Note: I’m brain dumping this on an airplane. It is, as usual, poorly organised and not thought through. Simply imagine a fat man typing on a cramped economy seat with coffee spilt all over my wife-beater typing away on a 1990s thinkpad. With that image in mind, the rambling below shall surpass your expectations.

Note: My ramblings are a combination of multiple websites, experiences, X posts, that I’ve read, studied, all whilst changing diapers on a lack of sleep for several months – thus this is may not be original, my comments are a combination of multiple sources and I do not remember where to provide the credit and thus, if you’re someone reading this where I’m borrowing what you’ve said and didn’t credit you – apols in advance.

Playing around with the LLMs and AI

I am absolutely loving this, it’s been absolutely wonderful to be able to create tools, programs, websites, that I have been dreaming about doing for the past 2 decades. I am no longer limited by time, capability, resources, access to people, information, money. I am only limited by my own mind.

Frankly, no one has an excuse to not be able to under the basics of any company going forward (or at least pretend to). But the future is a beautiful question mark.

Online I see the question of whether the role of fresh grad analysts will be gone?

Yes and no – I’m not sure.

It is now easier to:

• Aggregate data on a company and an industry
• To gather data to easily understand a company and an industry and what the historical trends were.
• To put together information packs, presentations, financial models etc.

All of this data comes from what is publicly available.

Every excel monkey that comes across, believes that their financial model is better than the one made by the previous excel monkey – I know I’m that monkey.

But these are just the bones – this isn’t the flesh, the brains. It misses the philosophy, culture and politics of humans.

But what about the data that isn’t public?

This is the human advantage.

I work under the assumption that when one talks to a company’s management, key team members, industry experts you are able to garner insights that are readily available in the public domain. This isn’t insider trading, but this provides an understanding of how an industry and company operate as not everything is 0 & 1’s despite what the tech overlords may say. There are still humans that impact how well or poorly a business can perform.

Now combine these insights garnered from in person meetings with what can be generated from the tools available today on an AI/LLM tool then holy moly – would that become the ultimate superpower? I think so – for now.

Humans are still going to be the difference maker.

Consider this our Augmented Intelligence period.

*cavaet – 99.9% of fresh grads won’t be able to spot this.

Applying this thought:

I generally have the viewpoint that companies and organisations held together and driven forward by 1 to 10 people at most.

Yes it is easier to pick a stock that performs yes in an industry that has a tailwind. But which one is that going to be?

In an industry that is flat-lining or declining – which individuals are going to ensure that their businesses can still remain competitive – pick up market share and then should the tides turn – be the winners or losers.

Understanding this is what is going to be provide a more complete understanding.

Here’s a random one, in 1 country, a ceramics manufacturer/distributor/brand owner is #1, growing market share etc etc #’s are decent, management is clean, shareholders are clean, but the stock price remains a dog for the past x # of years – why? It can’t grow faster than “GDP” because the Chinese ceramics are illegally imported, sold at -50% of the local market price, and not tracked on the government databases and therefore cannot be measured from the public data. This amazing well-managed company cannot grow at 2-3x GDP, because it operates in a country where the government sector are easily bribed and there are leakages everywhere. Talking with the management enough times – this information becomes known.

PoS companies will “always” be their respective fair value.

Researching names is so much more fun these days.

One thought I’ve had is that, in theory, shouldn’t this should make capital markets more efficient?
If companies are consistently analysed with these datapoints, should good companies that consistently perform well be trading at fair value, or a premium to their peers? Yes that’s already the case, but by how much more could the best performers trade above their peers? 1x? 2x? 10x? 100x? And by trading at these higher valuations, wouldn’t the best players then continually issue raise capital to capture more market share from their competitors? And thus applying this reflexive theory wouldn’t the top companies just continue to trade higher and higher?

Shouldn’t PoS (if you are new to me, this is a industry term used by highly specialised investment professionals, “Piece of Shit”) companies be constantly trading at a discount?

Would this theory apply to all capital markets or only those where 1) there is information readily available to the public 2) where capital can flow freely

stage 1) information flow
This would rely upon company disclosures
easy access to industry information i.e. how many tennis shoes are sold daily

stage 2) capital flowing freely
if capital can flow freely in/out of a country i.e. the US, HK, Singapore then it’ll be easier for companies consistently performing on a “fundamental” basis to be consistently bid higher to be trading at a premium to the “market” and to its peers

If capital cannot flow freely in/out of that said country then perhaps the potential valuation of companies that perform will be limited to the capital pool available in the country.

This is already the case – but they how much more will countries that limit capital flow be at a discount to peers that do allow free capital flow with minimal friction.

If the best in the world are eating market share of regional/local players – how is this going to be priced?

Yet to be thought through

Speed

Will this become more important? Looking through history and personal experiences, crisis occur faster & capital markets crash and recover faster. Even in smaller markets like Thailand – yes Thailand is tiny – accept it, what Thailand trades in a day is what 1 tech stock trades in 30 minutes.

I am curious what it will look like when AI trades against AI in the future, we’re seeing this with the HFTs over the past decade in the US and more recently in Thailand over the past few years. Perhaps it only exacerbates the lows & highs, the lows will be amusing when liquidity & confidence disappears.

Countries/regions

Can a country be easily modelled?

We have the usual GDP data….but as we all say, GDP growth doesn’t mean the stock market will perform well and vice versa…anyone in the markets for long enough recognises that the market prices in what GDP should be over the next 3-6 months.

And does GDP full encompass an economy?

Thailand? Everyone talks about the black economy being ~50% of reported GDP.

How do you measure this? Track this?

Taxes, imports and exports, logistics, the bribery culture – how is this going to be measured? Can it be quantified? This goes back to the comment that these new tools are again just the bones and is missing the flesh & brains of the body.

I’ve been prompting a little bit but wonder if it’s futile – or my imagination just isn’t there yet.

Have fun nerds

Ok here’s some highschool prompts’ for you to play with (again I don’t remember where I copied this from, I’ve edited some parts but thanks to whomever the original source is/are)

Researching is a function of how well one prompts now. It’s stupidly easy to produce a relatively decent report nowadays, but it will be predicated upon the quality of ones question (see at the end of the post)

Short form analysis
ROLE:

Act as an elite equity research analyst at a top-tier investment fund.
Your task is to analyze a company using both fundamental and macroeconomic perspectives. Structure your response according to the framework below.

Input Section (Fill this in)

Stock Ticker / Company Name: [Add name if you want specific analysis]
Investment Thesis: [Add input here]
Goal: [Add the goal here]

Instructions:

Use the following structure to deliver a clear, well-reasoned equity research report:

1. Fundamental Analysis

• Analyze revenue growth, gross & net margin trends, free cash flow
• Compare valuation metrics vs sector peers (P/E, EV/EBITDA, etc.)
• Review insider ownership and recent insider trades

1. Thesis Validation

• Present 3 arguments supporting the thesis
• Highlight 2 counter-arguments or key risks
• Provide a final verdict: Bullish / Bearish / Neutral with justification

1. Sector & Macro View

• Give a short sector overview
• Outline relevant macroeconomic trends
• Explain company’s competitive positioning

1. Catalyst Watch

• List upcoming events (earnings, product launches, regulation, etc.)
• Identify both short-term and long-term catalysts

1. Investment Summary

• 5-bullet investment thesis summary
• Final recommendation: Buy / Hold / Sell
• Confidence level (High / Medium / Low)
• Expected timeframe (e.g. 6–12 months)
  Formatting Requirements
• Use markdown
• Use bullet points where appropriate
• Be concise, professional, and insight-driven
• Do not explain your process just deliver the analysis

Then you get one set of results

Long form analysis
“You are an equity research analyst. Produce a rigorous, source-backed investment memo on xxxx with a clear Buy, Hold, or Sell call.
Rules for research and writing

1. Use only verifiable, recent sources. Prioritize official filings, earnings materials, investor presentations, regulatory documents, reputable industry data, and high quality media. Cite every non-obvious fact with a link and date.
2. Separate facts from interpretation. Tag each paragraph as Fact, Analysis, or Inference.
3. Use precise dates. Avoid vague time references.
4. Quantify claims. Show math for derived metrics. Use tables where helpful.
5. Note uncertainty. Call out missing data and state assumptions.
   Deliverables
   A) Executive summary (8 to 12 bullets): snapshot, thesis, rating, price targets and time frames, key drivers, key risks, near-term catalysts, and what would change the call.
   B) Full memo with sections 1 through 15 below.
   C) Appendix: source list with links and dates, data tables, and a simple operating model.
6. Thesis framing (purpose: define what must be true to create value)
   State the core investment question in one sentence.
   List 3 to 5 thesis pillars that would make the stock attractive.
   List disconfirming evidence to test that could break the thesis.
7. Market structure and size (purpose: size the prize and trajectory)
   Quantify TAM, SAM, SOM. Segment by product line, customer size, industry, and geography.
   Identify growth drivers: regulation, replacement cycles, macro activity, technology adoption.
   Estimate current penetration and runway. Compare against peer adoption curves.
8. Customer segments and jobs to be done (purpose: map who buys and why)
   Describe mix by size band and industry. Identify buyer roles and budget owners.
   Detail core workflows and pain points. Explain mission criticality.
   Assess switching costs and vendor lock-in by segment.
9. Product and roadmap (purpose: evaluate product-market fit and durability)
   Summarize core modules and adjacent products. Call out differentiators.
   Compare depth vs breadth versus best point solutions.
   Explain implementation time, integrations, configurability, and typical time to value.
   Provide quality and reliability signals: uptime, incident history, mobile performance.
   Roadmap credibility: stated milestones versus delivery track record.
10. Competitive landscape (purpose: position the company)
    Identify direct and indirect competitors by segment and size.
    Compare pricing, packaging, and feature gaps. Include switching friction and contract terms.
    Summarize win or loss reasons from reviews, case studies, and disclosed data.
11. Go-to-market and distribution (purpose: test scalability of new-logo engine)
    Break down demand sources: inbound, outbound, partner referrals, marketplaces.
    Sales productivity: ramp, quota attainment, conversion rates where disclosed or inferred.
    Role of channels and partnerships: integrations, OEMs, platforms.
    Services and customer success model. Training and community as moat.
12. Retention and expansion (purpose: quantify durability of revenue)
    Report gross and net dollar retention by cohort and segment if disclosed or estimable.
    Explain logo churn drivers and timing. Provide a churn curve if possible.
    Identify expansion vectors: seat growth, module attach, usage-based add-ons.
    Discuss contract length, renewal mechanics, and price increase policies.
    Include reference-call insights or credible review synthesis.
13. Monetization and embedded finance if applicable (purpose: understand usage economics)
    Revenue streams and pricing model. For payments or fintech: share of customers active, GTV penetration, take rate by tender type, blended margin, cost stack, fraud exposure, and who holds credit risk.
    Revenue recognition: gross vs net. Seasonality and cyclicality.
    ARPU uplift from usage products. Payback on onboarding.
14. Unit economics and efficiency (purpose: test scalability with profitable growth)
    CAC, payback period, magic number, LTV to CAC by segment if available or estimable.
    Contribution margin by line: software vs usage vs services.
    Cohort profitability and cash contribution over time.
    Implementation and support cost over customer lifetime.
15. Financial profile (purpose: link operations to financial outcomes)
    Revenue mix and growth by component. Gross margin by line. Operating leverage path.
    Rule of 40 and efficiency trends. GAAP to cash flow bridge.
    Leading indicators: billings, RPO, backlog.
    SBC, dilution, and share count trajectory.
    Liquidity, working capital needs, and path to FCF breakeven and target margin.
16. Moat and data advantage (purpose: assess defensibility)
    Workflow depth and data lock-in. Network or ecosystem effects if present.
    AI or analytics differentiation with measurable outcomes.
    Integration footprint and practical switching costs.
17. Execution quality and organization (purpose: evaluate management and operating cadence)
    Leadership track record and stability. Org design and succession.
    Engineering velocity: release cadence, defect and incident rates where available.
    Customer sentiment: CSAT, NPS, peer review sites, and community signals.
18. Risk inventory and mitigants (purpose: make downside explicit)
    Macro, regulatory, competitive, operational, and concentration risks.
    Payments, credit, or compliance risks if relevant.
    Implementation complexity and time-to-value risks.
    For each risk, propose leading indicators and mitigations.
19. Valuation framework (purpose: value with cross-checks)
    Public comps table: growth, gross margin, operating margin, Rule of 40, EV to revenue, EV to gross profit. Normalize for any usage or payments reporting differences.
    DCF with explicit drivers and sensitivity bands.
    Cross-checks: cohort NPV math, S-curve adoption, unit economics to enterprise value sanity checks.
20. Scenarios, catalysts, and monitoring plan (purpose: set expectations and triggers)
    12 to 24 month bear, base, bull cases. Specify NRR, new logos, pricing or take rate, margins, SBC, and share count. Assign probabilities that sum to 100 percent.
    Near-term catalysts: product launches, pricing changes, partnerships, market entries, M&A, regulatory outcomes.
    Early warning indicators: churn spikes in small cohorts, backlog slippage, uptime incidents, pricing pushback.
    What would change my mind: three positive and three negative triggers.
    Output format
    Executive summary
    Rating with price targets and time frames
    Investment thesis and variant perception
    Detailed sections 1 through 15
    Tables and charts embedded
    Source list with links and dates
    Appendix with model assumptions and calculations
    Quality bar
    No generic claims. Back important statements with numbers and citations.
    Label any speculation as Inference.
    Be concise and structured. Prefer bullets and tables.

“

You’ll get another completely different report.

And then depending upon what you use, perplexity, claude, gemini, grok, wokegpt, etc etc and what level you’re paying for you’ll get a different set of results.

Note: Use Deep Research mode.

Note: I do use different prompts from the above – for now it’s my Human IP and I’m keeping it to myself for now….it’s remarkable.

Note: To get in further into finance nerd mode – if you then need to make a presentation for whatever half-a$$ed Investment Committee members that pay no real attention to investment analysis nor ask decent questions do the following:

1) Copy and paste the result of the longer prompt into Notebook LM

2) Generate the presentation – boom presentation generated! But damn it, there’s that damn watermark from NotebookLM, want to remove the notebook lm logo?

3) Go to canva – put in a simple instruction, it’ll remove the watermark on each slide, WOOHOO!

4) Voila – You have a full company pack within 30 minutes from scratch.

5) Buy me a somtum waffle below

I advise that people get used to being a nerd again. Get used to working with .md’s and .txt’s again. Get used to use cmd on your windows pc or Linux (i’m on both) it works far faster compared to working on m’soft, google which are slow in comparison. The amount of data you can get through outside of big tech walls is amazing…

Enjoy it’s an amazing period for us, if you’re not continually playing around with this you’re missing out having Augmented Intelligence.

Note – I believe that I’m heavily influenced by Readmultiplex.com and the owner of it Brian R is all over X.com – have a look.

If this has helped you, please consider donating and buy me a Somtum Waffle 🙏🧇