Warning: preg_replace(): The /e modifier is deprecated, use preg_replace_callback instead in ..../includes/class_bbcode_alt.php on line 1270
The Actuary.com Actuarial Science Forums - Discuss Actuarial Exams , Jobs, Tests, Study Materials, Books, Courses, Seminars, Schools - Actuarial - Ask the Professor http://www.actuary.com/actuarial-discussion-forum/ Discuss anything with a professor of actuarial science. en Thu, 19 Oct 2017 10:54:59 GMT vBulletin 60 http://www.actuary.com/actuarial-discussion-forum/images/misc/rss.png The Actuary.com Actuarial Science Forums - Discuss Actuarial Exams , Jobs, Tests, Study Materials, Books, Courses, Seminars, Schools - Actuarial - Ask the Professor http://www.actuary.com/actuarial-discussion-forum/ Explanation of Value at Risk by importance sampling http://www.actuary.com/actuarial-discussion-forum/showthread.php?64539-Explanation-of-Value-at-Risk-by-importance-sampling&goto=newpost Fri, 13 Oct 2017 09:14:03 GMT Hello Everybody, I am studying... Hello Everybody, I am studying statistics applied to actuarial science, specially non life insurance mathematics. I am studying the method of importance sampling to get an estimator of the value at risk.

I am confused with somethings. I would like to understand how is the usage of value at risk in practice in the insurance field. Because I don�t have clear the reason fo simulate this risk measure, I mean If have my data set I can adjust my dataset to any loss distribution and the I can compute the probability of ruin and estimate the value at risk. But Why I need to perform a simulation? How to get the initial capital the premium rate? which values can I use? why is the importance of the importance sampling?

I am clear that I need to estimate aprobability of loss, in some cases the probability of loss can be very small depending of the initial capital, then with importance sampling method I will give more "importance" to the important values with low probability. But I am very unclear with the benefits in practice about the usage of this method.

Regards ]]>