[Retric]

I would probably go for some middle ground such as if (year > 0) then shares * Square root (years).

Just because someone has held a stock for 20 years does not necessarily mean they are currently interested in the long term. But, it probably points in that direction.

[rocqua]

You might want an upper limit, so either logistic growth [1] or something like 1 - exp(-(T + T_0)) feels like it makes more sense.

I especially like logistic growth for having slow growth at the start and end, only growing quickly in the middle.

[1] https://www.khanacademy.org/science/biology/ecology/populati...

[SomeStupidPoint]

You don't need an upperbound as long as your function is sufficiently slow growing.

Example:

Log(years) doesn't have a bound, but is unlikely to cause problems over any reasonable time scale.

[jshaqaw]

Or it means they are so utterly passive and unthinking that they are just the voters entrenched management dreams about.