Comment by nabla9

For synthetic distributions Mandelbrot's Multifractal is good enough.

You may consider using real historical distributions derived from real data. You can record different distributions from different time periods and different economic conditions.