Lognormally distributed share returns: Difference between revisions
From ACT Wiki
Jump to navigationJump to search
imported>Doug Williamson m (Spacing 22/8/13) |
imported>Doug Williamson (Layout.) |
||
Line 1: | Line 1: | ||
If share returns are lognormally distributed it means that the logarithm of [1 + the share return] has a normal probability distribution. | If share returns are lognormally distributed it means that the logarithm of [1 + the share return] has a normal probability distribution. | ||
Normal distributions have infinitely long ‘tails’ both upside and downside - so implying unlimited downside potential when used for modelling share returns. | Normal distributions have infinitely long ‘tails’ both upside and downside - so implying unlimited downside potential when used for modelling share returns. |
Revision as of 13:31, 6 May 2016
If share returns are lognormally distributed it means that the logarithm of [1 + the share return] has a normal probability distribution.
Normal distributions have infinitely long ‘tails’ both upside and downside - so implying unlimited downside potential when used for modelling share returns.
But the theoretically worst outcome for a share investor is to lose the whole of their investment - in other words a negative return of -100%.
It is not theoretically possible to suffer a return of worse than -100%.
Lognormal distributions - unlike normal distributions - also have a limited downside, so they do not suffer from this theoretical shortcoming.