The Low Beta Anomaly: Why higher returns doesn’t have to equal higher risk

To earn higher returns, we need to assume higher risks…this is what we’ve been taught.  

While I agree with this premise as it pertains to asset classes (i.e. stocks have dramatically outperformed safe Treasury bills historically), I disagree as it pertains to individual stocks.

In his paper titled “Risk Neglect in Equity Markets”, Malcolm Baker of Harvard Business School assembled two portfolios from 1967; one consisted of the 30% of US stocks with the lowest beta; another consisted of the 30% with the highest beta. By the end of the period, $1 in the low-beta portfolio has grown to $190 while the high-beta portfolio grew to just $18.

This is a difference of more than 5% per year.  

In addition, the low-beta portfolio was less volatile with a maximum drawdown of 35% compared to the high-beta portfolios maximum drawdown of 75%.

Higher risk = higher returns…not so! At least not with stocks

Research provides additional support – but with caveats.

A study by iQuant.pro analyzes various forms of Beta (and over various time frames) as they pertain to the large-cap S&P 500 Index from Jan, 1974 through January 2018.  

We applied the following forms of Beta to the S&P 500 Index (on an equal-weight basis):

  1. 2-Year Beta

  2. 3-Year Beta

  3. 4-Year Beta

  4. 5-Year Beta

  5. 2-Year Up Beta (Beta in Up Markets)

  6. 3-Year Up Beta

  7. 4-Year Up Beta

  8. 5-Year Up Beta

  9. 2-Year Down Beta (Beta in Down Markets)

  10. 3-Year Down Beta

  11. 4-Year Down Beta

  12. 5-Year Down Beta
     

The results were interesting, to say the least.

The following table shows the performance and Sharpe Ratio per quintile of each form of Beta noted above as applied to the S&P 500 Index since 1974:

As you can see, the outperformance of low-beta stocks is dependent on the timeframe over which Beta is calculated.

As you can see, the outperformance of low-beta stocks is dependent on the timeframe over which Beta is calculated.

While stocks with the lowest Beta displayed less risk and higher Sharpe Ratios across the Board, the same cannot necessarily be said for historical returns. The largest difference in returns came from the 5-year Up Market Beta – something that is intuitively hard to understand.  Since 1974, stocks with the lowest 5-year upmarket beta outperformed their high-beta counterparts by 3% per year with a much higher Sharpe Ratio (0.92 versus 0.47).  The largest difference in efficiency (measured by Sharpe Ratio) came from the 5-year total Beta ratio.  Interestingly, stocks with the highest 2-year down market beta outperformed stocks with the lowest 2-year down market beta.
 

So what does this all mean?

These results show that models, such as iQuant’s newest Smart Beta Model (which focuses on stocks with the lowest 5-year Beta for reasons noted above), have the potential to unlock the disconnect between high performance and volatility. Stock portfolios built with low beta stocks tend to outperform high-beta growth stocks, which equals a nice nights sleep without worrying about big swings in the market.

Ready to fall asleep knowing that your stocks are in good hands? Test drive any of our 14 models, including the new Smart Betal Model, as part of the iQuant community.