Turbulence beneath a calm surface

If you looked just at share market indices like the S&P 500 or the MSCI World index, you might think that share markets are currently enjoying a period of relative stability. This is shown in the graph below, measuring the daily volatility of both the S&P 500 index and the MSCI World index over rolling 65-weekday periods (approximately 3 months).

The graph above shows that the recent volatility of the share market indices has been slightly below the historical average (the long run average has been 0.85% for the MSCI World index and 1.04% for the S&P 5oo index). However, despite the stability of the indices, we have recently been seeing a lot of extreme price swings in individual share prices.

To evaluate how unusual this is, in the graph below we have looked at the 3-month daily volatility of the average constituent stock of the S&P 500 index at this time each year for each of last 30 years and compared this to the volatility of the index as a whole.

The most recent observation (the 3 months up to last Friday) stands out as being unusual in that we’re seeing significantly greater-than-average share price volatility whilst also seeing lower-than-average index volatility.

Historically, the typical relationship between index volatility and individual stock price volatility has been that the average member of the S&P 500 index has been about twice as volatile as the index itself, implying that general market effects explain about 25% of the movement in a typical company’s share price, while company-specific or sector-specific effects explain the remaining 75%.

But right now, the average member of the S&P 500 index is showing 2.9 times as much volatility as the index itself, which implies that general market effects are only explaining about 12% of the variability in its share price.

This also shows up in a high dispersion between the returns of different stocks. The graph below shows two different ways of measuring the dispersion between the returns of different stocks (in the S&P 500 index). The black line takes a 50-day average of S&P Global’s index that measures the one-day dispersion of returns amongst S&P 500 members (“dispersion” is the cross-sectional standard deviation of returns amongst index members on any given day). The green line is forward-looking, using a CBOE index that uses the pricing of stock options and stock index options to work out the level of future share price dispersion that is priced into the options market. Both indices show that dispersion is the highest it’s been over the period that these indices have been published except for a brief period during the outbreak of covid, when dispersion briefly spiked even higher. 

The high dispersion during a period of modest index volatility implies that the average correlation between any randomly selected pair of stocks is a lot lower than it has typically been in the past.

We have recently been seeing this frequently on days when the market rises or falls significantly. In historically normal times, if the market was down by (say) 2%, you’d expect that most stocks would decline, but that the more market-sensitive stocks would typically fall by more than 2% while the more defensive stocks might fall by a lesser amount (say, -1%). However, in the past few months, the pattern has been more extreme: The market has often seemed to pivot on days when the market moves significantly, such that many of the most defensive stocks will actually rise when the market declines (and fall when the market rises). For example, so far this year Verizon has typically moved 50% in the opposite direction of whatever the S&P 500 index has returned, as shown in the graph below.

 The fact that Verizon (along with many other defensive stocks) is typically moving in the opposite direction to the S&P 500 index indicates that it negatively correlated to most other stocks. Examples like this reduce the average correlation between stocks.

The high volatility of individual stocks combined with low correlation between many stocks implies that we are in a market environment where diversification will provide greater-than-usual benefits. Conversely, it also indicates that we’re in a market environment where we see might some extremely bad or extremely good results from funds and portfolios that are more heavily concentrated.

Another aspect of the current market is the extremely high volatility shown by a few individual stocks. While the average volatility of stocks in the S&P 500 index is 37.9%, 25% of stocks are showing annualised volatility of over 44.5%. We see this in particular in the next tier of tech stocks (outside of the magnificent 7), which I discussed in my last article. As I showed in that article, the median (annualised) volatility for that next tier of tech stocks is currently over 50% (corresponding to typical daily price movements of +/- 3.1%). As I pointed out in that column, if we assume that markets are efficient, this level of volatility in a stock essentially implies that it is a “hero or zero crapshoot” with a high probability that the stock could end up worthless over a period of a decade or more, but a smaller chance of extremely good returns.

Of course, the other interpretation is that the high volatility may show that markets aren’t being efficient, and that prices are either being buffeted by random “noise” (e.g. imbalances created by ETF buying and selling) or are over-reacting to scraps of real information that don’t deserve such an extreme reaction. In either case, the obvious strategy is to lean against the market, by buying the stocks that are weakest on the day and selling the stocks that are strongest. The graph below shows how well this would have worked during a previous period of volatility – the later stages of the Global Financial Crisis and its aftermath. 

This graph overstates the potential returns that contrarian traders could have earned over the GFC, as it shows returns prior to transaction costs, and a strategy of buying the weakest stocks each day would involve extremely high levels of turnover, which would rack up significant transactions costs even in a low-transaction cost market such as the United States.

Nonetheless, it indicates that when you’re in a period of high stock volatility it is important to be price-sensitive when executing trades, to wait for good opportunities when you are thinking about adding to or reducing an equity investment, and to avoid getting panicked by adverse price movements.

Is this unusual market environment a warning sign?

Historically, significant share market corrections have often been preceded by periods of unusual share price behaviours associated with speculative or programmed trading. Although a glance at share market indices might give you the impression that market behaviour is quite normal, the high level of individual stock volatility indicates some dangerous currents beneath the “still waters” of stable share market indices.

The graph below (using the same underlying data as the second graph in this column) shows that the disconnect between individual stock volatility and the volatility of the S&P 500 is greater than it has been at any other time (in the last 30 years) other than during the peak of the previous tech-related capex boom in the year 2000.

High stock volatility can easily transition to high index volatility, as correlations don’t tend to be as stable over time as volatilities. If stocks remain volatile, but they start to move more in unison, this would lead to greater volatility in the share market indices, similar to what we saw in the year 2000.

The current environment echoes the year 2000 in many other (more fundamental) ways as well, including high valuations, and cyclically high profitability being driven by a technology sector boom in capital spending.

Share price volatility is currently highest amongst the companies that investors anticipate will benefit from capital spending on AI-related data centres. While media commentators like to repeat the trope about how the people selling picks and shovels do the best out of a gold rush, recent history tells a different story. For example, many of the companies selling fibre optical cable and routing equipment into the 1999/2000 internet/telecommunications spending boom lost most of their value over the following 3 years and (as far as I can see) none of the large oil drilling and oil field services companies that were doing so well during the oil drilling boom of 2008-2012 are worth as much today as they were in the early days of that boom.

Capex booms are invariably driven in part by good economics based on current market pricing. However, a currently-favourable equation for near-term investment returns does not mean that the capital spending will persist for long. Spending large sums on telecommunications and routing equipment seemed to make a lot of sense in 2000 when people were paying $3 per hour to access the internet at 56.6 kbps. Similarly, spending large sums on oil drilling seemed to make a lot of sense in 2008 when oil was selling at US$120 per barrel. Similarly, today it is possible to model very strong returns from investing in data centres and GPU server racks if we assume that top tier AI models will continue to sell access to their AI models at US$2 per million tokens.

However, the market pricing that stimulates investment never seems to survive the investment that it stimulates. Regardless of how confidently some pundits might talk about the future of many companies supplying the AI capex boom, the high share price volatility of these companies indicates how difficult it is to even guess at what profits they may be able to sustain in the future.

 

Nicholas Bagnall is Chief Investment Officer at Te Ahumairangi Investment Management.

Disclaimer: This article is for informational purposes only and is not, nor should be construed as, investment advice for any person. The writer is a director and shareholder of Te Ahumairangi Investment Management Limited and an investor in Te Ahumairangi Global Equity Fund. Te Ahumairangi manages client portfolios (including Te Ahumairangi Global Equity Fund) that invest in global equity markets, and hold shares in Verizon, which was mentioned in this column.