Key Points

  • Momentum, breadth, sentiment, earnings, valuation and macro conditions currently support a bias within the U.S. equity market toward large caps over small caps
  • Small caps' 18-year outperformance phase may be ending
  • Large caps' defensive nature could serve as ballast to portfolios during pullbacks

Many investors are wondering whether the stock market has come too far too fast. The latest consolidation brought the S&P 500 down only 2%, but the average stock was down more than that. Although we remain steadfast believers in this eight-year secular bull market, we have been reminding investors about the power of disciplined rebalancing around and diversification within strategic allocation. For those investors who like to take a more tactical approach, we recently changed our bias within U.S. equities toward large caps. This report details why, in a chart-heavy, word-light format.

The tides have recently shifted and now it's a confluence of factors—including momentum, breadth, sentiment, earnings, valuation and macro conditions—that we believe support a bias toward large caps over small caps. The slightly greater defensiveness of large caps could serve as ballast for portfolios if the so-far mild equity market pullback has legs.

Momentum/trend

Coming out of the early-2016 correction, small caps outperformed large caps initially; with a second burst of outperformance in the immediate aftermath of the U.S. presidential election. Since the second week in December, however, the Russell 2000/Russell 1000 (small caps/large caps) ratio has given back about half those relative gains.

I keep tabs on the trend work from Ned Davis Research (NDR), and as you will see in the charts below, all three time periods have the models favoring large caps over small caps- with the longest-term trend model just recently confirming the short- and intermediate-term models. As you can see in the accompanying table for each, these signals have historically led to outperformance by large caps over small caps.

Russell 1000/2000 short-term model

Russell 1000/2000 short-term model table

Russell 1000/2000 intermediate-term model