Just want to make sure we are talking the same language. Let's imagine we have a betting slip which pays €2 if a coin toss comes up Heads and zero otherwise. A fair (efficient?) price for this slip "reflecting the risk" would be €1. However, for investors anyway, they should want compensation i.e. a premium for the risk as well as a reflection of it. They might only give you 80c.If the market was efficient, the price would reflect the risk without a premium.
All the evidence shows that over the last 100 years (for successful economies) equities have outperformed "risk free" assets considerably. But it is impossible to tell to what extent that was because there was a risk premium or to what extent it was because things came up Heads. I used the qualification "successful economies" and indeed the Equity Risk Puzzle was based on the US and given the very propitious environment for economic development (technology etc.) over the last century, my own view is that there is no puzzle. The 6% p.a. outperformance of equities could indeed be a 1% p.a. risk premium, which is what the authors claimed was justified, and 5% p.a. because Heads came up (10 times in a row at least for the US IMHO).