This post is essentially a reference to my month-old tweets.

1. Nature ≠ NPJQI
2. Risk measures like VaR were partly responsible for 2008. (https://t.co/t2T67Kas4m)
3. Quadratic speedups for monte carlo are boring.

— Sanketh Menda (@sgmenda) March 3, 2019

Academics have been warning us about the dangers of VaR since 1995: (https://t.co/1NhdjgxRb2) (https://t.co/1NhdjgxRb2). If you want a more extreme take on it, see Taleb: (https://t.co/rTlJNBIXLw). The basic idea is that VaR is extremely sensitive to model specification.

— Sanketh Menda (@sgmenda) April 9, 2019

After 2008, Taleb also spoke to Congress about the risks of using VaR. (Ignore the description under the video.) You can see the full hearing here. More generally, see the Criticism section on VaR’s Wikipedia page.