Timeline for Analytical solutions to limits of correlation stress testing
Current License: CC BY-SA 2.5
7 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Sep 20, 2010 at 4:14 | comment | added | shabbychef | ah, I think I see: you want to add a fixed amount to every off-diagonal element? I think my other answer can be adapted to that case. | |
| Sep 19, 2010 at 22:02 | comment | added | Eduardas | Same as before - I am looking for an analytical solution for correlation matrix stress testing. Solutions have been suggested for stress testing utilising multiplication but not addition of stress levels (addition was my original question above). Either way, thank you everyone. | |
| Sep 19, 2010 at 19:55 | comment | added | shabbychef | what exactly is your question, then? | |
| Sep 19, 2010 at 12:15 | comment | added | Eduardas | Indeed this is the same proof as the one provided by Paul Rubin on or-exchange.com. Multiplying each correlation by a constant has a very significant drawback though - it is a very very small stress test. It is possible to apply a much larger stress test by adding a constant to each matrix, not multiplying. However, it seems the analytical solutions for that are a little less obvious. | |
| Sep 18, 2010 at 20:55 | history | edited | shabbychef | CC BY-SA 2.5 |
had misplaced the I
|
| Sep 18, 2010 at 20:11 | history | edited | shabbychef | CC BY-SA 2.5 |
add part about whether k is positive or negative.
|
| Sep 18, 2010 at 5:03 | history | answered | shabbychef | CC BY-SA 2.5 |