The basic idea of Monte Carlo simulation method: Assuming that the change of asset price depends on a random distribution, the random price in the target time range is generated by computer simulation, the distribution of asset income is constructed in turn, and then the VaR is calculated.
Because a certain model (such as Brownian motion model) can generate random data to describe the asset price path, the generation of data is a disorderly process (true random), which will consume a lot of time and affect the simulation efficiency.
In order to improve the efficiency of simulation, the amount of data must be fundamentally reduced. In order to reduce the amount of data, the introduction of subtraction factor is equivalent to adding an access condition to the adoption of data, and the standard for introducing subtraction factor may be to exclude some extreme conditions. It also changes from true random to pseudo-random, which improves the efficiency, but also sacrifices the accuracy of the results.
Answer I hope you can be satisfied!