The SPIVA report, which measures the performance of actively managed funds against their respective benchmarks globally, produces a finding that is both consistent and sobering: between 75% and 90% of actively managed funds underperform their benchmarks over extended periods. This result holds across geographies, asset classes, and market cycles. It persists despite the considerable analytical resources, experience, and market access that professional fund managers bring to bear. Understanding why requires examining what a benchmark actually measures, how financial markets differ from ergodic systems, and why survival rather than outperformance is the correct primary objective for any strategy designed to compound wealth over the long run.
What a Benchmark Actually Measures
The S&P 500 index is commonly described as a representation of the average performance of the US equity market. This description is accurate but incomplete in a way that matters enormously for understanding why beating the benchmark is structurally difficult.
The value of the index at any given moment is a snapshot of the collective decisions of all market participants active at that precise time. It is a fossilised record of a specific instant, capturing the aggregate of buying and selling activity from the participants who are present. It does not represent the average experience of all participants across time. It represents only those who have survived to that point.
The marathon analogy makes this concrete. Imagine a marathon with thousands of entrants. If average completion times are calculated only from runners still on the course at each checkpoint, those averages will not reflect the full population of entrants. They will reflect only those who have not yet dropped out. As the race progresses and more participants withdraw due to injury, exhaustion, or any other cause, the average time at each checkpoint increasingly reflects the surviving population rather than the original field. The average improves not because all runners are getting faster but because slower runners are progressively removed from the calculation.
Financial markets work the same way. Approximately 90% of market participants exit within a decade. The benchmark at any given moment reflects the performance of the surviving 10%, not the experience of the full population of participants who started. A fund manager attempting to beat the benchmark is not competing against a fixed hurdle. They are competing against a dynamically recalculated average that continuously removes underperformers from its composition, leaving only the survivors to represent it.
The Non-Ergodic Nature of Markets
This survivorship dynamic is a specific expression of a more fundamental property of financial markets: they are non-ergodic. In an ergodic system, the time-averaged experience of a single participant converges to the ensemble average across all participants at a single point. The long-run experience of one investor, given sufficient time, would mirror the cross-sectional average of all investors at any moment.
Financial markets do not have this property. The long-run experience of an individual market participant diverges systematically from the cross-sectional average observed at any single point, for the precise reason that the population of participants changes continuously. New participants enter with capital and assumptions formed from a different market history than the incumbents. Participants exit, voluntarily or otherwise, when their capital is impaired beyond the threshold of recovery. The composition of the market at any moment does not represent a stable population whose collective experience approximates any individual’s long-run trajectory.
The practical implication is that the performance benchmark, constructed from the current population of market survivors, is not a meaningful representation of what any individual participant will experience over time. The individual faces a sequence of market conditions, regime shifts, and fat-tail events whose cumulative path dependence determines their long-run outcome. The benchmark, recalculated from a continuously refreshed survivor pool, does not face this path dependence in the same way. It absorbs the losses of departing participants into its reconstitution process rather than carrying them forward as impairments to its compounding base.
The Absorbing Barrier and the Risk of Ruin
Central to understanding long-run market participation is the concept of the absorbing barrier: the level of capital impairment from which a participant cannot recover and must exit the market entirely. Once a participant reaches this barrier, their future participation ends. They cease to influence market prices, cease to benefit from future trends, and cease to compound whatever capital they might have preserved. The absorbing barrier is not merely a bad outcome. It is a terminal state.
The asymmetry between recoverable losses and unrecoverable ones is severe. A 20% loss requires a 25% gain to return to the prior peak. A 50% loss requires a 100% gain. A 75% loss requires a 300% gain. The arithmetic of recovery deteriorates non-linearly as losses increase, and at some level of impairment, the recovery required exceeds any realistic expectation of future performance. The participant has effectively reached the absorbing barrier even if they have not formally exited.
This asymmetry defines the primary risk management objective for any strategy seeking long-run outperformance: not maximising returns in any given period, but avoiding the absorbing barrier across all periods. A participant who survives through numerous market cycles, including the cycles that eliminate the majority of their competitors, compounds from an intact capital base across the full history of their participation. A participant who reaches the absorbing barrier early in their career forfeits all the compounding that would have followed. The expected long-run difference between these two outcomes is enormous, and it is determined not by the magnitude of the returns achieved in favourable periods but by the severity of the losses incurred in adverse ones.
Geometric Compounding and the Primacy of Sequence
The mathematical expression of this principle is the distinction between the arithmetic mean return and the geometric compounded return, best expressed through CAGR. The arithmetic mean of a return series treats each period’s return as an independent contribution to the average. The geometric mean, the CAGR, accounts for the compounding base at each step and reflects the actual rate at which wealth grows over time.
The two diverge when returns are volatile, and the direction of that divergence is always unfavourable: the geometric mean is always less than or equal to the arithmetic mean, and the gap between them increases with volatility. This is variance drain: the mathematical consequence of compounding through a volatile return series is that the long-run wealth outcome is always lower than the arithmetic mean return would predict. A strategy with a high arithmetic mean return but high volatility may produce lower long-run CAGR than a strategy with a lower arithmetic mean return but lower volatility, precisely because the compounding base is eroded more severely by the larger drawdowns in the high-volatility strategy.
The sequence of returns compounds this further. A strategy that incurs its large losses early in its history compounds from a reduced base during all subsequent periods. A strategy that incurs its large losses late in its history has had the benefit of compounding from a larger base during the earlier periods of gains. The terminal wealth outcome differs materially between these two sequences even when the arithmetic mean return and standard deviation are identical. Path dependence is not a theoretical curiosity. It is the mechanism through which the timing and severity of losses determines long-run wealth outcomes, independent of any aggregate performance metric.
The TTU TF Index: Survival as a Selection Criterion
The TTU Trend Following Index provides a specific and instructive illustration of these principles. Its construction differs from indices such as the SG Trend Index in one critical dimension: inclusion requires a minimum track record of fifteen years. The SG Trend Index, by contrast, selects constituents on the basis of assets under management rather than longevity of track record, which means its composition is influenced by recent performance and capital flows rather than by demonstrated survival across multiple market cycles.
The consequence of this difference in construction is visible in Figure 1, which compares the performance of the TTU TF Index against both the SG Trend Index and the S&P 500 TR Index. The TTU TF Index delivers substantially higher returns over the comparison period. The outperformance is not attributable to superior short-run performance by its constituents in any given period. It is attributable to the compounding benefit of fifteen or more years of continuous participation by managers who have navigated the full range of market conditions, including the conditions that eliminated the majority of their contemporaries.
The SG Trend Index, by contrast, undergoes constituent changes driven by fluctuating performance levels and assets under management. Managers who underperform are replaced by managers who have recently outperformed. This reconstitution process introduces a performance dilution effect: the index systematically removes participants at low points and replaces them with participants at high points, which impairs its long-term compounding in exactly the same way that survivorship bias impairs the benchmark comparisons discussed earlier.
Figure 1: Comparative Performance of TTU TF Index, SG Trend Index and S&P500TR Index
Note: For more information about the TTU TF Index, refer to the monthly Trend Following Performance Reports prepared on Top Traders Unplugged in the Blog posts. https://www.toptradersunplugged.com/blog
The Barbell Architecture of Long-Run Survival
The strategy that allows trend followers to survive long enough to benefit from the compounding advantages illustrated by the TTU TF Index is not a strategy optimised for maximum return in any single period. It is a barbell strategy that rigorously protects realized capital on one end while actively pursuing the large gains available from Outlier trends on the other.
The realised capital end of the barbell is non-negotiable. Position sizing, stop-loss discipline, and broad diversification across uncorrelated markets and timeframes collectively ensure that no single adverse event, and no cluster of adverse events, can impair the capital base to the point of reaching the absorbing barrier. This requires restraint in the use of leverage and an acceptance that individual trade returns will be modest relative to the portfolio’s full potential, because the leverage and concentration required to maximise individual trade returns are incompatible with long-run survival across the full distribution of market conditions including fat-tail events.
The unrealized equity end of the barbell is where the compounding advantage is generated. As trends develop and produce profits above the initial capital base, those profits become the resource for increased participation in continuing Outlier trends. The asymmetric exit structure of the strategy, cutting losses short and letting profits run, ensures that the return distribution is positively skewed: frequent small losses offset by infrequent large gains. Over thousands of trades and multiple market cycles, this positive skew, combined with the compounding advantage of an intact capital base maintained through disciplined survival, produces the trajectory of geometric returns that the TTU TF Index exemplifies.
The Source of Long-Run Outperformance
The difficulty of consistently exceeding performance benchmarks is not a mystery when the mechanics of benchmarks and compounding are properly understood. Benchmarks are survivorship-filtered snapshots of a non-ergodic system. They continuously remove underperformers from their composition, producing an average that reflects only the current survivor population rather than the full historical experience of all participants. Any strategy attempting to beat this benchmark must not merely perform well in favourable conditions. It must survive the conditions that eliminate the majority of its competitors, compound from an intact capital base across the full history of its participation, and accumulate the geometric compounding advantage that only long-run survival makes possible.
Trend followers who prioritise survival, manage the absorbing barrier risk through disciplined position sizing and stop-loss architecture, and maintain broad diversification across the full universe of liquid markets are structurally positioned to achieve this. The compounding of modest returns over thousands of trades across multiple decades, without the catastrophic impairment that ends most market careers early, is the mechanism through which the benchmark is eventually exceeded. The key is not brilliance in any single period. It is endurance across all of them.