Hi
I have attached the Polycab Report in the my first post. That is generated using perplexity.
About the prompt, there no single prompt that can get you the whole thing. Since we are using free tools, there is a lot of back and forth. You won’t get a direct report unless you have the paid version.
Ok, here is the sample prompt I use for industry analysis in Gemini which give around 25-35 pages of deep research (Use Deep Research option to get detailed report)
Based on deep research, Prepare a formal industry coverage report on the India’s Speciality chemical sector, with key focus on Specialty chemical within Chemical industy
The report must strictly use only the most recent and credible data from authoritative sources such as government publications, leading financial data providers , and reputable industry associations etc.
Structure the report as follows:
- Industry Overview:
- Summarize the current size, structure, and recent growth trends of the industry.
- Highlight key macroeconomic and regulatory factors affecting the sector.
- Growth Drivers:
- Identify and explain the main factors propelling industry growth (e.g., demographics, policy reforms, technology adoption, export trends).
- Support each driver with the latest available statistics or examples.
- Include 5 liner Poter’s 5 forces analysis of industry
- Key Players:
- List the top 5–10 companies by market share, revenue, or other relevant metrics.
- Provide a brief profile of each, including core business areas, recent performance, and market positioning.
- Distinct Advantages of Leading Participants:
- Analyze what sets the leading companies apart (e.g., scale, technology, distribution, brand equity, regulatory compliance).
- Use recent examples or case studies where possible.
- Industry Challenges & Risks:
- Outline the primary risks and challenges facing the industry, citing recent developments or data.
- Outlook:
- Provide a forward-looking perspective on the sector, referencing credible forecasts and expert opinions.
Instructions:
- Cite all data points and claims with their original sources.
- Use only the most recent data (preferably from the last 12–18 months).
- Avoid using unverified or secondary sources (e.g., blogs, forums).
- Present insights in a concise, business-professional tone.
- Reports all financials numbers in Crore
- Cross verify numerical data after completion of the report
I am also attaching the report i got using this prompt PDF for proper understing and link for formatted report
Chemical Speciality.pdf (270.0 KB)
Next step for making these report in executive summary with professinal look is using Calude AI, to summarise and layout these detailed report into professional format
Hope this helps
Disclaimer : None of above report is any recommendation of the stock in anyway, its just for educational purposes, I am invested in E-Mudhra and Aether with Long term view (very old holdings) these sample reports are for users understing on how AI can benefit in deep research with a click of a button
I extensively used perplexity for general research but for deep research I dont get detailed view specially with free version.
Hallucination in AI is another pressing isssue that refrain me in using Company level research.
My view is DRHP, Annual Report, Earning call transcripts and Credit rating agency reports are more than sufficient to get view on a company with cutting all the extenal noise and biases of the analyst. and for this Notebook LLM does the best work.
Over the months my approach is to use data above data along with Screener.in data for evaluating individual companies using notenook LLM
I have some predefined format which I am looking at for years now for checking trends in Operating statement and Key balance sheet items along with share holding pattern, for these I use Screener.in csv file to update the tables, using external link data feature in google sheet
In absence of ability to link csv file data in Notebook LLM I was doing manual work for this particular section in my report.
Yesterday only, I figured out how to convert CSV file in text format which notebook LLM recognise with tables, now I am able to use screener.in data with for financial trend and peer analsysis with confidence…
Hey, this Gemini trick worked like a charm, but I did it a little different. Perplexity Pro Deep Research is hands down the best LLM to use for Company Analysis and gathering raw data. You get loads and loads of data, unless you do not restrict the model. Once you get that Indepth Research, I shared it to Gemini and asked it to rearrange the data to a format that I had loaded beforehand. I added a few more prompts, like “Do not shorten important attributes that are critical to the analysis of the company”. Within seconds, it gave me the exact format I asked for.
Later, i fed the formatted data into Claude and gave it a context of what kind of a Research Report I am looking for. Gave some info on the layout, font and theme and asked it to create a Research Report for an Investor Meet in 30mins. The result is really good. I am sure with the paid version, it can be improved further.
aurionpro_research_report.pdf (411.9 KB)
comprehensive_nuvama_report.pdf (90.0 KB)
Can you share full report you got using Perplexity pro to get feel of the difference how these 2 AI prepare compile reports?
Actually that is the problem with Perplexity. Even though I have the paid version, it always creates apps or html files that include the code for the report. I first have to host the app on the web and then share the website link to access the report. So I cannot really share the results.
But you can try it yourself. Copy paste the exact prompt from Gemini into Perplexity and see what it creates. It is totally different than Claude.
I tried using AI to back test Investment thesis based on Fundamental parameters like High div yield and low PE, PSE stocks performance for last 6 month and 1 years to validdate my thesis and see historical performance.
It worked like a charm,
Research that would have taken hours to just get data and then shortlisting the stock based on criteria and then preare table would have taken full day, was done just with click of a button.
I am attching one such report, formatted version (all made usign AI, Data colletion, presentation, validation of criteria)
01 Recommendation.pdf (104.3 KB)
I am also attaching back testing results prepare by the AI
Investment thesis back testing.pdf (94.1 KB)
Investment thesis 6 months.pdf (67.4 KB)
I use following financial matrix which automatically comes from screener.in for further analysis of trend and ratios of the comapnies so short listed
00 Fin Info Coal India.pdf (601.9 KB)
Finally this matrix is also prepared automitically from export data fucntion of screener.in which helps to understand bird view of valuation multiples and other important data point that help me for final decision
01 Fin Info Coal India.pdf (145.9 KB)
All these automation tool and prompts are preared once, later on these are replicated on any stock. It hardly take 3-5 minutes to fetch all the reports for a stock, once template are set.
Disclaimer: Attached report(s) for coal india is only to illustrate how to use AI and other automation tool for better research and analysis, its not an investment recommendation.
I am invested in the stock, with less than 5% allocation of portfolio.
You can continuously update your tool, whenever business update and financial published by your coverage company like Maxpos
Microsoft Word - Aerohub- Garuda.pdf (587.8 KB)
ure
This is a tech company with niche products. As it is well-known, it is very difficult to do initial breakthroughs in tech products. It is more difficult for a small company. But thanks to Maxposure’s technical skill they have made breakthrough in inflight audio infotainment. And for pilot project, they have received approval from a reputed airlines like Garuda of Indonesia. This is a very big achievement.
They are also doing well in advertisement field through their subsidiary.
All these things augur well for the future of the company and has brought them visibility.
Sorry I did not understand. Can you please elaborate?
Yeah this AI feature in Screener.in looks promising, but still haven’t tried in detail..
Will post my feedback once tested.
if you want highly readable deep dives into a company (history, major milestones, founder stories etc) - do check my site https://empor.top and I’d love to hear feedback.
LINK OF PROMPT DATED 07-08-2025 ON SCRIBD.
It covers Cosmo First Ltd. The link covers both the Prompt and the report generated by CHATGPT.
I found the prompt and the report quite useful and comprehensive.
PROMPT :
You are a senior equity research analyst at Motilal Oswal Institutional Equities. You are preparing a detailed, 25-page Initiating Coverage report on Cosmo First Ltd, covering business model, growth outlook, industry dynamics, management, SWOT, financials (historical and projected), valuations, and key risks.
Objective: Generate a 25-page equity research report (professional/institutional-grade), suitable for institutional investors (mutual funds, PMS, FIIs, insurance companies).
Data Available:
• Audited FY25 financials
• Forecast required for FY26–FY28
• Current market price as of August 2025
• Industry benchmarks for valuation multiples
Write in the tone/style of: Motilal Oswal/ICICI Securities/Axis Capital/Jefferies India reports — data-driven, jargon-heavy, clear visual formatting, and institutional tone.
You are a Senior Equity Research Analyst at Motilal Oswal Institutional Equities. Write a full 25-page “Initiating Coverage” equity research report on Cosmo First Ltd, with the title: “Riding the Specialty Film Supercycle”.
Use the following structure and content guidelines:
Investment Rationale Summary (1 page)
Company Snapshot & Key Financial Metrics (1 page)
Business Model Deep-Dive (2 pages)
Industry Landscape & Outlook (2 pages)
Management & Strategy (1 page)
Competitive Positioning (1 page)
Financial Analysis: Historical Performance (2 pages)
Financial Projections FY26–FY28 (3 pages)
Valuation & Target Price Justification (2 pages)
SWOT Analysis (1 page)
Key Risks & Mitigants (1 page)
ESG Framework & Sustainability (1 page)
Technical Charts & Shareholding Pattern (1 page)
All credit for the prompt goes to its original author. Any modifications will be posted separately.
https://claude.ai/public/artifacts/c032403e-2a4b-497b-acda-d537916688a5
Above is the link of the Research Report generated by CLAUDE AI based on the above PROMPT after removing the name of the research house. That was included to ensure a certain format for the report. Removing it does not alter the report.
To me the report looks very comprehensive. However, going thru the details one may find some important information that is missing. Will be thankful if you please bring it up so that the prompt could be modified. I plan to use the prompt for reports on other shares I hold with necessary modifications in the PROMPT.
