Are Small Caps Expensive?

The recent underperformance of small cap stocks has me wondering whether the outperformance that small cap stocks has enjoyed over the last decade could be coming to a close.

Over the last ten years, the S&P 500 has gained 7.16 percent and the S&P 600 Small Cap Index has earned 9.81 percent a year, an outperformance of 2.65 percent per year.

There is nothing particularly unusual about small cap stocks outperforming large cap stock since the academic data (CRSP) shows that the annualized outperformance of small versus large stocks has historically been 2.18 percent per year.  The last ten years has been better than average, but not outrageously so.

Still, some market observers are saying that small cap stocks are expensive, so I decided to take a look at the data and see what it says.

The first chart shows the price-to-book ratio of the S&P 500 large cap index and the S&P 600 small cap index since 1995, the earliest point that I could find data.  The P/B ratio compares the market value of a stock with the accounting value of a stock, known as the equity or book value.

The P/B is computed by subtracting the company liabilities from the assets, much like determining the equity in your home (the market value minus your mortgage).

2014-04-16 PB Ratios

What we see is that large-cap stocks over this period have always been more expensive than small cap stocks, especially in the late 1990s.  At the height of the tech bubble in tech stocks, the P/B for large cap stocks was 4.88 and 2.40 for small cap stocks.

You could say that small cap traded at a 50 percent discount to large cap (or that large cap traded at a premium).  The following chart shows a history of the discount for the study period.

2014-04-18 Small to Large

From here, I wanted to see if the relative discount said anything about the subsequent returns of each asset class.  To test this, I did a regression that compares the initial discount pictured above with the subsequent five year returns for the two indexes to see if one outperformed the other.

In short, I found that the relative discount in valuation between the two indexes was a good signal for subsequent outperformance (for this period anyway).  So, small caps were cheap relative to large cap stocks, they tended to outperform in the next five years.  If small cap stocks were expensive relative to large cap stocks, they tended to underperform.

The following scatterplot shows the relationship in more detail.  The x-axis shows the relative value between small cap and large cap (0.50 is cheap and 0.90 is expensive) and the y-axis shows the annualized over or under-performance of small cap versus large cap.

2014-04-18 Scatterplot

Generally speaking, you can see that when small cap stocks are cheap relative to large cap stocks, small cap stocks tend to outperform (the upper left data points on the chart).  When small cap stocks are more expensive relative to large cap, they tend to underperform (the small right data points on the chart).  The r-square of 0.61 shows that the signal is relatively good – the starting relative value explains 60 percent of the subsequent over or underperformance.

Today, the price-to-book ratio for small cap stocks is 2.25 and the price-to-book ratio for large cap stocks is 2.61, or a ratio of 0.86 that signals that the discount is small right now.

Does that mean we should exit small cap stocks?  As you might expect, I don’t think so.  First, our data set isn’t long enough and we don’t have access to longer-term data.  We would also have to test other indexes beyond S&P and other subsequent periods like seven or 10 years.  Since I’ve never seen this kind of data, we may do that, but finding older and reliable data is more important.

Second, just because the discount is small right now, doesn’t mean that small caps will underperform.  We can see from the chart that there are a number of periods where the discount is at 0.80 or higher and small cap still outperforms large cap.

Many of the periods where the discount is high and the subsequent returns are low are from the tech bubble, so we can’t say whether this is a true phenomenon or specific to his time frame (hence the need for longer-term data).

At the same time, we won’t ignore this data either.  Some clients have expressed interest in increasing their small cap exposure, probably because small caps have outperformed large cap do dramatically in the last 12 months.  It seems unlikely to me that we would increase our exposure to small cap right now.

Regardless, I can say that we will investigate this further and not just between large and small cap stocks.