After the Report
By Patricia Horotan, Founder & Chief Strategy Officer of BlueMatrix
In a recent Financial Times essay marking the bicentenary of photography, David Campany reflects on a paradox. Photography’s repeated encounters with technological obsolescence have not diminished its cultural significance. On the contrary, they have often prompted a reconsideration of what the medium is actually for.
When photography emerged, it challenged painting’s role in representing the visible world. Painting did not disappear, but found new possibilities beyond representation. Photography itself would undergo a similar transformation as digital technologies assumed many of its practical functions, prompting artists and intellectuals to reconsider what made the medium distinctive. There is something revealing about a medium discovering new possibilities just as its established purpose becomes less indispensable.
Read the essay - Photography at 200: Why the camera still matters in the age of AI.
Investment research may be approaching a similar moment. Its medium, though, is not quite the report, even if the two have become difficult to tell apart.
For decades, the research report has been the principal expression of an analyst’s work. An entire industry has organized itself around its production, distribution and consumption. The report has become so closely identified with research that the distinction between the two is rarely examined.
Yet the value of research has never resided entirely in the information it contains. Facts do not interpret themselves. A financial statement does not explain what matters about it, nor does a valuation model establish whether its assumptions are reasonable. Somewhere between the evidence and the conclusion, an act of judgment takes place.
The industry has always understood this. Some of the most valued interactions between investors and analysts take place not through published reports, but in conversations. Investors seek time with analysts to challenge their assumptions, test the reasoning behind their published conclusions, and understand how firmly those conclusions are held. The report communicates what an analyst thinks. The conversation often reveals why.
For much of the industry’s history, this distinction presented little difficulty. The report provided a durable record of published research, while the relationship between analyst and investor allowed for further inquiry. Not everything needed to be written down. The reader could speak to the author.
But the reader is changing.
Increasingly, investment firms are using AI to consume research, synthesizing views across analysts, comparing interpretations and identifying changes in conviction. The technology can process an extraordinary volume of material, but it is largely confined to what has been made available to it. The conversations through which analysts have historically explained and defended their reasoning are not ordinarily part of that material.
An AI system may accurately reproduce an analyst’s conclusion, compare it with consensus and summarize the published rationale. But the assumptions that give the conclusion its meaning, the degree of conviction behind it, or the circumstances under which it might change may never have been fully articulated in the report.
The problem is not necessarily that the machine has misunderstood the document. It may have understood it perfectly. The document was simply never the entirety of the research.
That gap has a second consequence. When reasoning is invisible, it becomes harder to tell where an idea originated, and AI is making that harder still. Reports can be assembled, financial information summarized and plausible interpretations generated at a scale that would once have been unimaginable. As the appearance of research becomes easier to reproduce, distinguishing an original intellectual contribution from its subsequent repetitions becomes more difficult.
Human authorship alone does not settle the question. An idea is not valuable simply because a human expressed it, any more than a conclusion becomes credible because it appears in a professionally produced report. Its value depends on the reasoning that sustains it and the evidence that informs it.
Rather than concentrating exclusively on making documents more accessible to machines, we may need to consider how the reasoning behind them can be preserved and made available as well. Not every assumption, qualification or change of mind belongs in a published report, but neither should the intellectual value of those elements disappear simply because the reader is no longer human.
The report itself remains essential. Writing is part of the intellectual process, and the discipline of constructing an argument often reveals weaknesses that would otherwise remain hidden. But alongside the report, there may be room for another kind of record, one that preserves the reasoning that develops across publications and over time.
Imagine being able to follow the handful of assumptions on which a thesis depends: how confidently each is held, what evidence would weaken it, and how it has shifted across six months of publications. Such a record would not replace the report, but extend its usefulness. It would allow the reasoning behind an investment thesis to be examined independently of any single document, while preserving its relationship to the analyst who developed it.
None of this diminishes the conversation. The value of time with an analyst lies in the exchange itself: the challenge, the follow-up question, the judgment applied to circumstances no record could anticipate. A clearer record of published reasoning makes that exchange better informed, not less necessary.
It would also change what can be recognized.
Today, the familiar format of a research publication makes an original insight and its tenth restatement look much the same. Once ideas can be followed as they develop, it becomes possible to see who identified a shift first, whose assumptions held, and whose reasoning others came to adopt. For analysts, that may be the most consequential change of all: original thinking, supported by rigorous reasoning, acquiring a visibility that the document alone could never provide.
For decades, the investment research industry has built sophisticated systems to preserve and distribute what analysts write. Much of the reasoning that gives their work its value has remained elsewhere, in conversations and in the accumulated understanding of those who follow them.
As AI becomes a new reader of research, that separation becomes harder to sustain. The report will continue to matter, but it may no longer be sufficient on its own.
The history of painting and photography suggests that a practice need not abandon its familiar form for its possibilities to expand. Sometimes a new technology simply reveals dimensions of that practice that its form could never fully express.
The report has given research its form. What comes next may give its ideas a life of their own.