Equity Risk Premium

CFA Level II
#1

Hi there,

I am a little unclear about some of the basic theory behind the equity risk premium.

As i understand it, if the economy is in good times, expansion, profitable, the equity risk premium should be lower. In poor times, there is more risk, so the premium goes up.

If that is the case then why does survivorship bias increase the equity premium. Are we not including the strong performing companies which therefore should reduce it?

Thanks

#2

Further, if we have a string of historical inflation surprises and productivity, then surely this indicates a healthy economy which would = low premiums. But as they are surprises and not normal if we were to predict the forward ERP would we not need to adjust it back upwards to take into account more normal risk levels? The answer suggests we would need to adjust downwards.

#3

Historical Equity Risk Premium (ERP) = Excess return over risk free rate. Higher equity returns result in higher ERP (assume that Rf is flat). Inflated historic performance as result of survivorship bias or a string of positive surprises results in inflated historical ERP. That is why historical ERP should be adjusted downward. Expected/Forward ERP = Risk perception/assessment in the future. And this is what you are thinking about when you say “As i understand it, if the economy is in good times, expansion, profitable, the equity risk premium should be lower.” Also, keep in mind that GGM and Ibbotson-Chen models are supply-side models and use long-term risk free rate, and CAPM and Fama-French are equilibrium/multifactor models and use short-term/current risk free rate.

1 Like
#4

Cheers Sceptic! It helps

#5

Survivorship bias means that the failed firms (or poor performers) aren’t included in whatever index you’re using to proxy the market (in other words, the strong performers with higher returns are overrepresented relative to poor performers). This means that the market proxy is composed of returns for firms that performed better relative to the ones that were excluded. The Rm is then biased upward because the firms that did not “survive” were not included (lower returns excluded). A bigger ERP is the result with survivorship bias.