Roles in AI prompts: Anchoring perspective with context
In our earlier discussion, we introduced several pillars of professional AI prompting and how they transform vague requests into structured drafts. Among these, “role” stands out as particularly important as it directs AI models to research, analyse, prioritise and communicate as a professional would in specific situations. When paired with context, roles provide a powerful framework for precision and relevance in investment analysis.
Why roles matter
Assigning a role gives AI a defined professional identity. For example, asking it to act as an equity analyst versus a retail investor fundamentally changes how the model processes information and structures reasoning.
Consider a simple request: “Write an analysis of the bond market.” Without further instructions, the model will produce a broad overview, often too general for practical use.
Compare it with the following prompt: “Act as a senior fixed income portfolio manager addressing institutional clients after the Federal Reserve’s (Fed) rate cut. Explain how this move affects investment-grade (IG) spreads, high-yield (HY) valuations and duration strategies in the US. Reference past easing cycles by the Fed and highlight practical risks for institutional bondholders.”
This framing compels the AI to highlight spreads, yield curves, and credit risk, the very factors a fixed income manager would prioritise.
Context amplifies role
Roles gain power when paired with context. If role answers "who is speaking?", context answers "under what conditions and for whom?"
For example, the same portfolio manager role could generate very different outputs depending on whether the context specifies institutional investors, retail clients or an internal risk committee. Context provides the environment in which the role operates, such as market conditions, audience needs or compliance requirements. As a result, the analysis becomes both professional and relevant.
In practice, a well-crafted role prompt includes a degree of context by design. Specifying “Act as a portfolio manager addressing institutional clients after the Federal Reserve’s rate cut” aligns both the perspective and the setting and produces an analysis tailored to a defined situation.
Real AI prompt examples in action
Financial journalist:
“You are a senior financial journalist specialising in technology and business and writing for an institutional investor audience. Produce a ~300-word analysis of [Company]’s investor event on [Date], covering market-moving announcements, strategic initiatives, financial projections and market reaction. Conclude with three investor takeaways. Cite transcripts, analyst reports and market data.”
Here, the role shapes tone and perspective, while context ensures relevance to institutional investors.
Fixed income:
"Act as a chief fixed income strategist presenting to institutional investors in the US, following the Federal Reserve's latest monetary policy decision. Analyse the implications for corporate credit markets, focusing on how changing rate dynamics affect spread compression opportunities in investment-grade (IG) bonds versus high-yield (HY) securities. Discuss optimal duration positioning strategies across the yield curve, drawing parallels to previous Fed cycles to illustrate potential sector rotation patterns. Address key downside risks including credit quality deterioration and liquidity concerns in secondary markets. Provide references.”
Here, the role ensures the model speaks with a portfolio manager’s lens, while context, such as Fed policy and historical precedent, keeps the analysis focused and meaningful.
Venture capital:
“Act as a senior venture capitalist at a Tier 1 fund evaluating UK SaaS startups for Series A investment opportunities. Analyse three high-potential companies that demonstrate strong product-market fit and scalable business models. For each startup, provide: company overview with founder background and market positioning, key financial metrics (such as growth rate, revenue, gross margins and burn rate), competitive landscape analysis, go-to-market strategy assessment and critical investment risks. Provide credible references.”
By defining both role and context (growth stage, revenue thresholds, geography), the AI is guided to produce realistic due diligence rather than generic commentary.
Together, these examples show how roles and context create outputs that are credible and tailored to professional needs. Thus, instead of starting with generic drafts, you begin with draft documents already shaped around the right perspective and operating environment. This reduces revision time and improves consistency.
Avoiding Pitfalls
The most effective approach uses one precise role paired with relevant context. Vague roles ("financial expert") or missing context ("write about inflation") produce diluted outputs. Stacking multiple roles also confuses the model.
Want to develop systematic prompting approaches that deliver consistent, professional results? Contact EquityEdge Studio for effective AI prompting training for investment professionals.
EquityEdge Studio is organising a free AI prompting webinar on 30 September at 2pm BST. Sign up below:
