Understanding Warehoused Risk and Why Stops are Critical Risk Management Tools for Classic Trend Followers

Understanding Warehoused Risk and Why Stops are Critical Risk Management Tools for Classic Trend Followers This blog explores the concept of warehoused risk and how effective portfolio management can enhance returns while mitigating risk. When trading, we often focus on individual strategies and their performance metrics, such as drawdown (DD) and compound annual growth rate (CAGR). However, the real power of trading lies in managing a portfolio of these strategies. This blog explores the concept of warehoused risk and how effective portfolio management can enhance returns while mitigating risk. Importantly, it demonstrates a curious conservation law in investing: “Risk cannot be eliminated from a portfolio; it can only be transferred within it.” The only way to release risk from a portfolio is by eliminating the risk contribution of an individual return stream by exiting that position. We will demonstrate why stops, though often criticized as inefficient for mitigating risk, are essential at the portfolio level. They provide risk release valves to mitigate current risk and allow the portfolio to absorb new risk in the future. Not having stops in place can detrimentally affect portfolio performance, especially when market conditions arise that have never been seen before in backtests. These new, unforeseen environments can expose the warehoused risk that exists in a portfolio, which may not have been previously observed. Single Strategy vs. Multiple Strategies Consider a single trading strategy on a single market. This strategy might produce a return profile with a 20% drawdown and a 7% CAGR. If we multiply this strategy five times, trading five identical systems on the same market with the same allocation, the drawdown increases to 100%. This happens because the drawdown in each system occurs simultaneously, resulting in a linear relationship between leverage and drawdowns. Multiplying the position size by 5x causes the drawdown to increase fivefold. However, the relationship with CAGR is not linear. While multiplying the strategy 5x might achieve a drawdown of 100%, the corresponding CAGR might only increase to around 30%. This is because CAGR is a path-dependent and nonlinear metric, where increased volatility suppresses compound growth. Now, what happens if we diversify our approach? Suppose we develop 10 different, uniquely configured trend-following (TF) strategies for the same market. Each of these strategies has a 7% CAGR and a 20% drawdown occurring at different points in time. The portfolio of these 10 strategies might then produce a CAGR of 40% but with a drawdown of only 40%. This is because the risks are spread out due to the lack of perfect correlation between the strategies. The drawdowns of each unique strategy do not coincide, and each return stream offers correlation offsets at different points in time across the entire time series. Furthermore, the CAGR is increased as the volatility drag associated with the drawdown of the entire ensemble at 40% is far less than the alternative of 100%. By diversifying, we distribute risk across various strategies, each contributing to the overall performance at different times. This risk spreading, due to the lack of perfect correlation, allows for a more stable and robust portfolio, enhancing returns while managing drawdowns more effectively. The Principle of Warehoused Risk Warehoused risk is a crucial concept in portfolio management. It represents the total risk present in all return streams, calculated as if each return stream went to zero at a specific point in time. By summing the potential risk contributions of each return stream, we arrive at the total risk summation, known as the warehoused risk of that portfolio at a given moment. This theoretical limit adheres to the conservation law that risk can never be eliminated, only transferred within the portfolio. Another term used to describe warehoused risk is “Portfolio Heat.” Many investors overlook this potential risk lurking in their portfolios—the risk of total collapse if all the risk held by a portfolio is suddenly released at once. Think of warehoused risk as akin to a “risk sponge.” The more we diversify and add new return streams to a portfolio, each with the same risk contribution, the more we pack warehoused risk into the portfolio. Despite the presence of warehoused risk, the risk investors typically pay attention to are measures such as the Sharpe ratio, Sortino Ratio, MAR (maximum drawdown), Ulcer Index, and other risk metrics. These measures are always far lower than the warehoused risk that actually resides in a portfolio. They reflect how individual risks within a portfolio offset each other and how risk events are dispersed across the time series of different return streams. However, these metrics often understate the actual risk potential within a portfolio. These risk measures typically assess the volatility of portfolio returns over time, a consequence of how discrete return streams interact. Some risks cancel each other out, resulting in a net portfolio variance measure, such as standard deviation or maximum drawdown, occurring at specific points in time. However, these proxy measures understate the possible risks if a new market regime emerges—one that has never been experienced in backtests and significantly alters these risk metrics, reflecting the higher warehoused risk inherent in the portfolio. Understanding warehoused risk helps investors recognize the potential hidden dangers within their portfolios. By acknowledging this risk and managing it through diversification and strategic use of stops, we can create more resilient portfolios that are better prepared for unexpected market conditions. The Role of Stops in Portfolio Management Trend followers often use stops as a critical tool for managing portfolio risk. Stops should be seen as risk release valves for the entire portfolio, preventing warehoused risk from becoming overwhelming due to any contributing return stream in unfavourable market conditions. While some traders argue that stops detract from performance compared to other exit measures, in portfolio management, stops are crucial for maintaining the positive skew of the entire collection of return streams and managing total portfolio heat. Consider this: in a portfolio comprising potentially thousands of return streams, there is always the possibility that many return streams could suddenly become positively correlated, potentially

Navigating the Complexities of Risk in Trend

Navigating the Complexities of Risk in Trend In a Non-Linear World View Straight Lines with Caution: Risk Management Principles for Trend Following The ability to adeptly navigate risk stands as a crucial skill for both investors and traders. The quest for robust investment strategies has traditionally leaned on a set of established risk metrics, with the Sharpe Ratio, Sortino Ratio, and Standard Deviation at the forefront. These metrics have provided a foundational framework for assessing the risk-adjusted performance of various investment approaches, offering a semblance of predictability and control in the inherently unpredictable nature of financial markets. The Sharpe Ratio, for instance, has been widely revered for its simplicity and effectiveness in conveying the amount of excess return per unit of risk, with risk quantified as the standard deviation of returns. Its appeal lies in its straightforwardness, allowing for a quick comparison of different investment opportunities under a common risk-return lens. Similarly, the Sortino Ratio refines this concept by focusing solely on downside risk, aligning more closely with the typical investor’s aversion to losses. Standard Deviation, on the other hand, offers a direct measure of volatility, serving as a proxy for the uncertainty inherent in investment returns. Despite their widespread adoption and inherent virtues, these traditional metrics are not without limitations, especially when applied to the intricate domain of trend-following strategies. Trend-following models, characterized by their reliance on capturing sustained directional market movements, embody a unique set of attributes that challenge the applicability of conventional risk assessments. The core of these strategies lies in their path-dependent nature, where the sequence and timing of market trends significantly influence their performance outcomes. The primary shortfall of metrics like the Sharpe and Sortino Ratios in this context is their inherent assumption of symmetry and normality in return distributions. These metrics do not distinguish between upward and downward volatility, treating all fluctuations around the mean as equal contributors to risk (Refer to Figure 1). This oversimplification glosses over the nuanced dynamics of trend-following strategies, where the asymmetry of returns—frequent small losses punctuated by occasional large gains—is a defining characteristic. Moreover, the path-dependent nature of trend-following strategies introduces a layer of complexity that traditional metrics are ill-equipped to handle. The success of these strategies hinges not just on the magnitude of market movements, but on the sequence in which these movements occur. A series of small gains followed by a significant upward trend can result in a vastly different outcome than the same trend occurring in reverse order. This aspect of path dependence is critical in understanding the risk and potential of trend-following models, yet it remains conspicuously absent from the risk assessment toolkit provided by traditional metrics. Figure 1: Three Different Strategies with Three distinct Paths of Returns with Identical Sharpe Ratios and Standard Deviations. Without seeing the NAV chart and only viewing the classic risk-return metrics, an investor would be indifferent: all three strategies would look equally good, however the impact of these paths on a return series have serious consequences for compounded wealth. In essence, while traditional risk metrics like the Sharpe Ratio, Sortino Ratio, and Standard Deviation have served as valuable tools in the arsenal of investors and traders, their application to trend-following models reveals inherent limitations. The dynamic and complex nature of these strategies, thriving on the ebbs and flows of market trends, demands a more nuanced approach to risk evaluation—one that considers the asymmetry of returns and the pivotal role of path dependence in shaping investment outcomes. The Limitations of Conventional Risk Metrics The limitations inherent in conventional risk metrics such as the Sharpe and Sortino Ratios extend beyond their mathematical formulations to the very core of how we perceive and measure investment risk. These metrics, while elegant in their simplicity, often fall short of providing a holistic view of an investment’s risk profile, especially in the context of specialized strategies like trend-following. The Sharpe Ratio, for instance, has been a linchpin in the arsenal of risk assessment tools, offering a succinct measure of risk-adjusted performance. By dividing the excess return of an investment by its volatility, it ostensibly provides a clear indicator of the return an investor can expect per unit of risk undertaken. However, this metric’s reliance on standard deviation as a proxy for risk introduces a critical blind spot: it does not differentiate between positive and negative volatility. In the realm of trend-following strategies, where profits often stem from “riding” prolonged market trends, this failure to distinguish between beneficial volatility (upside) and harmful volatility (downside) can lead to misleading interpretations of an investment’s true risk profile (Refer to Figure 2). Figure 2: The geometry, or the paths of returns, have significance for compounded returns. Contrary to popular opinion there are better geometries for compounded wealth apart from straight lines. Given that Sharpe ratios penalise beneficial volatility, we observe that Example 1 (the straighter line) has a far higher Sharpe than Example 2 but a far lower CAGR. The direction of the volatility is crucial for wealth generation which can be observed when comparing the terminal wealth of strategies with negative skew and positive skew. Moreover, the Sharpe Ratio’s implicit assumption of normally distributed returns does not hold water in the unpredictable seas of financial markets, where extreme events (often referred to as “black swan” events) are not as rare as traditional models would suggest. This discrepancy becomes even more pronounced in trend-following strategies, which, by design, aim to exploit these very outliers—large, sustained market moves—rendering the Sharpe Ratio’s insights less applicable, if not entirely moot, in evaluating such strategies. On the other hand, the Sortino Ratio, often touted as an improvement over the Sharpe Ratio, narrows its focus to downside volatility, ostensibly aligning more closely with investors’ natural aversion to losses. By considering only the negative deviations from the mean return, the Sortino Ratio aims to provide a more relevant measure of “bad” risk. While this adjustment marks a step towards a more nuanced understanding of risk, it, too, is not without its shortcomings. Specifically,