Upside versus Downside Stocks

There is more upside in downside, and more downside in upside stocks than expected.

April 2023. Reading Time: 10 Minutes. Author: Nicolas Rabener.


  • Stocks can be ranked by their upside and downside betas to the S&P 500
  • Results in strong sector biases and factor exposures
  • Excess returns from upside stocks were negative, zero for downside stocks


Most capital allocators use correlation to identify strategies that may add diversification benefits to their portfolios. However, average correlations are often misleading. For example, a market-neutral merger arbitrage strategy is uncorrelated to stocks in good times, but when stocks crash then mergers between companies fall apart. In this case, diversification benefits fail to materialize when most needed (read Merger Arbitrage: Arbitraged Away?).

A perhaps better metric for evaluating diversifying strategies is downside betas to the S&P 500, which represent the betas calculated by only using the negative returns of the stock market. If a strategy has a negative downside beta, then it would generate positive returns when stocks decline.

Although downside betas are most useful for reviewing asset classes or funds, the concept can also be used for stock selection, which we explore in this article.


We consider our universe as all stocks traded in the US stock market with a market capitalization above $1 billion. We calculate the upside and downside betas for each stock for the trailing 12 months, and create two portfolios by selecting the 10% of stocks with the highest upside and 10% of stocks with the lowest downside betas. These portfolios are rebalanced on a monthly basis and stocks are weighted equally.

We observe that the median beta of the upside beta portfolio was 1.9 in the period from 2002 to 2023, which can be compared to a leveraged position in the stock market. In contrast, the median beta of the downside beta portfolio was 0.4. It is interesting to observe that the downside beta was never negative during those 20 years, ie almost all stocks were positively correlated with the stock market. It seems that the diversification benefits from stocks are limited.