Goldman Sachs Doesn't Understand Stocks

The stock market posted a small loss on Wednesday, with the Dow Jones Industrial Average (^DJI 0.07%) and the S&P 500 (^GSPC -0.09%) down 0.31% and 0.20%, respectively.

Goldman Sachs' influential Portfolio Strategy group published a new report last Friday with its key takeaway at the top of the first page:

Zero rates support S&P 500 rising to 2100 by year-end, but stock-picking to be challenging.

Nonsense.

Here's how Goldman justified its warning (my emphasis):

Although financial conditions eased, market uncertainty is likely to remain high as investors refocus on the timing of a "first" Fed hike, China's economic slowdown, and oil price volatility. The uncertain market environment ensures that macro forces will stay top of mind for fund managers. Unfortunately, elevated macro risks mean a challenging stockpicking environment as macro forces outweigh stock fundamentals.

They have things topsy-turvy: All other things equal, it's precisely when "macro forces outweigh stock fundamentals" that the stock-picking environment is likely to be most favorable.

That's if you accept Benjamin Graham's quote that "though the stock market functions as a voting machine in the short run, it acts as a weighing machine in the long run." (Ben Graham was Warren Buffett's mentor and the father of security analysis.)

In other words, long-term stock returns reflect a company's fundamentals rather than market sentiment or any other extraneous factor, including the timing of the first Fed hike, etc.

As such, old-fashioned stockpickers welcome an environment that distracts the market from focusing on stock fundamentals, causing stock prices to stray from stocks' intrinsic value.

Undaunted by this logic, Goldman took out a quantitative bludgeon:

During the last month, stock correlations skyrocketed as indiscriminate macro pressures pummeled the S&P 500 index and weighed on the stock-picking environment. [...] As correlations spiked, return dispersion -- measured as the cross-sectional standard deviation of stock returns -- fell, leaving bottom-up stock pickers with a diminished opportunity set for generating alpha.

Simply put, as stock prices move in closer unison (higher correlations), the differences across individual stocks' returns narrow (lower return dispersion). That's perfectly true, and it explains why stock pickers' trailing short-term returns have become less differentiated. However, it doesn't tell us anything about stock pickers' prospective long-term returns and their ability to beat the market.

To illustrate this, the following table shows U.S. equity hedge funds' 3- and 5-year performance versus the S&P 500 following quarters that were notable for:

• Decreasing returns dispersion
• Returns dispersion significantly below the historical average
• Poor recent hedge fund performance (Goldman looks at the past 12 months)

Hedge funds' cumulative returns versus the S&P 500*

Source: Bloomberg. Hedge funds' performance is based on the HFRI Equity Hedge (Total) Index. *On a total return basis. 

Yes, hedge funds' underperformance over the past five years has been breathtaking (last row), but looking at all instances, it clearly shows that low returns dispersion doesn't predict poor long-term results. (I'd be very curious to see the same results for fund managers with the lowest portfolio turnover, as I don't think equity hedge funds, as a group, are the best representation of "stock pickers.")

Earlier this month, Warren Buffett was asked whether he saw the recent pullback as a reason to buy. He responded:

We're buying because we like what we're buying in relation to its long-term prospects. If we like it at X a few weeks ago and if it's 95% of X now, I like it even better.

There was no mention of returns dispersion.