. We consider this question both in the context of a simple binary classi er, which seeks to predict the direction of the next mid-price movement, and a probabilistic classi er, which seeks to predict the probability that the next mid-price movement will be upwards. To implement these classi ers, we t logistic regressions between the queue imbalance and the direction of the subsequent mid-price movement for each of 10 liquid stocks on Nasdaq. In each case, we nd a strongly statistically signi cant relationship between these variables. Compared to a simple null model, which assumes that the direction of mid-price changes is uncorrelated with the queue imbalance, we nd that our logistic regression ts provide a considerable improvement in binary and probabilistic classi cation for large-tick stocks, and provide a moderate improvement in binary and probabilistic classi cation for small-tick stocks. We also perform local logistic regression ts on the same data, and nd that this semi-parametric approach slightly outperform our logistic regression ts, at the expense of being more computationally intensive to implement.
Описание: 

Исследуется влияние имбэлансов стакана. Анализируется логистическая регрессия от перекосов имбэланаса. Вроде как на 95% уровне значимости модель работает. Использовали топ-10 акций по ликвиндности на NYSE. Кроме того, отдельно смотрели акции с большой относительной величиной тика (0.01/price)  и малой. Лучше работает в бумагах с большими тиками.