LSTM forecasting foreign exchange rates using limit order book
LSTM forecasting foreign exchange rates using limit order book
复制标题
DOI：
10.1016/j.frl.2021.102517
复制
发表时间：
2021-10
影响因子：
10.4
通讯作者：
Katsuki Ito;H. Iima;Y. Kitamura
中科院分区：
文献类型：
--
作者：
Katsuki Ito;H. Iima;Y. Kitamura
我们使用长期和短期记忆（LSTM）来预测外汇市场的日内回报。作为预测器，我们使用限价订单簿中的事件。与其他模型相比，我们的模型预测了1分钟中间报价回报的运动。当我们考虑买卖价差时，这种预测不会带来经济收益。这一结果表明，这些事件可以有助于价格发现和研究的市场有效地设置价差。
We use long and short term memory (LSTM) to predict intraday returns in foreign exchange markets. As predictors, we use events in the limit order book. Compared to other models, our model predicts the movement of a 1-min midquote return. When we consider the bid-ask spread, this prediction does not bring economic gains. This result indicates that these events can contribute to price discovery and the studied markets efficiently set the spread.