# Portfolio Choice in a Multi-Asset World

### Portfolio Choice in a Multi-Asset World

The course work asks you to choose a portfolio of four publicly traded assets and create a portfolio according to an investment idea.
Explain your choice of assets and your investment strategy.
Get the assets’ daily prices for the last 3 years of trading approximately.
You may choose one from the following three portfolio strategies and pick four suitable assets that provide a good fit for this strategy:

• ESG portfolio (assets focused on sustainability, be it bonds or stocks)
• Brand equity portfolio (assets focused on entities (firms) that have high brand equity, i.e., whose value proposition is based on its product brand name)
• Bond portfolio (bonds – corporate or sovereign – as an asset class instead of equities)
##### For each stock, calculate the performance over the past year:

I. the annual rate of return
II. the annualized standard deviation

• Plot all four stocks’ past year average performance in a graph in RI – σi space. Does any stock dominate the others on the basis of the mean-variance criterion (taken for each stock in isolation)? Discuss.
• Compute the correlation coefficients between each pair of stocks and present the results in a 4 x 4 correlation matrix and explain what this may imply for your portfolio.
• Compute the Minimum Variance portfolio of the four-asset universe, i.e., allowing for positions in all four of your chosen assets. Consider whether it makes a 2 difference for your selected assets (and data sample) to restrict short selling or not.
• Choose an appropriate proxy for the risk-free rate and explain your reasoning. Using this proxy, compute the Tangency Portfolio in your four-asset universe.
• Evaluate your Tangency Portfolio against a common benchmark. Which benchmark is reasonable to compare your portfolio to? How do Sharpe Ratios compare?
• General discussion: this “optimal” Tangency Portfolio came from a small selection of stocks. Does it look like a reasonable portfolio to offer possible clients? Are there aspects that you would hesitate to implement directly without further study? Are there data issues that you believe may have affected your findings?