Big Tech, narrow market

Good morning. US job openings edged up in April. Despite the fact that no one particularly trusts this data series, it was widely discussed yesterday as a further signal that the labour market remains tight and there might be another rate increase in our future. Booooo. All eyes are now on Friday’s jobs report. Well, not quite all eyes. Mine are on tech stocks. Email me: robert.armstrong@ft.com.

Tech, alone

Here is a chart of the relative performance of the six superstar tech stocks (Apple, Amazon, Google, Meta, Microsoft, Nvidia) against the S&P 500 with those six names removed:

The recent staggering outperformance of the mega-techs looks a lot like the run-up they enjoyed during the early pandemic days of 2020. But things are different this time. For most of the last mega-tech rally, the rest of the index was rising, too — just not nearly as fast. This time, the mega-techs are carrying the index on their back. Here is the performance of the S&P 500 less the super six:

Narrow markets are widely believed to be a bad omen. We have argued in the past that narrowness is a noisy signal that, on its own, does not concern us too much. But — as with most stock market indicators — there are as many views as there are ways to cut the data.

Our view that narrowness was not too big a deal was based on a measurement of breadth that looked at the number of stocks making new highs versus those making new lows. Adam Turnquist of LPL financial, by contrast, defines breadth by the proportion of S&P 500 stocks above their 200-day moving average. He finds a strong correlation between narrow markets and weak index performance over the following year. He divides the market into quartiles by breadth, using data reaching back to 1991. We are now in the fifth, or narrowest, quartile of markets:

Don’t panic yet. Brian Belski of BMO looks at past periods when the top five stocks in the S&P 500 have reached peak relative performance (one standard deviation above normal), and found subsequent returns were just fine over the next six and 12 months. He also looked at historical periods, going back to 1990, in which the number of stocks outperforming the index dropped sharply, and found that during the subsequent 12 months, S&P 500 returns were about average.

Given that different ways of defining breadth lead to different results, one cannot help suspect there is an element here of torturing the data until it confesses. Unhedged continues to suspect that breadth on its own is not a reliable indicator. It needs to be accompanied by an account of why the market is narrow, and that account also needs to help explain why there should be a future market decline.

The 2020 Big Tech rally corresponded with rapidly falling interest rates. The outperformance was therefore widely attributed to “long duration” — because much of growth stocks’ cash flows are far in the future, a lower discount rate increases their value disproportionately, it was said. We never much liked this explanation, but in any case, it doesn’t hold now. In 2023 Big Tech has leapt against a background of roughly sideways movement in long bond yields. So what is driving the super six now? There are two explanations, one reassuring and one troubling. 

The first explanation, which Unhedged has leaned on in the past, is that Big Tech stocks’ strong free cash flow, high barriers to entry, and central place in the modern economy make them sensible things to own when the world is starting to wobble. Despite moments of doubt, we still believe this (though the run up in the super six’s valuation makes us believe it a little less; Apple’s price/earnings ratio, for example, has gone from 20 to 29 this year).

The second explanation is that the super six are rising on a narrative. The narrative, roughly speaking, is that AI is going to be a bigger deal than the internet, the PC, the printing press, the wheel and fire combined, and that the mega-techs are best positioned to take advantage of this. I don’t know anything about AI, but I have some experience with market narratives. They are not tightly linked to the facts, and they don’t last for ever.

Corporate boards and stock performance

There is a large amount of literature on the influence of management skill on stock performance. Most of it, in my experience, is unsatisfying. Because executive excellence is hard to measure directly, most studies treat it as a residual. In other words, for a given company over a given period, you take out every measurable factor that could explain stock performance (market returns, industry performance, and so on) and attribute whatever is left to good leadership.

The influence of a good board of directors can be measured in a slightly different way, however. Because clever directors are in demand, good ones tend to serve on multiple boards or as executives at other companies. One can compare companies whose board members have lots of connections to other firms to companies with less well-connected boards, and see if the former perform better. Several studies (see for example here and here) have found that they do.

Yin Luo of Wolfe Research has had another look at this idea in a recent report, and his results are interesting. He looks at companies where the board has strong connections to firms that are highly successful, as measured by a range of financial measures, but where the company itself ranks poorly on those measures. The thesis is that the board should be able to use its connections and experience to improve lagging performance.

Looking across a large universe of companies, Luo finds that the thesis holds. The presence of board connections with strong companies predicts improving performance in area such as profit growth, margins, and valuation.

One particularly interesting result: companies with strong board connections to companies with, for example, high returns on equity — but which did not have high ROE themselves — had notably lower downside risk than those with high ROEs. The average maximum stock drawdown for the well-connected firms is in green below; that of high ROE firms in red. Luo breaks the results out by period:

This is what a good board is supposed to do, after all: keep a company from falling into really bad trouble.

I wondered, on reading the study, about the direction of causality. Do good directors improve corporate performance — or seek out promising companies to serve? I put the question to Luo, and he thinks it’s probably some of both. I myself lean to the latter view, on the basis that it is easier to recognise potential than it is to fix companies without it. That makes a connected board a signal very much like insider stock purchases. For investors, of course, the direction of causality doesn’t matter: the point is that connected directors might be a useful buy signal, especially at underperforming companies.

One good read

Economic liberals and Brexit.