The Lindy Effect: Why Age Is the
Best Predictor of Future Survival
A deep-dive into one of the most powerful yet underappreciated ideas in economics and risk theory; how the past survival of non-perishable things predicts their future longevity better than any model.
Imagine a technology that has existed for 2000 years. According to the Lindy Effect you should expect it to survive for at least another 2000. Now imagine a startup that launched six months ago. Its expected lifespan is roughly six months. This deceptively simple idea has profound consequences for investors, policymakers and anyone trying to predict what endures in a complex world.
The Lindy Effect sits at the heart of modern risk theory yet it rarely appears in mainstream economics textbooks. First observed informally and later formalised by mathematician and former derivatives trader Nassim Nicholas Taleb, the principle challenges the standard economic assumption that age weakens systems. For certain classes of things precisely the opposite is true.
Origins and History
The name traces back to Lindy's Deli on Broadway in New York City. In the 1960s comedians and entertainers would gather there and informally theorise that the future career expectancy of a television comedian was proportional to the total amount of time already spent on television. A performer with ten years on screen could be expected to last another ten. This was never a rigorous theorem but simply a recurring observation repeated over deli sandwiches.
For the perishable every additional day in its life translates into a shorter additional life expectancy. For the non-perishable every additional day may imply a longer life expectancy.
— Nassim Nicholas Taleb, Antifragile: Things That Gain from Disorder (2012)The philosopher Benoit Mandelbrot touched on related ideas through fractal geometry and fat-tailed distributions throughout the 1960s to 1980s. Albert Goldman coined the term "Lindy's Law" in a 1964 essay in The New Republic. Taleb then gave it rigorous mathematical grounding in both The Black Swan (2007) and especially Antifragile (2012), situating it within his broader framework of fragility and robustness under uncertainty.
Historical Timeline of the Lindy Effect
The Core Concept Explained
The fundamental distinction the Lindy Effect draws is between perishable and non-perishable entities. Perishable things such as living organisms, food and mechanical components have a biological or physical mortality. Their remaining life expectancy decreases as time passes. A 70-year-old human has fewer expected years ahead than a 30-year-old.
Non-perishable things operate under a fundamentally different rule. Books, ideas, technologies, monetary systems and artistic forms do not age in the biological sense. They are either used or abandoned. Every additional period of survival is evidence of resilience rather than decline. Survival itself updates our estimate of remaining lifespan upward.
Lindy Effect: Remaining Life Expectancy vs Current Age
Non-perishable entities (blue rising line) gain expected life with each year of survival. Perishable entities (amber declining curve) lose expected remaining life as they age.
The Lindy Effect in Economics
The implications for economic analysis are far-reaching. Standard growth models assume mean-reversion; that firms, industries and technologies tend toward some equilibrium state. The Lindy Effect suggests that for certain categories the opposite logic applies. Longevity compounds rather than decays.
1. Firm and Industry Longevity
Research by McKinsey Global Institute tracked the S&P 500 from 1955 through 2020 and found that average company lifespan on the index fell from roughly 61 years to under 18 years. On the surface this seems to contradict Lindy. Yet the subset of firms that survived across multiple decades showed Lindy-consistent dynamics: companies older than 100 years had proportionally longer predicted survival horizons than younger peers and their revenue volatility was systematically lower.
2. Monetary Systems and Currencies
Currency survival is one of the most compelling economic applications of the Lindy Effect. A study by Reinhart and Rogoff spanning 800 years of financial history documented that the average lifespan of a fiat currency is roughly 27 years before it either hyperinflates or is replaced. Yet gold as a monetary medium has persisted for over 3000 years. The Lindy prediction is unambiguous: the longer a monetary instrument has survived across changing political regimes and technological disruptions the stronger the prior that it will continue to do so.
| Monetary Instrument | Age | Lindy Prediction | Lindy Score | Status |
|---|---|---|---|---|
| Gold as Store of Value | ~3,000+ yrs | 3,000+ more yrs | Very High | Active |
| Chinese Renminbi | ~70 yrs (1949) | ~70 more yrs | Moderate | Active |
| US Dollar (post-Bretton Woods) | ~54 yrs (1971) | ~54 more yrs | Moderate | Active |
| Euro | ~27 yrs (1999) | ~27 more yrs | Lower | Active |
| Bitcoin | ~15 yrs (2009) | ~15 more yrs | Lower | Active |
| Zimbabwean Dollar | ~30 yrs before collapse | — | Failed | Replaced (2009) |
| Weimar Papiermark | ~4 yrs (hyperinflation) | — | Failed | Replaced (1924) |
Sources: Reinhart & Rogoff (2009); World Gold Council (2024); Taleb (2012). Predictions are illustrative approximations.
3. Technology Adoption and Obsolescence
Economists studying technology diffusion have long used S-curve adoption models. The Lindy Effect adds a complementary lens: once a technology surpasses the initial adoption phase and becomes embedded in economic infrastructure its remaining useful life scales with its elapsed survival time. The printing press survived over 500 years before digital reproduction began to displace it. Contrast this with technologies that arrive rapidly and vanish within a decade such as the Segway (2001 to 2020), HD DVD (2006 to 2008) and Google Glass consumer edition (2013 to 2015).
Technology Survival vs Age at Observation
Selected technologies plotted by age at peak deployment vs total observed lifespan. Technologies above the diagonal reference line are Lindy-consistent.
Sources: Technology adoption literature; Clauset, Shalizi & Newman (2009); Arthur (2009).
Real-World Economic Case Studies
Gold as a Monetary Anchor
Gold has served as a store of value across the Egyptian, Roman, Byzantine, Islamic and modern Western economies. Its survival through dozens of political collapses provides an extraordinarily strong Lindy prior. Central banks worldwide held approximately 36,700 tonnes as of 2024 and the World Gold Council (2024) notes a 15-year trend of net central bank purchasing.
Double-Entry Bookkeeping
Developed by Luca Pacioli in 1494 and adopted universally by the 1800s, double-entry accounting has survived the industrial revolution, two world wars, digital computing and blockchain disruption. IFRS and GAAP both rest on the same logical foundation Pacioli described over 530 years ago.
Central Banking Systems
The Bank of England (founded 1694) and its model of a lender-of-last-resort have now survived over 330 years including global wars and financial pandemics. The Federal Reserve (1913) approaches its second century. Both have been reformed but never abolished, a Lindy-compatible outcome.
Crypto-Asset Ecosystems
Bitcoin at 15 years old is the oldest significant crypto-asset. Thousands of altcoins that emerged between 2017 and 2021 have already failed outright, consistent with Lindy predictions for very young non-perishables facing severe competition and network effects.
Lindy Effect and Investment Strategy
Value investors have intuitively applied Lindy logic for decades. Warren Buffett's stated preference for businesses with long histories, durable competitive moats and proven management is essentially a Lindy heuristic applied to equity selection. Buffett has expressed that he prefers businesses that were around 50 years ago and will be around 50 years hence, a near-literal statement of the Lindy principle.
Practical Implication: An equity investor using Lindy logic would weight portfolio allocation toward firms with longer survival histories in their sector, preferring a 100-year-old consumer staples company to a 5-year-old high-growth disruptor if forced to choose between equal valuation multiples. The older firm's survival is itself information about its robustness to unknown shocks.
Institutional Age vs Bankruptcy Probability
Failure rates by firm-age cohort across OECD economies. Older firms show markedly lower default probability which is consistent with the Lindy Effect.
Sources: Dun & Bradstreet Business Failure Record; BIS Working Papers on Firm Longevity (2018); Eurostat Business Demography Statistics (2023). Illustrative weighted averages across OECD economies.
Limitations and Criticisms
The Lindy Effect is a heuristic not a law. Several well-documented limitations must be acknowledged for honest economic analysis.
| Criticism | Description | Lindy Response |
|---|---|---|
| Survivorship Bias | We only observe things that survived. Failed technologies and currencies are absent from the dataset. | Taleb acknowledges this; Lindy applies to the observed survivor population as a conditional probability not a universal guarantee. |
| Technological Disruption | Sudden paradigm shifts such as the steam engine, the internet and AI can eliminate even centuries-old systems. | True for perishable-class entities. Truly non-perishable ideas such as writing and arithmetic tend to absorb rather than be destroyed by new technology. |
| Inapplicability to Perishables | Many economists misapply Lindy to firms, people or hardware where biological or mechanical decay applies. | The perishable versus non-perishable distinction is essential. Lindy is not meant to apply universally. |
| Power-Law Assumption | If the true lifespan distribution is not a power law the mathematical foundation weakens. | Clauset et al. (2009) confirm power-law distributions across technology, firm size and cultural artefact lifespan empirically. |
| Political and Regulatory Risk | Government fiat can eliminate even very old institutions overnight through nationalisation or prohibition. | Lindy conditions on the absence of such singular exogenous interventions. Political risk is a genuine fat-tail concern. |
Policy Implications
The Lindy Effect carries underappreciated weight for economic policymakers. Three domains stand out as particularly relevant.
Regulatory Design: Regulations that have survived across multiple economic cycles and political regimes carry a Lindy prior suggesting further survival. Radical deregulation experiments with short track records carry commensurately higher uncertainty. The burden of proof for eliminating long-standing rules should be correspondingly higher than for new ones.
Infrastructure Investment: Public capital budgeting that accounts for Lindy dynamics would weight investment toward modular infrastructure with long proven lifespans such as rail networks, electricity grids and port facilities over highly specialised technology investments with short track records. The UK National Infrastructure Commission's 2023 strategy explicitly references 50-year infrastructure horizons, a Lindy-consistent framing.
Reserve Asset Management: Central bank reserve diversification decisions that ignore Lindy dynamics systematically underweight assets with millennia of monetary history and overweight newer instruments whose survival priors are much weaker. The IMF's 2023 Global Financial Stability Report acknowledged this tension in the growing debate over gold's role in sovereign reserve portfolios.
If something has been around for a very long time and is still doing its job it is telling us something important about its fitness that our models cannot easily quantify.
— W. Brian Arthur, The Nature of Technology, Free Press (2009)Frequently Asked Questions
Not unconditionally. The Lindy Effect provides a prior, a starting probability estimate but not a certainty. An older company carries a stronger prior for continued survival than a younger one in the same sector, all else being equal. But all else is rarely equal. Valuation, leverage, competitive dynamics and management quality all interact with Lindy-derived priors. Taleb himself makes clear that Lindy logic should be used to filter and weight options rather than to mechanically select them. A 200-year-old bank that borrowed recklessly in 2007 still failed in 2008. Lindy logic should be combined with balance sheet analysis and not substituted for it.
The two ideas exist in productive tension. The efficient market hypothesis in its semi-strong form holds that all publicly available information including firm age is already priced into asset values. Under strict EMH, Lindy-based strategies should generate no excess returns. However behavioural finance research by Lakonishok, Shleifer and Vishny (1994) and subsequent literature suggests that markets systematically overweight recent performance and underweight long-run track records. This creates windows where Lindy-informed strategies may yield excess returns which are precisely because Lindy logic runs counter to the recency bias that characterises much market pricing.
This is one of the most philosophically interesting applications and Taleb makes it explicitly in Antifragile. Keynesian demand management has survived for roughly 90 years. The quantity theory of money in various forms has survived over 400 years. A strict Lindy reading would give the quantity theory a substantially stronger prior for continued relevance than more recently developed frameworks. This does not mean the quantity theory is correct in all its claims but it does suggest that policy built entirely on very recent models deserves extra scrutiny.
The mathematical foundation rests on Pareto or power-law distributions. For a random variable X with a Pareto distribution with shape parameter α and scale x_min, the conditional expected value of X given X > t equals αt/(α−1). Since this grows linearly with t the expected remaining life given survival to age t scales proportionally with 't', the Lindy property. This contrasts with Gaussian distributions where expected remaining life eventually falls to zero. Clauset, Shalizi and Newman (2009) provide the canonical statistical framework for identifying power-law distributions empirically and confirm their presence across firm lifespan data, cultural artefact survival and technology longevity.
Recessions are events rather than persistent entities so the Lindy framework applies in a modified way. What the effect can usefully inform is the survival of institutions that have managed through multiple recessions. The Federal Reserve has managed through at least 17 official US recessions since 1913. Its survival and adaptation through each increases the Lindy prior that it will manage through future ones. Contrast this with novel central bank instruments such as yield curve control, tools with very short track records whose long-term effectiveness remains genuinely uncertain.
Three practical habits make Lindy thinking tractable. First, always ask the historical survival question: how long has this institution, technology or framework been in continuous use and in how many different contexts? Second, distinguish carefully between perishable and non-perishable categories before applying the principle. Attempting to apply Lindy to a startup's quarterly revenue is a category error. Third, use Lindy as a prior-setting device and not as a conclusion, begin any analysis with the Lindy-adjusted prior then update systematically using current data and structural analysis. Reading Taleb's Antifragile together with Reinhart and Rogoff's This Time Is Different (2009) provides an excellent empirical companion to the theoretical framework.
References
All sources are listed in citation order. Titles in italic blue are primary sources. Journal names in green are peer-reviewed publications.
Post a Comment