Comment by scythe

>Not really. It has fluctuated a lot. You can pick starting and ending points a few years apart and come up with very different results relative to actual inflation.

I think he's talking about really long periods of time. It's true that gold is an extremely volatile investment, whose price can seemingly quadruple or be cut in four at any time. But if you look over periods where the price of gold increased by more than 20x, this becomes a lot less important when you try to estimate things like the average rate of inflation. If you work with a ten-year moving average of the price of gold the problem is also reduced. Gold is the only metal whose sulfide is unstable under standard conditions (101.3/293.15).

In other fields, this is called a "low-pass filter".