Why investment management may become one of the highest leverage industries for AI transformation
Few industries depend more heavily on information processing, analytical throughput, continuous monitoring, and institutional research capacity. Every investment organisation is fundamentally constrained by how much high quality research its team can consistently produce, maintain, and act upon over time. Research quality influences portfolio quality. Research capacity influences scalability.
Yet despite this, investment management remains relatively early in AI adoption compared to many other industries.
Large institutions such as BlackRock, JPMorgan, Goldman Sachs, and Morgan Stanley have already committed substantial resources towards AI infrastructure, internal research tooling, monitoring systems, and knowledge platforms. These firms understand that AI is not simply another productivity tool. Over time, it may materially affect the scalability and economics of investment management itself.
Outside the largest institutions, however, adoption remains uneven. Many smaller and mid sized investment firms are still approaching the category cautiously, despite potentially having the most to gain from increased research leverage.
There are good reasons for this hesitation.
In investment research, 95% accuracy is often not good enough. A partially incorrect financial figure, an outdated filing reference, or a weakly validated research output can materially distort valuation assumptions and investment conclusions. Unlike many other industries, errors inside investment workflows compound directly into capital allocation decisions.
This creates a very different AI adoption environment from general business automation.
The question is not whether AI can produce outputs. The question is whether AI can materially expand research capacity while preserving analytical quality and trust.
At Perthshire, this became an area of deep interest because we are investors ourselves. We did not approach the problem as an external software vendor searching for a financial services use case. We approached it from inside the investment research process itself.
As we explored the space further, we began speaking with equity research analysts and investment professionals across multiple regions including Bucharest, Karachi, London, and broader international markets. The goal was simple: understand where research teams are actually spending time, where operational bottlenecks emerge, and where AI may create genuine leverage inside investment workflows.
Several themes emerged consistently throughout those conversations.
Most investment research remains significantly more manual than many people outside the industry realise. More importantly, research capacity remains one of the primary constraints on the scalability of investment firms.
The Hidden Constraint Inside Investment Management
Research bandwidth ultimately limits scalability
Most investment firms are constrained by research bandwidth long before they are constrained by ambition.
A portfolio manager can only allocate capital effectively if the organisation can continuously monitor companies deeply, process new information quickly, revisit assumptions regularly, and maintain analytical continuity across portfolios and watchlists. As coverage expands, the operational burden compounds rapidly through filings, earnings releases, industry developments, model maintenance, investment memos, and portfolio reviews.
Historically, increasing research capacity required increasing analyst headcount. More AUM generally meant broader teams, more sector specialists, larger operational support functions, and increasingly expensive research infrastructure.
This creates a structural scaling problem inside investment management.
AI may fundamentally change this equation. Not because it replaces investors, but because it may materially increase the effective capacity of each analyst and each investment team.
A well equipped analyst may eventually be able to:
- monitor substantially more companies
- process filings faster
- onboard into unfamiliar industries quicker
- maintain broader watchlists
- revisit theses more frequently
- identify second order developments earlier
At scale, this has important implications for the economics of investment firms themselves.
What We Learned From Speaking With Analysts
The workflows are still far more manual than expected
One of the more interesting findings from our discussions was how operationally fragmented many research workflows still are, even inside sophisticated environments.
Over the past several months, our Founder and Chief AI Officer, Usman Zuberi, has been speaking with research analysts and investment professionals across multiple markets and institutions around the world.
Much of the public discussion around AI in finance assumes investment firms already operate on highly structured digital research systems. In reality, many workflows remain heavily manual, highly repetitive, and deeply dependent on analyst discipline.
Sebastian, Bucharest
Sebastian, an equity research analyst based in Bucharest, described spending significant time manually transferring financial information from annual reports into Excel models and validating historical figures across filings. The work itself was not intellectually differentiated. The problem was trust.
Even where automation tools existed, outputs still required extensive validation before they could be relied upon operationally. Small inconsistencies between reporting periods or filing references could materially distort historical trend analysis and valuation assumptions.
This reinforced a recurring theme we heard repeatedly across conversations.
In investment research, partially correct is often operationally unusable.
This helps explain why many investment firms remain cautious despite recognising the broader opportunity around AI assisted research. The industry is structurally unforgiving when it comes to inaccuracies.
Laiba, Karachi
Laiba, an analyst based in Karachi, highlighted a different but equally important challenge.
Much of her research time was spent understanding unfamiliar industries, navigating fragmented regulatory material, and extracting useful information from large document sets. In several cases, AI materially accelerated the process of building contextual understanding around industries such as biotechnology, agriculture, fermentation, and specialised industrial sectors.
The value was not that AI produced final investment conclusions automatically.
The value was that it dramatically reduced the time required to understand an industry well enough for deeper research to begin. This distinction matters because one of the largest hidden costs inside investment research is onboarding into unfamiliar sectors.
Laiba also described situations involving large customs and regulatory classification documents where locating small but relevant information points required reviewing hundreds of pages manually. This highlighted another important insight from our conversations.
One of the highest leverage opportunities in investment research may simply be reducing research friction.
The firms that can process information more effectively, across broader universes, with greater continuity, may develop structural analytical advantages over time.
Three High Leverage AI Use Cases in Investment Research
Through our discussions and internal research, three use cases emerged consistently as especially important for investment firms. Not because they are technically impressive, but because they directly affect research scalability and organisational capacity.
1. Expanding Analyst Coverage Capacity
The ability to support larger AUM with leaner teams
This is likely the single most important opportunity.
Most investment organisations are capacity constrained. Analysts can only monitor a finite number of companies deeply while maintaining quality and continuity. As coverage expands, monitoring requirements become increasingly difficult to sustain consistently.
Portfolio companies continuously generate:
- annual reports
- earnings transcripts
- investor presentations
- market announcements
- management commentary
- regulatory developments
- industry news flow
AI assisted monitoring systems may allow analysts to maintain substantially broader coverage universes without proportional increases in operational burden. This matters because research capacity ultimately influences opportunity discovery, portfolio responsiveness, monitoring quality, and organisational scalability.
Historically, increasing AUM often required increasing research teams almost linearly.
AI may partially decouple that relationship.
A smaller investment team with strong AI assisted research capability may eventually support substantially larger AUM than would historically have been operationally feasible. For smaller and mid sized investment firms in particular, this could become strategically significant.
2. Accelerating Industry and Company Understanding
Reducing the onboarding cost of unfamiliar sectors
One of the largest hidden costs inside investment research is the time required to develop contextual understanding in unfamiliar industries.
Before meaningful analytical work begins, analysts often need to understand industry structures, value chains, competitive dynamics, regulatory environments, historical developments, and major market participants. This process is time intensive and often slows opportunity exploration.
AI systems appear particularly strong at accelerating contextual understanding and synthesising fragmented industry information.
Several analysts we spoke with described meaningful productivity improvements when using AI to:
- understand technical industries
- map supply chains
- summarise sector developments
- identify key market participants
- accelerate background research
This becomes especially valuable for firms with limited specialist coverage. Historically, narrow analyst capacity constrained how broadly firms could explore opportunities outside existing areas of expertise.
AI may materially reduce that limitation.
Over time, firms may be able to evaluate broader opportunity sets without needing to build large specialist research teams across every sector.
3. Institutionalising Research Knowledge
The hidden value is long term research continuity
Most investment firms contain far more knowledge than their systems are able to capture effectively.
Research often becomes fragmented across analyst notes, disconnected models, historical memos, internal documents, emails, meeting notes, and archived coverage files. As analysts move roles or priorities shift, large amounts of institutional knowledge become difficult to retrieve or reuse.
This creates hidden inefficiency across investment organisations.
AI enabled research systems may allow firms to preserve and retrieve prior investment reasoning more effectively, track thesis evolution over time, compare historical management commentary, and maintain stronger continuity across research teams.
The long term advantage may not simply come from generating more research.
It may come from compounding institutional knowledge more effectively over time.
The firms that can process more information, preserve more knowledge, and maintain broader analytical continuity may ultimately compound informational advantages over time.
The Bigger Opportunity
Why this matters at the organisational level
The most important implication of AI in investment management is not automation for its own sake. It is the potential expansion of research capacity across the organisation.
Investment management is fundamentally a research intensive industry. The firms that can process more information, maintain broader coverage, revisit assumptions faster, preserve institutional knowledge better, and monitor portfolios more consistently may ultimately develop meaningful structural advantages over time.
This is why large institutions are investing heavily already. They understand that AI may eventually reshape the operating leverage of investment firms themselves.
For smaller and mid sized firms, the implications may be even more significant. Historically, scale advantages in investment management often came from larger analyst teams and deeper research infrastructure. AI may compress part of that advantage by allowing leaner firms to operate with substantially greater research capability than their headcount alone would traditionally allow.
That does not eliminate the importance of judgement, experience, or investment discipline.
But it may materially expand what investment organisations are operationally capable of achieving.
Final Thoughts
At Perthshire, we believe investment management may become one of the most important long term AI transformation opportunities. Not because AI will replace investors, but because it may materially expand the research capacity, scalability, and operational leverage of investment firms.
As investors ourselves, we are particularly interested in how AI can help investment organisations:
- scale coverage intelligently
- improve monitoring continuity
- preserve institutional knowledge
- expand analytical capacity
- support larger AUM without proportional operational expansion
The firms that benefit most may not necessarily be those pursuing the most aggressive automation narratives. More likely, they will be firms that use AI to build stronger, more scalable research organisations.
If your investment firm is exploring AI transformation across research and portfolio workflows, speak with Perthshire’s CTO to discuss how AI may support the scalability and research capacity of your organisation.