Is everything really speeding up? (turbulence is measured, not proclaimed)

The slide that opens every seminar

You have seen it. The average tenure of a company in the S&P 500 was 33 years in 1964; it fell to 24 by 2016; and the projection takes it to 12 years by 2027 —with half the index replaced within a decade. It comes from the consultancy Innosight —the classic version is the 2018 edition, and it is recalculated every so often— and it is probably the most quoted management statistic of the century. It usually travels with its BCG cousin: across 35,000 US-listed companies since 1950, the risk that a public company drops off the market within the next five years went from 5% in the 1960s to 32% in the 2015 measurement —almost one in ten public companies leaves the board every year, four times as many as in 1965.

The figures are real. The conclusion drawn from them —"competition has turned ruthless, everything is accelerating, your company is next"— is, to put it mildly, hasty. Because when the academics went looking for that acceleration with a magnifying glass, they found something far more interesting than the slide.

This three-part series is about competing when the environment moves. But before buying tools for turbulence it is worth knowing how much turbulence there actually is —yours, not the conference circuit's. That is what this first part is for. The second looks at the central tension of strategy in motion: how much to commit and how much to keep open. The third takes up the evidence on which form of organization holds up best in the storm —with a result that will make more than one crisis committee uncomfortable.

Six thermometers, three answers

The trouble with "has the world become more turbulent?" is that it is not one question: it is six, and they do not give the same answer.

Thermometer 1: index turnover (rising). This is Innosight's. It measures how long a company stays in the S&P 500. Its catch: a company leaves the index when it goes bankrupt, yes —but also when it is acquired, when it merges, when it loses relative market capitalization, or when the index committee swaps it for a more representative name. In years of cheap M&A the index turns over faster without anyone having been "disrupted". It is a thermometer of financial reshuffling, not of competitive death.

Thermometer 2: mortality among listed companies (rising, with the same catch). The 32% exit risk BCG measured. A team at the Santa Fe Institute (Daepp and colleagues, 2015) applied a biologist's mathematics to it: analysing tens of thousands of US-listed companies between 1950 and 2009, they found that the "half-life" of a public company —the time it takes for half of a cohort to disappear— is around ten years, and that the risk of dying is strikingly independent of age: a 50-year-old company has, against all intuition, an annual risk close to that of a 5-year-old one. But the study itself underlines the fine print: "dying" includes being acquired. In this data, half the deaths are weddings.

Thermometer 3: the persistence of advantage (this one really is shortening). This is the one that matters strategically, and its result is uncomfortable in a different way. Robert Wiggins and Timothy Ruefli tracked 6,772 companies across 40 industries for 25 years with a precise question: how many achieve sustained superior economic performance, and for how long? First answer (2002): very few, and almost never for long stretches —sustained superior performance is the statistical exception, not the norm. Second answer (2005, in a paper with the best title in the field: Schumpeter's Ghost): periods of superior performance are getting shorter over time, the phenomenon is not confined to technology —it shows up across a wide range of industries— and the companies that stay on top do so less and less through one durable advantage and more and more by stringing temporary advantages together, one after another. Rita McGrath would later turn that observation into a thesis: competitive advantage as a wave you catch, ride and abandon in time.

Thermometer 4: the markers of hypercompetition (not rising). Here it gets interesting. In 1994 Richard D'Aveni popularized the word "hypercompetition": the era of sustainable advantage was over. The claim became common sense. In 2003, Gerry McNamara, Paul Vaaler and Cynthia Devers put it in the dock with 114,191 observations of business-unit results —some 5,700 a year, across 908 industries— between 1978 and 1997, testing four markers: is advantage eroding faster? is business mortality rising? are markets more volatile? is growth drying up? Their title says it all: Same as It Ever Was. They found no systematic trend towards greater hypercompetition. The instability of results rose in the mid-1980s… and then fell, returning to late-1970s levels. Turbulence showed up in episodes and sectors, not as a universal tide.

Thermometer 5: aggregate dynamism (falling!). The killer blow to the universal-acceleration story came not from strategy but from labour economics. Decker, Haltiwanger, Jarmin and Miranda documented that the US economy has become less dynamic over recent decades: the rate at which new firms are created has been falling since the 1980s, and employment in young firms dropped from 20% to 10%. Less entry, less exit, less reallocation: in the aggregate, American capitalism was moving more slowly, not faster, precisely while the seminars were announcing the opposite.

Thermometer 6: uncertainty (rising since 2012 —and more so for us). What did rise, measured seriously, is uncertainty: the difficulty of knowing what comes next. The World Uncertainty Index (Ahir, Bloom and Furceri), built for 143 countries since the 1950s, shows spikes at every crisis —9/11, 2008, the eurozone, Brexit, the trade war— a rising trend since 2012, and its all-time high in the second quarter of 2020. With two findings that will not surprise a reader in this part of the world: uncertainty is systematically higher in emerging economies than in advanced ones, and its shocks have real effects —on the order of 3% of the variance of growth over two years. Turbulence and uncertainty are not the same thing: an environment can change a great deal in predictable ways, or change little and turn illegible. It is the second that has grown most.

What the data show once you read them together

Lined up side by side, the six thermometers tell a more precise story than the slide does:

First: universal acceleration is an aggregation myth. There is no evidence that everything has become faster for everyone. There are sectors and periods of extreme turbulence (electronics in the 1980s, media in the 2000s, retail in the 2010s) sitting alongside sectors where the rules of 1990 remain intact. The flaw in the Innosight slide is not its numbers: it is selling the turbulence of the 90th percentile as if it were everybody's weather. And measures built on market exits inflate the drama, because they count every acquisition as a death.

Second: what has genuinely shortened is the life of each individual advantage. Even if the aggregate climate has not warmed uniformly, the Wiggins and Ruefli evidence is hard to dismiss: sustaining superior performance on a single advantage is rarer and rarer, in more and more industries. The strategic consequence is not "run faster": it is changing the unit of planning —from the advantage (singular, defensible, eternal) to the portfolio of advantages (plural, temporary, chainable). How that portfolio is managed is what Part 2 is about.

Third: for a reader in this region, the news is a different one. The uncertainty that global indices record as a recent phenomenon —policy regimes that shift, rules that get rewritten, prices that carry no information— is the historical habitat of any Argentine company. That has a bitter reading and an advantageous one. The bitter: the strategy textbooks were written for environments that look very little like ours. The advantageous: the capabilities this series describes —reading the regime, staging commitments, deciding in a decentralized way— are not theory here: they are the price of admission, and local organizations that survived decades of macro turbulence have, without knowing it, muscles their peers in stable economies are only now discovering they need.

Measuring your own turbulence (the only weather that matters)

If turbulence is local, the operational question is not "is the world accelerating?" but "how fast does my business turn over —and is it changing regime?". That can be measured, with indicators within reach of any management team:

Product turnover. The measure economists use to capture turbulence at firm level is brutally simple: how many products were added and discontinued in the period, over the total catalogue (that is how Aghion, Bloom, Sadun and co-authors build it in the study that headlines Part 3). Calculate it for your company and for your category: if half of what you sell today did not exist five years ago, you operate in high turbulence, whatever the economy-wide average says.

Competitor turnover. Who were your five relevant competitors ten years ago, and who are they today? A stable list describes one regime; an unrecognizable one, another. Run the same exercise with customers and channels.

The useful life of your advantages. The home-made version of Wiggins and Ruefli: take your last three real commercial advantages —a winning product, an unbeatable cost, an exclusive arrangement— and measure how long each took to erode. That duration, not the S&P 500's, is your strategic clock: it sets how much time you have to pay back the next bet.

Signals of a regime change. The dangerous turbulence is not the chronic kind —you have already adapted to that— but the change of regime: the moment a stable business turns turbulent, or the reverse. The early indicators are rarely in the income statement, which arrives late by design; they are in the churn of small customers, in the pilots the entrants are running, in distributor margins, in what your competitors are recruiting for. Part 2 of the previous series put it another way: the expensive mistake is managing one territory with another territory's method. The strategist's first job is to notice that the territory changed —while noticing is still cheap.

There is a serious academic objection to this whole programme, and it is worth facing head-on: Lawrence Hrebiniak and William Joyce showed forty years ago that adaptation is not a pure choice —the environment genuinely constrains (determinism) and management genuinely chooses (strategic choice), at the same time, in proportions that vary by industry. Measuring your own regime is exactly that: knowing how much room to choose the environment has left you, so you can spend it where it pays.

The thermometer, what it really measures, and its catch.

Questions for Monday

For your core business, with data rather than impressions:

That leaves the question the Innosight slide dodges and Wiggins and Ruefli serve up: if each individual advantage lasts less and less, is the answer to commit less —keep everything light, everything reversible, everything optional? Modern instinct says yes. The evidence says something finer: total flexibility is as lethal as total rigidity, because a company that commits to nothing builds nothing. Part 2 goes there: what to commit, what to keep open, and the tools —real options, discovery-driven planning— for not choosing blind.