Exercise

# Which distribution?

It's often hard to initially select how to represent a loss distribution. A visual comparison between different fitted distributions is usually a good place to start.

The `norm`

, `skewnorm`

, `t`

, and `gaussian_kde`

distributions are available. Their fitted estimates of the available investment bank portfolio `losses`

from 2007 - 2008 are displayed in the `plt.figure(1)`

object, which you can show.

Create a new figure and plot a histogram of portfolio `losses`

using `plt.hist(losses, bins = 50, density = True)`

. **Using this histogram for comparison**, which distribution(s) in `plt.figure(1)`

fit `losses`

best?

Instructions

**50 XP**

##### Possible Answers

- Normal
- Skewed Normal
- T
- Gaussian KDE
*Both*T and Gaussian KDE