Survivorship bias and the effect on ERP

[300hoursoverandovera]

Can some one explain why ERP has an upward bias with survivorship bias?

[300hoursoverandovera]

Also, how does the failure to incorporate dividends bias ERP downwards.

And why is ERP inversely related to valuation ratios?

[MrSmart]

If you are calculating the ERP from a broad market index like the SP500, then the companies that do badly drop out of the index, biasing the actual returns on equity upwards, and thus the ERP. Remember that the proxy for risk (ERP) here is historical return.

By the same note, using only price changes, and not total return, would give you lower percieved risk, and thus lower ERP. If ERP is low, then P/Es are high, and asset prices are high. Lower risk means higher price, check the breakdown of GGMs or any multiple.

[300hoursoverandovera]

@MrSmart

If you are calculating the ERP from a broad market index like the SP500, then the companies that do badly drop out of the index, biasing the actual returns on equity upwards, and thus the ERP. Remember that the proxy for risk (ERP) here is historical return.

By the same note, using only price changes, and not total return, would give you lower percieved risk, and thus lower ERP. If ERP is low, then P/Es are high, and asset prices are high. Lower risk means higher price, check the breakdown of GGMs or any multiple.

Thanks for the explaination Mr.Smart. I am still a bit confused though.

So let us say return is high based on survivorship bias, you could relate that to CAPM formula and say Re is higher because Rm goes up?

And if P/E is high that would mean using GGM model your Re has to be lower, thereby meaning ERP is lesser?

Using the above meaning, we apply for dividend yeilds as well?

Could you clarify if my approach to this is correct?

[300hoursoverandovera]

Also, could u explain the reasoning behind events such as War, that have an impact on the ERP?

[MrSmart]

@300hoursoverandovera

Thanks for the explaination Mr.Smart. I am still a bit confused though.

So let us say return is high based on survivorship bias, you could relate that to CAPM formula and say Re is higher because Rm goes up?

Yes.

And if P/E is high that would mean using GGM model your Re has to be lower, thereby meaning ERP is lesser?

P/E should be the output, unless you’re solving for r.

In this case, adjusting the P/E upwards will bring the r, and therefore the ERP down, all else equal.

Using the above meaning, we apply for dividend yeilds as well?

I don’t get this part, but ERP should be calculated with dividends included in your return, else the ERP would seem understated, as you’re neglecting one component of equity returns.

In times of war, the forward looking ERP should be much higher, but that also depends on the movement of the Rfr, as that would be driven up as well. You cannot use historical returns in this case. To illustrate, if you were to solve for the implied ERP using the price level of a broad market index in that country, then the discount rate should be a lot higher (whether due to higher Rfr or not, you need to solve for ERP to figure that out) from the depressed stock prices, and thus reflects the higher percieved risk, which would not nessecarilly reflect higher expected returns, as predicted by the CAPM in this case.