Market returns over long periods are not distributed evenly across companies. A substantial body of research finds that a small minority account for the great majority of the aggregate gain.
This describes research findings rather than advising on any approach. Investment decisions warrant regulated advice.
The skewness finding
The result that reframes the problem.
Studies examining the long-run performance of individual listed companies have found that the median company substantially underperforms the market, and that aggregate returns are driven by a small proportion of extreme winners.
A meaningful proportion of companies produce negative returns over their lifetimes, and a large number underperform cash.
Which means the average return of an index is not the typical return of a company within it, and those two figures differ enormously.
Why this makes selection hard
The implication is uncomfortable.
If most of the return comes from a few companies, then missing those few produces substantial underperformance even with a portfolio of otherwise reasonable choices.
A concentrated portfolio has a high probability of excluding the extreme winners, since they are rare by definition.
Which means the distribution of outcomes for a stock picker is itself skewed — most underperform, some substantially outperform, and the median outcome is worse than the average.
That is a different situation from one where selection is merely difficult, and it explains part of why the aggregate evidence on active management looks as it does.
What it implies about diversification
The mechanical consequence.
Holding more companies increases the probability of including the eventual winners.
Which is a different argument for diversification from the usual one about reducing volatility, and arguably a more consequential one.
It also explains why the arithmetic of index tracking works — an index holds everything, so it holds the winners by construction.
The identification problem
Why hindsight is misleading.
The companies that produced extraordinary returns are obvious in retrospect and were not obviously distinguishable in advance from many similar companies that failed.
Survivorship bias in the stories told about them compounds this, since the failures are not written about.
Which means learning from the winners produces characteristics shared by many companies that did not succeed, and the resulting pattern is not predictive.
What professional investors do about it
Worth noting because it is instructive.
Venture investing explicitly assumes this distribution, holding many positions on the expectation that most fail and a small number produce the return.
Which is a rational response to a skewed distribution and requires the capacity to hold many positions and to tolerate a high failure rate.
Applying that approach with a small number of holdings does not work, since the arithmetic depends on the number.
The behavioural dimension
A related difficulty.
Holding the eventual winners requires holding through substantial declines, since companies that eventually produce extraordinary returns frequently experience severe falls along the way.
Research on investor behaviour finds a persistent tendency to sell winners and hold losers, which is precisely the opposite of what capturing skewed returns requires.
Which means that even correct selection can produce poor outcomes if positions are not held, and the holding is arguably harder than the selection.
What the finding does not establish
Being careful about the limits.
It does not mean nobody can select successfully, and some investors have long records that are difficult to attribute to chance.
It does not address timescales shorter than the long periods studied.
And it says nothing about what any individual should do, which depends on circumstances, objectives and the alternatives available.
What it does establish is that the difficulty is structural rather than a matter of effort, which is worth knowing before deciding how much effort to apply.
What this implies about concentrated positions
A specific situation worth noting.
Employees holding shares in their own employer frequently accumulate a large concentrated position, sometimes alongside a salary and a pension from the same company.
Which concentrates several distinct exposures on a single organisation, and the skewed distribution of company outcomes applies to that holding as it does to any other.
Restrictions on selling frequently apply, and understanding them and the resulting exposure is worth doing deliberately rather than by default.
Costs on frequent trading
A practical drag that compounds the difficulty.
Each transaction incurs the spread, any commission, and in some jurisdictions a transaction tax.
Which means frequent trading imposes a cost that must be overcome before any selection skill produces a net benefit.
Research examining individual investor trading records has consistently found that more active traders underperform less active ones after costs.
The finding holds across markets and is one of the more robust results in the literature on individual investing.
The information question
A structural point about what an individual is competing against.
Prices reflect the aggregate of what participants know, and professional participants have research resources, data access and analytical capacity that an individual does not.
Which means acting on publicly available information is acting on information already reflected in the price.
This does not mean individuals cannot invest successfully; it means the advantage is unlikely to come from analysing the same public information faster or better.
Where individuals do have genuine advantages, they are generally in patience and in the absence of pressure to perform over short periods.