What this chart shows is that the idea you can get reliable six percent returns every year is utter garbage, completely contradicted by the facts, and that there are extended periods of negative returns entirely possible. If you want to rationalize a continued belief in compound interest, go nuts, but I should warn you that this irrationality is likely to cost you, in your real life, quite dearly, vs. actually engaging with the truth of the market return rate. Your choice. Doesn't much matter to me in the end.
But as you can see there isn't a single example on this chart where someone could add to their portfolio over ~50 years and entirely lose money.
> there are extended periods of negative returns entirely possible
I mean, the very worst extended period is the 20-year range from 1961-1981 where -2.0% was lost year over year. I never argued against this, just that over the long run it is smoothed out by positive returns over time.
The comments you're making are much more emotional than fact-based:
- "utter garbage"
- "rationalize a continued belief in compound interest"
- "irrationality is likely to cost you, in your real life, quite dearly"
- "actually engaging with the truth of the market return rate"
The chart itself, and the way virtually every American structures their retirement accounts doesn't reflect your interpretation.
Even if you don't want to call it a flawed assumption, at best it's misleading to argue "compound interest is a myth" with the evidence of linking a chart that shows that's only true if investing in a way in which "nobody invests that way".
If you link to a chart that shows when compound interest _could_ lose, but only under conditions in which no one actually invests, then I don't consider that sufficient support to dismiss it as a "myth". Actually, if anything, it supports how true compound interest is. As long as you're willing to make the standard assumption of diversifying deposits over time to build it every year (i.e., dollar-cost averaging), which virtually every person investing does.