To provide a technique for calculating an execution probability of a limit order by creating a probability model accurately reflecting a market trend.

In the execution probability calculation system, execution information and quotation information of a securities exchange are stored in a market data storing part, limit order information is extracted on the basis of time series change of quotation information and stored in a limit order DB, determination of execution, cancelation, unexecution of each limit order is carried out on the basis of the execution information and cancel information and a determination result is stored in a determination result DB, predetermined attribute information relevant to executed and unexecuted limit orders is calculated on the basis of the information of the market data storing part, remaining time of trading relevant to each limit order is calculated, the determination result of execution/unexecution relevant to each limit order, its quantity, the remaining time, and the attribute information are statistically processed to create the probability model per brand, and the probability model corresponding to the current attribute information, the remaining time of trading, and the quantity of the limit order is applied to calculate the probability.