1. Can a portfolio manager diversify away to a negligible level both the common factor risk and the specific risk of a portfolio?
2. Is a bottom-up stock-picker portfolio with 50 stocks riskier than a top down factor-betting portfolio with 500 stocks?
3. Can a portfolio manager use a multifactor risk model to predict if an active portfolio is at greater risk of underperforming its benchmark than outperforming it over the next period? That is, does the "risk model" predict "risk" in the conventional sense?