Skip to main content

Power-Law Returns in Modern Growth

Quick definition: Power-law returns describe the phenomenon where a tiny fraction of growth stocks generate the majority of portfolio returns, creating incentive for concentrated positioning and acceptance of significant volatility.

Key Takeaways​

  • Empirical research shows that the top 10 stocks in a broad market index account for 40-50% of total returns over decade-long periods, while the median stock underperforms
  • Technology and growth sectors display more extreme power-law distributions than value sectors because winner-take-most competition creates massive gaps between leaders and followers
  • The concentration of returns in a few businesses stems from exponential growth math: A business doubling in value repeatedly creates orders-of-magnitude larger absolute gains than one doubling once
  • Growth investors face a choice between diversification, which reduces volatility but captures fewer power-law winners, and concentration, which increases volatility but maximizes exposure to outsized returns
  • Understanding power-law distributions explains why concentrated growth portfolios can generate superior long-term returns despite periods of severe underperformance

The Mathematics of Exponential Value Creation​

The power-law distribution emerges naturally from exponential growth mathematics. Consider two businesses, Alpha and Beta, both valued at $1 billion:

Alpha doubles in value three times: $1B → $2B → $4B → $8B. Total return: 8x or 700%.

Beta doubles in value once: $1B → $2B. Total return: 2x or 100%.

Both achieved growth through doubling; Alpha simply doubled three times while Beta doubled once. Yet Alpha generated 8x the absolute value creation and 7x the shareholder return. This difference grows more extreme with each additional doubling.

Now extend this to a portfolio of 50 stocks, each beginning at $1 billion and each doubling in value over a decade. If all 50 doubled once, the portfolio compounded at 100%. But if five businesses double three times, 10 double twice, and 35 double once, the portfolio return exceeds 300%. The concentration effect is powerful.

Growth sectors display more extreme power-law distributions than value sectors because competitive dynamics in growth markets are more winner-take-most. In mature industries, numerous competitors coexist with stable market shares. In growth markets, network effects, switching costs, and scale advantages permit a single leader to capture enormous market share and eliminate most competitors. The winner grows 40% annually; the runner-up grows 5%; competitors exit.

This dynamic means that accurately identifying which businesses will achieve exponential growth—which will double three or four times versus which will double once or zero times—is the dominant source of returns. Missing a 1000x winner and instead investing in a 2x winner costs thousands of basis points of return.

Historical Evidence of Power-Law Concentration​

Academic research quantifies this concentration. Analysis of returns from 1983 to 2023 shows that the top 10 holdings of the S&P 500 accounted for roughly 40% of total returns while the bottom 200 destroyed returns. The median stock in the broad index underperformed Treasury bonds.

The concentration intensified in technology sectors. Studies of venture capital returns show even more extreme distributions: The top 1% of investments account for 75% of all venture returns. Most venture investments fail or produce modest multiples; a tiny fraction generate massive home runs.

This finding has profound implications. It means that through sheer randomness, most growth stock investors will underperform. If 20% of growth stocks become 10-bagger winners and 40% become 0.5x losers, the math of the power law means that a random selection of stocks in a well-diversified portfolio will be dragged down by the failures and miss the massive winners.

Alternatively, investors who concentrate portfolios in areas where they believe power-law winners will emerge can capture outsized returns—if their selection accuracy improves even modestly above random chance.

The Concentration Dilemma​

Growth investors face a genuine tension between diversification and concentration. A fully diversified growth portfolio—holding all 2,000 or so publicly traded growth companies globally—will achieve market-average returns plus or minus fees. This diversification minimizes volatility and tracking error. Most investors will underperform this broad portfolio.

Conversely, a concentrated portfolio holding 20-30 carefully selected growth stocks maximizes exposure to power-law winners but introduces significant volatility. In periods when the chosen stocks underperform, concentrated portfolios suffer large drawdowns. The concentrated investor may underperform for years before their thesis companies achieve power-law scale. Many concentrated investors abandon their approach during these dry spells, selling winners prematurely and locking in losses.

The optimal approach depends on personality, capital availability, and investment thesis quality. An investor with limited capital can justify concentration because diversification with small sums leaves each position too small to meaningfully impact returns. An investor with confidence in their ability to identify power-law winners should concentrate. An investor uncertain about their selection skill or temperamentally unsuited to volatility should diversify broadly.

The Quality Dimension​

Not all growth stocks are equally likely to achieve power-law expansion. Certain characteristics correlate with extreme scaling:

First, competitive moats: Businesses with network effects, switching costs, or economies of scale compound faster because competitors cannot replicate their position. Amazon built unassailable scale in cloud infrastructure; smaller competitors cannot match.

Second, large addressable markets: Doubling from $1B to $2B is easy in a $5B market; further doubling becomes difficult. But doubling from $1B to $2B to $4B to $8B is plausible in a $100B market. Businesses attacking massive categories—cloud infrastructure, digital payments, artificial intelligence—have more room to scale.

Third, pricing power: Businesses that can raise prices without losing customers enjoy improving margins during scaling. Software-as-a-service businesses often achieve this. Commodity e-commerce does not.

Fourth, capital efficiency: Doubling without proportional capital requirements generates exponential returns. Capital-intensive businesses that double require matching capital, limiting the exponential effect.

Sophisticated growth investors screen for power-law probability, not just growth rate. A business growing 30% in a massive market with defensible advantages is more likely to achieve power-law scaling than a business growing 50% in a commodity market with weak competitive positioning.

Portfolio Implications​

Understanding power-law distributions informs portfolio construction for growth investors. If you accept that power-law winners are critical, you can justify:

Concentrated portfolios of 15-30 holdings rather than 100+. Excessive diversification dilutes exposure to power-law winners.

Conviction sizing based on probability of power-law scaling rather than equal-weight approaches. Allocate more capital to businesses most likely to achieve exponential growth.

Long holding periods for conviction positions. Power-law scaling often requires 5-10 years; premature selling captures only modest returns rather than full exponential expansion.

Tolerance for volatility during periods when chosen holdings underperform. The portfolio will underperform in years when value or defensive stocks outperform; this is the cost of maximizing power-law exposure.

Selective rebalancing rather than mechanical rebalancing. Trimming winners to rebalance into underperformers locks in returns prematurely and prevents exponential compounding.

Conversely, diversified investors must accept that they will miss power-law winners. Their trade-off is lower volatility and more consistent performance. Neither approach is wrong; both depend on circumstances and temperament.

Next​

Explore growth opportunities beyond developed markets in Growth Investing in Emerging Markets.