نوع مقاله : مقاله پژوهشی
نویسندگان
1 دانشجوی دکتری تخصصی مهندسی مالی، گروه مدیریت، واحد رشت، دانشگاه آزاد اسلامی، رشت، ایران
2 گروه مدیریت بازرگانی، واحد رشت، دانشگاه آزاد اسلامی، رشت، ایران (نویسنده مسئول)
3 گروه مدیریت، دانشگاه گیلان، رشت، ایران
چکیده
کلیدواژهها
عنوان مقاله [English]
نویسندگان [English]
Considering the extreme fluctuations of the cryptocurrency market and also the importance of predicting the value at risk in such conditions, the purpose of the present study is to predict the value at risk in the cryptocurrency market and also to compare different models for predicting the value at risk. In addition, the impact of different distributions of model innovation terms has been investigated. In this research, we use different models to predict the value at risk of return of four well-known cryptocurrencies. The data used in the research covers the period from 1/1/2018 to 16/3/2022. This research uses CAViaR and DQR models that directly predict the return distribution quantiles as value at risk. In addition to the mentioned models, several types of common models have been used to predict value at risk. In order to check the performance of the used models, we have used the back-test method, which is one of the common methods for testing the performance of the models. The results show that the models that directly use the quantiles of the return distribution to predict value at risk (specifically CAViaR and DQR models) have a much better performance than other common models for predicting value at risk.
کلیدواژهها [English]