Research Question: Does a diversified portfolio benefit investors in terms of returns or volatility compared to the broader market? Is maximizing risk-adjusted returns or minimizing volatility more critical when constructing a portfolio?
The S&P 500 is a popular index tracking the top 500 US firms by market capitalization. It is often used as a benchmark for the US stock market and economy. This analysis uses the SPDR S&P 500 ETF, commonly known as SPY, as a benchmark since it seeks to mirror the S&P 500.
To understand the composition of the S&P 500, we used the Rvest package to scrape the list of companies in the index from the Wikipedia page (https://en.wikipedia.org/wiki/List_of_S%26P_500_companies). The image below shows the composition of the index by sector. Apart from ‘Industrials,’ we can see that the index is skewed towards service sector companies in the Financial, Information Technology, and Healthcare sectors.
To construct a robust portfolio, we must ensure that its constituents are not highly correlated. This minimizes the portfolio’s volatility. In finance, this is called ‘diversification’ and is analogous to the saying ‘don’t keep all your eggs in one basket’; in this case, don’t heavily invest in the same or similar sectors. We pulled data from 202301 to 202405 to identify sectors with lower correlation and constructed a correlation matrix below. All sectors are positively correlated since they are exposed to the same market risk, i.e., when the US economy does well, they tend to do well. However, we can also see that some sectors have a lower correlation; for example, IT and utilities have a correlation of 0.23. Similar correlations are also observed between IT and Energy/Consumer Staples. This aligns with the popular theory that growth stocks (service sector) tend to correlate poorly with value stocks (manufacturing/mature businesses).
However, we are still exposed to market risk since all stocks fall into the equity asset class. To minimize this, it is also essential to consider other asset classes, such as bonds and commodities, since investors move money out of the market and into bonds and commodities like gold during bad market conditions. Our analysis added popular bond and commodity ETFs to our portfolio.
As mentioned above, diversification is key to creating a portfolio with consistent returns. In our analysis, we decided to use the following financial instruments, which can be broadly classified into three main types: equity, fixed income and commodities.
Asset Class | Type | Name |
---|---|---|
Equity | Growth | Dell, Nvidia, Micron Technology, Uber, Amazon, Eli Lilly |
Equity | Value | General Electric, Hilton, Targa Resources, JPMorgan, Amex |
Bond | Fixed Income | Vanguard Total Bond Market ETF |
Commodities | Commodities | iShares S&P GSCI |
Data: We pulled pricing data from 2023-2024. The first 18 months, 202301 - 202405, would act as our in-sample data, allowing us to adjust portfolio weights per our strategy. The last 6 months, 202406-202412, would be an out-of-sample test set.
Strategies: To understand the role of diversification, we explore three strategies
The weights assigned during the training period are left as they are and applied to the test period to see how each strategy performs on future data.
During the training period, the Equal Weight Portfolio and Max Sharpe Portfolio have the highest returns, followed by the broader S&P 500 and Minimum Variance Portfolio. The drawdown chart below shows that the S&P 500 and Equal Weight Portfolio have the most significant drawdowns or downside risk. This indicates that there is some value to diversification, whether it is to minimize volatility (Min. Var. Portfolio) or to maximize risk-adjusted return (Max. Sharpe Portfolio)
We measured the Sharpe Ratio of all three portfolios and the S&P 500 to give an objective metric to complement the above charts.
S&P500 | Equal Weight | Minimum Variance | Maximum Sharpe | |
---|---|---|---|---|
Sharpe Ratio | 2.14 | 5.29 | 1.89 | 5.98 |
Unsurprisingly, the maximum Sharpe ratio portfolio had the highest value during training. But to evaluate each portfolio, it is necessary to look at its performance on future data. Thus, we simulate the portfolios using the saved weights on pricing data from 202406 to 202412.
S&P500 | Equal Weight | Minimum Variance | Maximum Sharpe | |
---|---|---|---|---|
Sharpe Ratio | 1.68 | 1.06 | 1.72 | 0.96 |
Evaluating the portfolios out of time gives a few interesting results
Research Question: Does a diversified portfolio benefit investors in terms of returns or volatility compared to the broader market? Is maximizing risk-adjusted returns or minimizing volatility more critical when constructing a portfolio?
From our analysis and the portfolios constructed, we can conclude that: