忽略极端情况最后但并非最不重要的一点是，VaR估算是基于不充分的数据来

Ignore extreme cases, last but not least, it is, VaR estimates are based on insufficient data to calculate, this is another limitation of VaR. Some economists argue that, based on historical data to predict the future value is not appropriate, because VaR no reference value (Damodaran, 2007). Although suitable for use historical data to estimate VaR, but still not enough data to estimate, in particular credit risk and operational risk. Therefore, the data is insufficient to limit the application of VaR model can solve the problem (Danielsson et al., 2000) by a bootstrap method. At the same time, there are other limitations, such as difficult to assess the validity of VaR backtesting and higher than the frequency of the frequency regulation, which limits the use of VaR models.

Ignoring extremes last but not least, VaR estimates are based on insufficient data, another limitation of VaR. Some economists argue that it is not appropriate to predict future value based on historical data because VaR has no reference value (Damodaran, 2007). Although it is useful to use historical data to estimate VaR, there is still not enough data to estimate, especially credit and operational risks. As a result, insufficient data limits the application of the VaR model, which can be solved by self-raising (Danielsson et al., 2000). At the same time, there are other limitations, such as the difficulty of evaluating VaR correctness and the higher regulatory frequency than the backtest frequency, which limits the use of the VaR model.

The last but not least point of neglecting extreme cases is that VaR Estimation is based on insufficient data, which is another limitation of var. Some economists believe that it is inappropriate to predict future value based on historical data, because VAR has no reference value (Damodaran, 2007). Although it is applicable to use historical data to estimate VaR, there is still not enough data to estimate, especially credit risk and operational risk. Therefore, the lack of data limits the application of VAR model, which can be solved by bootstrap method (Danielsson et al., 2000). At the same time, there are other limitations, such as the difficulty in evaluating the correctness of VaR and the higher frequency of supervision than the back testing frequency, which also limits the use of VAR model.<br>