Why Process Beats Prediction in Investing

Why Process Beats Prediction in Investing Every week, markets produce a fresh set of forecasts: where rates are heading, which currency will strengthen, what the next quarter holds. Most of those forecasts will be quietly wrong, replaced by new ones before anyone checks. Professional investors noticed this a long time ago, and they drew a conclusion that still separates them from the crowd: since prediction cannot be relied upon, the reliable thing has to be built elsewhere. It gets built in the investment process, the set of rules that defines how decisions are made, how risk is controlled, and how a portfolio adapts when conditions change. Prediction changes every week. Process remains. This article explains what that actually means in practice. The prediction trap The appeal of prediction is obvious: if you knew what markets would do, everything else would be easy. The problem is equally obvious once stated. Markets price in available information almost immediately, the variables that move them interact in ways no model fully captures, and the events that matter most are precisely the ones nobody forecast. This is not a temporary limitation waiting for better analytics. It is a structural feature of markets. The trap is not in making forecasts, which every investor implicitly does. The trap is in building a portfolio that only works if the forecast is right. A position sized for a confident prediction, without a defined response for the scenario where the prediction fails, is not a strategy. It is a bet with a story attached. An investment process starts from the opposite premise: the future is uncertain, and the portfolio has to work across several futures at once. That single shift changes every downstream decision. The evidence, for once, is unambiguous This is not a philosophical preference. Few questions in finance have been measured as thoroughly as whether professional forecasting ability translates into results, and the answer is consistently uncomfortable for the prediction camp. The most systematic measurement comes from the SPIVA scorecards published by S&P Dow Jones Indices, which have compared actively managed funds against their benchmarks for a quarter of a century. Over the 15-year period ending in 2024, roughly nine out of ten active US large-cap equity funds underperformed the S&P 500, and there was not a single US equity category in which a majority of active managers beat their benchmark. These are professionals with research teams, information advantages, and every incentive to be right, and as a group their market views subtracted value after costs. The persistence data is, if anything, harsher. Funds that do outperform in one period rarely keep doing so in the next, which is exactly the pattern you would expect if short-term outperformance were dominated by luck rather than repeatable forecasting skill. Decades of academic work on expert judgment point the same way: confident long-range predictions about markets and economies perform barely better than chance, while the forecasters’ confidence remains untouched by their record. None of this means markets cannot be analyzed or that all active management is futile. It means something more specific: returns that depend on being right about the future are built on the least reliable input available. Whatever edge a professional operation has, it has to live somewhere else, in structure, in risk control, in execution, in discipline. In process. What an investment process actually contains Process is one of those words that gets used vaguely, so it is worth being concrete. A real investment process answers, in writing and in advance, at least four questions. How are decisions made? Entries, exits, and position sizes follow defined criteria rather than conviction of the moment. Speed without structure produces inconsistent outcomes; a plan does not remove uncertainty, but it removes the emotional improvisation that uncertainty otherwise triggers. How is risk controlled? Exposure limits, concentration limits, and drawdown thresholds exist before the positions do, with defined responses when they are reached. We described this architecture in detail in our article on the structure behind a professional forex fund, and the logic applies to any strategy. How does the portfolio adapt? Conditions change, and a process specifies how change is detected and what adjustment follows: scheduled reviews, defined triggers, deliberate procedure. Adaptation on a schedule is a strength; adaptation under pressure is usually damage. How is the process itself reviewed? Even good rules age. A serious process includes a procedure for changing the process, calmly and with evidence, never in the middle of a drawdown. None of this is exciting, and that is rather the point. Excitement in portfolio management is a cost, not a feature. Drawdowns: where process proves itself If you want to know whether an investment process is real, look at how it treats losses. Returns attract attention, but drawdowns determine survival, and the distinction matters more than most performance discussions acknowledge. A strategy that compounds well over a decade is rarely the one with the most spectacular months. It is the one whose losing periods stayed shallow enough that recovery never required heroics and never forced a deviation from the rules. Deep drawdowns do their damage twice: once in capital, and again in behavior, because they push investors toward exactly the improvised decisions the process was built to prevent. Risk management is, in practice, drawdown management. The path matters more than the peak. This is also where prediction-driven investing fails most visibly. The forecaster who is right four times and then badly wrong once can end up behind the process-driven investor who was never spectacularly right about anything. Compounding rewards the absence of disasters more than the presence of brilliance. Risk is about outcomes, not fluctuations Underneath the process view sits a different definition of risk. Day-to-day volatility is what gets measured, because it is easy to measure. But risk, properly understood, is uncertainty that affects your ability to reach long-term goals: the possibility of losses too deep to recover from, of illiquidity at the wrong moment, of a portfolio that forces bad decisions under stress. The

From ESG to AI: Hype Cycle or Structural Shift in Investing?

From ESG to AI: Hype Cycle or Structural Shift in Investing? Financial markets have always been fertile ground for narratives. Over the years, entire investment frameworks have risen, peaked, and faded—often driven as much by storytelling as by substance. Few examples illustrate this better than the recent trajectory of ESG investing. A few years ago, ESG was everywhere. Asset managers, funds, and advisory firms rushed to demonstrate alignment with environmental, social, and governance principles. New products were launched, reporting frameworks multiplied, and ESG quickly became a commercial and marketing standard. Then, almost as quickly, the momentum faded. The Rise and Cooling of ESG Today, much of the ESG hype has dissipated. Many ESG-labelled products have been rebranded, consolidated, or quietly discontinued. Investors have shifted their focus, becoming more selective and increasingly sceptical of surface-level claims that lack measurable impact. This evolution does not mean sustainability has lost relevance. Rather, it highlights a familiar pattern in finance: when a concept becomes primarily a narrative tool instead of an operational discipline, disillusionment follows. AI Takes Centre Stage Now, a new theme dominates the conversation: Artificial Intelligence. From asset managers to analysts and technology providers, AI is being embraced across the investment industry. The enthusiasm is unmistakable. Yet this raises a critical question: is AI simply the next ESG—another hype cycle destined to fade? Where AI Is Already Changing the Game Unlike ESG narratives, AI is already delivering tangible applications—particularly in trading and, even more so, in the foreign exchange market. Machine learning models are increasingly used to: Optimize signal detection across complex market environments Adapt execution strategies dynamically Manage risk exposure in real time Process vast volumes of macroeconomic data, news flow, and central bank communications Some investment strategies now rely on AI-driven systems to interpret market sentiment and anticipate currency movements with a speed and depth that traditional models cannot replicate. Why AI Is Not a Shortcut That said, AI is not magic. Its effectiveness depends on data quality, model governance, and disciplined human oversight. Without these elements, AI risks becoming little more than a sophisticated buzzword—much like ESG did at its peak. Technology alone does not eliminate risk. It reshapes how risk is identified, measured, and managed. The Key Difference Between ESG and AI The crucial distinction lies in utility. While ESG often struggled to move beyond narrative alignment, AI offers concrete tools that directly influence decision-making processes. It enhances speed, consistency, and analytical depth—but only when embedded within a robust investment framework. Still, it is too early to declare AI a definitive structural shift. Finance has a long history of turning innovation into storytelling cycles: enthusiasm, saturation, disillusionment, and eventual correction. A Measured Perspective AI may indeed reshape how investments are managed—but only if applied with discipline, transparency, and accountability. Otherwise, it risks following the same arc as previous trends. In investing, technology should serve process—not replace judgment. Understanding this distinction is what separates durable innovation from temporary hype. This article is based on a recent market commentary originally published on LinkedIn. 👉 Read the original LinkedIn post here Paolo Volpicelli INCOME CAPITAL MANAGEMENT s.r.o.