This paper applies a forward-looking approach to the minimum variance portfolio optimization problem for a selection of 100 stocks. The purpose is to determine which market conditions favor this strategy of using option-implied information. Out-of-sample volatility, the Sharpe ratio and certainty equivalent return have been measured against eight benchmarks, including the equal-weighted 1/N and minimum variance portfolio based on historical estimates. Equivalent or superior performance is evident in terms of reduced volatility and higher certainty equivalent return. However, strict outperformance of the best benchmarks is only seen when option-to-stock volume ratios are high and information signals in the options market are strongest.