The conventional toolkit for measuring investment risk was built on a set of assumptions that trend-following strategies violate structurally. The Sharpe Ratio, Standard Deviation, and Sortino Ratio have served as the dominant framework for evaluating geometric return outcomes across the investment management industry for decades. Applied to trend-following models, they produce assessments that are not merely incomplete but actively misleading. Understanding why requires examining what these metrics actually measure, what they assume about the distribution of returns, and why those assumptions are incompatible with the mechanics of Outlier-targeting strategies.
What Conventional Metrics Actually Measure
The Sharpe Ratio expresses the excess return of a strategy per unit of volatility, with volatility measured as the standard deviation of returns. Its appeal is its simplicity: a single number that appears to summarise the efficiency of a strategy’s return relative to its risk. The problem is that standard deviation treats all deviation from the mean as equivalent. A large upward move and a large downward move of identical magnitude are indistinguishable from the perspective of standard deviation. Both contribute equally to the volatility measure, and therefore both penalise the Sharpe Ratio equally.
For a trend-following strategy whose return distribution is characterised by frequent small losses and occasional large gains, this symmetry assumption produces a systematic distortion. The large gains that define the strategy’s long-run performance are penalised in the same way as large losses. A strategy with a positively skewed return distribution, which is structurally what a well-designed Outlier Hunter produces, will appear worse by the Sharpe Ratio than a strategy with a negatively skewed distribution of identical mean return and identical standard deviation. The metric is penalising the very property that makes the trend-following strategy valuable.
The Sortino Ratio attempts to correct this by measuring only downside deviation rather than total standard deviation. The correction is directionally right but insufficient. The Sortino Ratio still assumes that returns are distributed in a way that makes standard statistical moments meaningful descriptors of the distribution’s shape. Trend-following return distributions have fat tails: the probability of extreme outcomes is materially higher than a normal distribution would predict. The Sortino Ratio, like the Sharpe Ratio, does not capture this property. It produces a more relevant downside measure than the Sharpe Ratio, but it still fails to describe the full character of the distribution or its implications for long-run compounded wealth.
Path Dependence and the Geometry of Returns
The deeper problem with conventional metrics is not that they measure the wrong type of volatility. It is that they are path-independent by construction, and trend-following returns are path-dependent by nature.
A path-independent performance metric aggregates outcomes across the full return history without regard to the sequence in which those outcomes occurred. It treats a strategy that produced returns of +10%, -5%, +8%, -3% as identical to a strategy that produced -5%, -3%, +10%, +8%, provided the aggregate statistics are the same. From the perspective of geometric compounding, these sequences are not identical. The sequence of returns determines the compounding base at each step, and therefore determines the terminal wealth produced by the strategy. A strategy that incurs its losses before its gains compounds from a lower base during the gain periods. A strategy that generates its gains before its losses compounds from a higher base. The terminal wealth outcome differs materially between the two sequences even when the arithmetic mean return and standard deviation are identical.
This is variance drain in practice: the gap between the arithmetic mean return and the geometric compounded return widens as volatility increases and as the sequence of returns produces unfavourable compounding dynamics. A strategy assessed as having adequate geometric return outcomes by conventional metrics can produce materially different long-run wealth outcomes depending on the sequence of its returns, a dimension that those metrics are structurally unable to capture.
For the Outlier Hunter, the sequence of returns is a defining characteristic of the strategy. Extended periods of small losses, accumulated while waiting for Outlier events that have not yet arrived, are followed by the large gains that define the strategy’s long-run performance. The path matters. The timing of the Outlier relative to the drawdown that preceded it determines the compounding base from which the Outlier gain compounds. Metrics that average across the full history without accounting for this sequence are not measuring what matters most for the long-run geometric return outcomes the strategy is designed to produce.
Drawdown Geometry as the Primary Risk Language
The metrics appropriate for evaluating trend-following strategies are those that speak the language of drawdown geometry rather than the language of symmetric volatility. The relevant questions are not how volatile a strategy is in aggregate, but how far it falls from its peak, how long it remains below that peak, and how efficiently it recovers. These are the dimensions of risk that determine whether a trend-following strategy survives long enough to capture the Outliers that justify its existence.
Maximum drawdown measures the largest peak-to-trough decline over the full history of the strategy. It is the most direct measure of the capital preservation question: what is the worst the strategy has done, and is that level of loss survivable for a participant who is committed to the process? Maximum drawdown alone is insufficient, however, because it provides no information about how long the strategy spent in that drawdown or how efficiently it recovered.
The MAR ratio addresses the return efficiency question by expressing the annualised CAGR of the strategy relative to its maximum drawdown. A high MAR ratio indicates that the strategy has generated strong compounded returns relative to the worst drawdown it has produced. This is a more meaningful measure of geometric return outcomes for a trend-following strategy than any volatility-normalised metric, because it directly relates the compounding engine of the strategy to its most severe capital risk.
The Calmar Ratio performs a similar function, expressing annualised compound return relative to maximum drawdown over a specified period. Its value as a complement to the MAR ratio lies in its sensitivity to the time frame over which it is calculated: a strategy’s MAR ratio over its full history may differ materially from its Calmar ratio over the most recent three years, and this divergence can be informative about whether the strategy’s drawdown characteristics have changed over time.
The Ulcer Index extends the analysis of drawdowns beyond their maximum depth to incorporate their duration. It measures the sustained discomfort of being in drawdown, weighting both how far below the peak the strategy is and how long it remains there. For trend-following strategies, which can experience extended flat or declining periods between Outlier events, the Ulcer Index captures the experiential reality of managing the strategy in a way that maximum drawdown alone does not.
The Serenity Ratio builds on the Ulcer Index by relating the strategy’s return to the drawdown discomfort measured by the Ulcer Index. A higher Serenity Ratio indicates that the strategy delivers meaningful returns relative to the duration and depth of the drawdown periods investors must endure to receive them. This is a practically useful measure for assessing whether the compounding benefit of the strategy justifies the patience it demands.
Conditional Drawdown at Risk (CDaR) examines the tail of the drawdown distribution: the average loss across the most severe drawdown scenarios, those that exceed a defined threshold. For a strategy targeting Outliers in a non-ergodic, fat-tailed market environment, the tail of the drawdown distribution is not a theoretical curiosity. It is the domain in which absorbing states, the drawdowns that permanently impair capital and end the strategy’s ability to recover, become possible. CDaR quantifies this domain in a way that maximum drawdown, which captures only the single worst historical event, cannot.
The Shape of the Return Distribution
Alongside the drawdown-based metrics, the shape of the return distribution provides essential information about a trend-following strategy’s structural properties. A histogram of trade outcomes for an Outlier-targeting strategy will exhibit positive skew: the majority of trades produce small losses or small gains, and a minority of trades produce the large gains that define the strategy’s long-run performance. The distribution has a long right tail. This positive skew is not an accidental property of the strategy’s historical record. It is the structural consequence of cutting losses short and letting profits run consistently across a diversified portfolio of markets and timeframes.
The practical importance of understanding this skew is that it inverts the intuitions built on experience with negatively skewed strategies. A strategy with negative skew produces frequent small gains and occasional large losses. It generates smooth equity curves and high conventional risk metrics but is vulnerable to the tail events that conventional metrics systematically underestimate. A positively skewed trend-following strategy produces the opposite: a choppy equity curve with frequent small losses, conventional metrics that appear relatively unimpressive, and the occasional large gain that drives long-run compounding. Evaluating the trend-following strategy through the lens of conventional metrics designed for negatively skewed distributions produces the wrong assessment. The equity curve and the return histogram together provide the visual evidence that the conventional metrics obscure.
Choosing the Right Analytical Lens
The inadequacy of conventional risk metrics when applied to trend-following strategies is not a minor technical issue. It determines whether a strategy is correctly understood, correctly evaluated, and correctly compared to alternatives. A practitioner who assesses an Outlier-targeting strategy using the Sharpe Ratio will systematically undervalue it relative to negatively skewed strategies whose apparent smoothness conceals fat-tail exposure. A practitioner who evaluates drawdown characteristics using MAR, Calmar, the Ulcer Index, and CDaR alongside the return distribution’s skew profile will have a substantially more accurate picture of what the strategy actually does and what it requires from the participant in terms of patience and process commitment.
The metrics appropriate for trend-following evaluation are those that measure the dimensions of risk that actually matter for long-run geometric compounding: the depth and duration of drawdowns, the efficiency of recovery, the tail properties of the loss distribution, and the shape of the return distribution. These are the measures through which the Outlier Hunter’s process should be assessed, and through which its compounding advantages over time become visible.
For a detailed technical exploration of the risk metrics discussed here, including visual illustrations of equity curve geometry and return distribution shapes, the full analysis is available at Aussie Turtles: https://www.aussieturtles.com/navigating-the-complexities-of-risk-in-trend/