Parallel scenario thinking in investment research | Steravindal

Why the quality of your research process matters more than you might think

There is a quiet pressure in investment research to arrive somewhere definite. When you have spent time reading reports, comparing balance sheets, and following a company or sector across several months, the natural instinct is to consolidate all of that effort into a single conclusion — a view you can act on, defend, and feel confident about. That instinct is not irrational; decisiveness has genuine value, and endlessly deferring a judgement is its own kind of failure. But there is an important distinction between a conclusion that has genuinely resolved uncertainty and one that has simply buried it. When a forecast feels confident because the uncomfortable possibilities have been quietly set aside rather than honestly examined, the confidence is cosmetic. The forecast looks tidy on the surface while concealing the real shape of the situation underneath. A more disciplined approach starts by accepting that two plausible futures can coexist in your thinking at the same time, and that holding them in parallel — rather than forcing a premature choice between them — is not a sign of indecision but of intellectual honesty.

The practical way to do this is to construct what researchers sometimes call parallel scenarios: two distinct but internally coherent accounts of how a situation might unfold, each built on a different set of assumptions about the key variables. The word "assumptions" matters here, because the most useful thing a scenario can do is make its own foundations explicit. Rather than asking which outcome is more likely in the abstract, you ask what would need to be true for each scenario to materialise. One scenario might rest on the assumption that a company's core market continues to grow at roughly its recent pace, that management executes a planned restructuring without significant disruption, and that the broader cost environment remains broadly stable. A second scenario might assume that market growth slows, that the restructuring encounters delays, and that input costs remain elevated for longer than expected. Neither of these is a prediction. Both are structured ways of organising what you know, what you are uncertain about, and what evidence would cause you to update your view. The discipline of writing them out in this way — even informally, even in a few sentences each — tends to surface assumptions you were previously making without realising it.

Where this kind of parallel thinking becomes particularly valuable is in stress-testing the logic of your own reasoning. Most analytical errors in investment research are not errors of calculation; they are errors of framing. A single-scenario approach tends to anchor you to one narrative, which then shapes which information you notice, which questions you ask, and which risks you take seriously. When you maintain two scenarios simultaneously, you are effectively obliged to steelman both. You have to ask what evidence would support the less comfortable view, and you have to be honest about whether you are genuinely weighing that evidence or merely noting it before returning to the conclusion you had already reached. This is harder than it sounds. Human reasoning has a well-documented tendency to seek confirmation rather than disconfirmation, and investment contexts — where you may have already spent time and energy on a thesis — intensify that tendency considerably. The antidote is not to pretend you have no initial view, but to build a second scenario with the same care and seriousness you brought to the first, and to track, over time, which of the two the incoming evidence is actually supporting.

The final and perhaps most practical benefit of scenario-based thinking is that it changes how you relate to uncertainty over time. A single forecast either comes true or it does not, and when it does not, there is a temptation to treat the miss as bad luck rather than as information about the quality of your process. Two honest scenarios, by contrast, give you something more useful: a map of the conditions under which each view was plausible, and a record of which assumptions held and which did not. When reality diverges from one scenario, you can ask whether it diverged because the scenario was poorly constructed or because the underlying circumstances genuinely shifted — and that distinction matters enormously for how you think about the next decision. Over time, this kind of structured reflection tends to improve the quality of your assumptions, sharpen your sense of which variables are genuinely important in a given situation, and make your thinking more robust to the inevitable surprises that markets produce. The goal is not to be right more often in some simple scorecard sense, but to understand your own reasoning well enough to learn from it.