At Fenyx Capital, we have anchored our risk management framework in Value at Risk (VaR), specifically employing a dynamic 6.5% monthly VaR at a 95% confidence level. This is not a choice made lightly, nor is it a concession to industry convention. It is, instead, a deliberate alignment with a framework that answers the three questions that matter most: how much might we lose, how likely is that loss, and over what timespan?
The appeal of VaR lies in its conceptual clarity. Unlike drawdown, which chronicles the wounds of yesterday, VaR casts its gaze forward. It is prospective rather than retrospective, concerned not with what has happened but with what could happen. This distinction is not semantic. It is fundamental. An investor who relies solely on historical drawdown is navigating by looking in the rearview mirror, a practice that may offer comfort but rarely prevents collision.
The architecture of uncertainty
To understand why VaR has become our cornerstone metric, one must first appreciate what it attempts to capture. VaR provides a statistical estimate of potential loss within a defined time horizon and at a specified confidence level. When we state that our portfolio operates with a 6.5% monthly VaR at 95% confidence, we are saying something precise: in nineteen months out of twenty, we expect our losses to remain below 6.5%. Conversely, in one month out of twenty, we may experience a drawdown exceeding this threshold.
This framing accomplishes something that many risk metrics fail to achieve. It translates the abstract language of volatility into the concrete terms of probability and magnitude. It answers the questions that keep investors awake: not merely “is this portfolio risky?” but rather “what is the worst I might reasonably expect, and how often might I expect it?”
The confidence level we have selected, 95%, represents a deliberate calibration. A 99% confidence level would be more conservative, constraining risk further and potentially dampening returns. A 90% confidence level would permit greater volatility, accepting more frequent excursions beyond the threshold. The 95% level strikes what we believe to be the optimal balance for our investor base: it provides substantial protection against adverse outcomes while permitting sufficient flexibility to capture market opportunities.
A benchmark grounded in reality
The selection of 6.5% as our VaR threshold is not arbitrary. It reflects a careful consideration of market history and the risk characteristics inherent in equity investing. The S&P 500, the most widely followed benchmark of U.S. equity performance, experiences monthly losses exceeding 6.5% approximately once every twenty months. This empirical observation provides a useful calibration point. It suggests that a 6.5% monthly VaR at 95% confidence aligns our risk tolerance with the natural volatility of equity markets themselves.
This alignment is instructive. We are not attempting to construct a portfolio that is dramatically more conservative than passive equity exposure, which would likely sacrifice returns unnecessarily. Nor are we seeking to assume substantially greater risk than the equity markets themselves exhibit, which would expose our investors to potential losses beyond what they might reasonably expect from a diversified portfolio. Instead, we are positioning ourselves within the established risk envelope of equity markets, while using our diversification across bonds and gold futures to potentially moderate that risk during periods of stress.
This benchmark also serves a communicative function. When we tell investors that our VaR target aligns with the historical frequency of significant S&P 500 drawdowns, we provide context that pure statistics cannot convey. Most investors have lived through multiple market corrections. They understand, at an intuitive level, what it means to experience occasional months where equity portfolios decline by more than 6%. By anchoring our VaR target to this familiar experience, we make risk tangible rather than abstract.
Forward vision in a backward-looking world
The criticism most frequently leveled against VaR is that it relies on historical data, and history, as every prospectus reminds us, is not predictive of future results. This objection deserves serious consideration. Markets do not repeat themselves with mechanical precision. Black Swan events, by their very nature, lie outside the boundaries of historical experience. A risk model calibrated to the placid markets of the mid-2000s would have offered little protection against the convulsions of 2008.
Yet this critique, while valid, misses a deeper truth. The alternative to using historical data is not some pristine method of divining the future. It is either the abandonment of quantitative risk management altogether, or the substitution of guesswork for analysis. Historical data, for all its imperfections, represents the only empirical foundation we possess. The question is not whether to use it, but how to use it wisely.
Our approach incorporates several safeguards against the limitations of backward-looking analysis. First, we employ a historical simulation method that captures the full texture of past market behavior, including the fat-tailed distributions and non-linear relationships that characterize financial markets. This stands in contrast to parametric VaR models that assume normal distributions, a dangerous simplification given the frequency of extreme events.
Second, we complement our VaR calculations with stress testing and scenario analysis. We simulate extreme conditions that may not appear in our historical dataset: sudden geopolitical ruptures, sharp liquidity contractions, coordinated central bank interventions. These exercises force us to contemplate scenarios that sit uncomfortably outside our comfort zone. They serve as intellectual fire drills, preparing us for contingencies that VaR alone cannot capture.
Third, we recognize that market regimes shift. The volatility of 2020 bears little resemblance to the quiescence of 2017. Our VaR calculations are updated daily, ensuring that our risk assessments remain responsive to current market conditions rather than anchored to outdated assumptions. This dynamism is critical. A static risk model is an artifact, interesting perhaps, but irrelevant to the task of managing capital in real time.
The tyranny of the point estimate
One might argue that VaR’s reduction of risk to a single number is itself a form of dangerous oversimplification. After all, a portfolio that loses exactly 6.5% in a bad month and a portfolio that loses 25% in a bad month might both violate a 6.5% VaR threshold, yet the implications for capital preservation are vastly different. VaR tells us nothing about the severity of losses beyond its threshold, the so-called tail risk that has undone countless sophisticated investors.
This objection has merit. VaR is not a complete description of risk. It is a boundary marker, not a map of the entire territory. For this reason, we do not rely on VaR in isolation. We monitor additional metrics: conditional VaR (which estimates the average loss in scenarios where VaR is exceeded), maximum drawdown, and the Sharpe ratio. Each metric illuminates a different facet of risk. Together, they provide a more complete picture than any single measure could offer.
Yet even with these limitations acknowledged, VaR retains a clarity that other metrics lack. It is intuitive in a way that concepts like skewness or kurtosis are not. An investor can understand what a 6.5% monthly VaR at 95% confidence means without needing a doctorate in statistics. This accessibility is not trivial. Risk management that cannot be communicated clearly is risk management that will not be adhered to under pressure. When markets turn volatile and fear clouds judgment, the simplicity of VaR becomes a virtue rather than a limitation.
The diversification complement
VaR, for all its virtues, is not a standalone solution. It must be paired with thoughtful portfolio construction. At Fenyx Capital, our diversification across equities, bonds, and gold futures provides a natural foundation for risk management. These asset classes respond differently to various economic scenarios. Equities thrive in periods of growth. Bonds provide ballast during downturns. Gold offers protection against currency debasement and geopolitical uncertainty.
This diversification creates a portfolio whose constituent parts may move in different directions, reducing the likelihood of catastrophic loss. When we calculate VaR for a diversified portfolio, we capture not just the individual risks of each asset class but also the correlations between them. A portfolio of highly correlated assets offers little true diversification, and its VaR will reflect this fragility. A portfolio of genuinely uncorrelated or negatively correlated assets will exhibit lower VaR, a mathematical expression of its resilience.
The synergy between diversification and VaR-based leverage management is profound. Diversification provides the structural foundation, the architecture of resilience. Dynamic VaR-based leverage provides the adaptive response, the ability to dial risk up or down as conditions warrant. Together, they create a portfolio that is both robust and responsive, capable of weathering adverse scenarios while remaining positioned to capture opportunities when they emerge.
The path forward
The choice to anchor our risk management in Value at Risk is not an assertion that VaR is perfect. It is an acknowledgment that in the realm of portfolio management, perfection is unattainable. Every risk metric involves tradeoffs. Every model rests on assumptions. The question is not whether a framework is flawless, but whether it is fit for purpose.
VaR, particularly when implemented dynamically and complemented with stress testing and thoughtful diversification, meets our requirements. It provides forward-looking risk assessment. It translates abstract volatility into concrete probabilities and magnitudes. It informs rather than prescribes, offering a foundation for judgment rather than a replacement for it. And crucially, it allows us to communicate risk in terms that our investors can understand and evaluate.
In an industry too often characterized by complexity that obscures rather than illuminates, this clarity is no small achievement. We have built our risk management framework around VaR not because it is fashionable, but because it works. It allows us to navigate uncertainty with a steady hand, adjusting our exposure as conditions change while maintaining a consistent risk profile. It transforms the abstract specter of potential loss into a measurable, manageable reality.
The markets will continue to surprise us. Volatility will spike and subside. Correlations will shift. Black Swans will land when least expected. But with a robust VaR framework guiding our decisions, we believe we are positioned to weather these storms while preserving capital and delivering risk-adjusted returns. In the end, that is what prudent risk management demands: not the elimination of uncertainty, but its careful navigation. VaR provides the compass. The journey remains ours to navigate.
Disclaimer: This article is for informational purposes only and does not constitute investment advice. Investors should conduct their own due diligence or consult with a financial advisor before making any investment decisions.
Photo by Loic Leray on Unsplash.
