This paper describes an empirical analysis of optimized portfolios and safe return rates across multiple investment time horizons using Telser’s Safety-First method. The analysis uses thirty years of historical monthly data for 81 different Fidelity® mutual funds and a blended money market fund rate. The Fidelity® funds represent a wide variety of investment factors, strategies and asset types, including bonds, stocks, commodities and convertible securities. A large synthetic return dataset was generated from this data by a Monte-Carlo random walk using cointegrated bootstrapping of investment returns and yields. Portfolio optimization was then performed on this synthetic dataset for safety factors varying from 60% to 99% and time horizons varying from one month to ten years. Results from portfolio analyses include the following: 1) there are no risk-free investments available to Fidelity® mutual fund investors, as even money market funds have risk due to yield fluctuations, 2) optimized portfolios are sensitive to both investment time horizons and safety factor confidence levels, 3) conservative, short-term investors are better off leaving their money in a money market fund than investing in securities, 4) optimized portfolios for longer term, more aggressive investors consist of a blend of both value and growth equities, and 5) the funds most often represented in optimized portfolios are those that have the best risk/reward ratios, although this rule is not universal. Two practical applications of this optimization approach are also presented.