Profiting From Bad Employee Relations
Robbert is a member of The Motley Fool Blog Network -- entries represent the personal opinion of the blogger and are not formally edited.
A few months ago I conducted a research on the relationship between employee relations (that’s not office romances) and stock returns. As it turned out, there were some interesting relationships between the two.
Some basic info on our research
We used the KLD database to determine how well each of the Russell 3000 companies took care of their employees. There were several determinants of good relations such as employee involvement in decision making; a bad characteristic could be workforce reduction for example.
Some examples of companies that scored positive on employee relations are Intel (NASDAQ: INTC) which had the all time top score, Southwest Airlines (NYSE: LUV), and Tellabs. When looking at Intel, it seems odd that is was #1 in KLD while it's #51 on the fortune list of 100 best companies to work for. Moreover Google (NASDAQ: GOOG) (#3 on the Fortune list) is well known for its good employee treatment. As it turns out, Google is only acknowledged for its working environment and employee involvement in the KLD data. Intel has the measurable advantages. This shows the limits of the KLD-database, although we could argue Google doesn't deserve to be acknowledged for what it does not do.
Southwest Airlines on the other hand is known for its employee treatment and actively tries to bring this image across, and as it turns out there's truth in it. Most airlines had a negative score in the database, which makes Southwest somewhat special.
Wal-Mart (NYSE: WMT) and Supervalu (NYSE: SVU), were among the bad performers on employee treatment. Especially Wal-Mart is known for its bad employee treatment. When you enter 'employee treatment' in Google, the second hit is a Wikipedia page named 'Criticism of Wal-Mart' where you can read about accusations of forced working overtime without pay and more. These Wal-Mart practices are captured by health & safety and other concerns in the KLD database. Supervalu on the other hand is a less obvious villain as pay is pretty good, and this is not fully captured in the database. Generally the consumer staples industry had the worst employee relations, while financial institutions and tech companies had the best employee relations. Below is a table to illustrate how the data looked like for the 5 companies mentioned.
| 2009 | workforce reduction | Retirement benefits | Health&safety | Union relations | Other | Employee involvement | Cash profit sharing | Score |
| INTC | +1 | +1 | +1 | +1 | +1 | 5 | ||
| LUV | +1 | +1 | +1 | +1 | 3 | |||
| TLAB | +1 | +1 | +1 | +1 | 4 | |||
| WMT | -1 | -1 | -1 | -3 | ||||
| SVU | -1 | -1 | -1 | -1 | -4 |
You should know that many companies had retirement benefit concerns or strengths throughout the KLD database, 34% of tickers in 2009 while only 6% had a score on union relations. This is why we only used companies with scores greater than +1 or lower than -1, because we didn't want to measure the sole effect of having a good or bad pension plan. The KLD-database might not seem very perfect, but there's no other database with so much employee relations data for publicly traded companies.
The time frame of the KLD statistics is 2003-2009, the corresponding returns are 2004-2010 to prove causality. Our hypothesis was that bad employee relations companies outperformed companies with good employee relations. So we put stocks with bad employee relations in a long portfolio and stocks with good employee relations in a short portfolio (equal size of course) and updated the portfolios each year. We also removed industry effects (by putting positive and negative scores in equally sized long-short portfolios) as we didn’t want to research the difference between different industries, but between employee relations. Finally we calculated (by regression in STATA) the risk adjusted returns using 3 models: CAPM, the three factor model (FF) and the four factor model (FFC).
The results
One thing which shouldn’t surprise us was that there was little to no causal relationship to be found between risk-adjusted stock returns and employee relations (see table below for data). So as an investor, you can as well ignore it. A very small risk-adjusted return could be made by going long good relations companies while shorting companies with bad employee relations. When only adjusting for stock volatility (CAPM), you’ll have a nice positive annual return of 0.39% (excl. transaction costs!) on the long-short portfolio. When removing all risk adjusting models and sector weighting there is a 23% return over the 7 years, that’s an average of 3% per year, with only 2 years of loss (2007 and 2005). This sudden performance after removing risk adjusting models means that companies with bad employee relations are generally more risky (and more profitable!) investments. We also found some fundamental support for this: companies with bad employee relations were generally in a worse financial position and had more debt than other businesses.
Returns Long-Short portfolio
Jan. 2004 - Dec. 2010
| Annualized alpha (p-value) |
Unweighted Annualized alpha |
|
| CAPM | 0.393% (0.797) |
0.234% (0.932) |
| FF | -0.157% (0.905) |
-1.144% (0.663) |
| FFC | -0.180% (0.890) |
-1.066% (0.645) |
The right column contains the results with industry effects. P-values indicate significance (the lower the better).
Surprise 1
Both the long and the short portfolio had a positive excess return (alpha) of more than 2%. Our best guess is that this had to do with the small average size of those companies in the KLD database. Still this outperformance shouldn't make sense because the FF and FFC models adjust returns for company size. The excess return could also be explained by the fact that we didn't allocate proportions in our portfolio's according to market capitalizations; all companies had an equal weight within their industry. So perhaps there is excess return to be made by going long companies in the KLD-database while shorting the S&P500.
Surprise 2
Businesses with bad employee relations which left the stock market generally did so right after a slide in stock price while companies with good employee relations which ended their listing mostly did so after a steep rise in its stock price. So if you can predict which of the good and/or bad employee relations companies would probably leave the stock market in a given year, your outperformance will be very large.
The bottom line is the returns don’t depend that much on employee relations, as it turns out, the efficient market theory really works for employee relations. It is common sense that employees work harder, but as we assumed at the start of our research, investors know this and value the stock accordingly. If you don’t care about risk though, it might be a good idea to try our long-short portfolio strategy.
Robbert Manders does not own any of the companies mentioned in this entry. The Motley Fool owns shares of Google, Intel, and SUPERVALU INC. Motley Fool newsletter services recommend Google, Intel, and Southwest Airlines. Try any of our Foolish newsletter services free for 30 days. We Fools may not all hold the same opinions, but we all believe that considering a diverse range of insights makes us better investors. The Motley Fool has a disclosure policy.If you have questions about this post or the Fool’s blog network, click here for information.
