N1 · Crypto Research & Investment
Collect your own data, build your own valuations, and deliver sector research others can verify. This track teaches methods rather than trading advice.
This track serves people seeking careers in research or investment.
Its scope is explicit: it does not recommend purchases or positions. It teaches forming judgments independently, understanding their unreliability, and producing work others can verify. Whether and how much to buy remain your decision and consequences.
Who This Track Serves
Look at where you are stuck rather than your job title. Any of these four situations qualifies.
You can read research but cannot write it. Others' arguments seem reasonable, but your own first paragraph stalls: which number starts the analysis, and when is it complete?
You write research consisting entirely of retelling. Structure and citations are complete, yet you collected no numbers yourself. Asked where a figure came from, you return to another article.
You make decisions but cannot separate judgment from luck. You cannot explain gains and blame losses on the market. You cannot learn or improve your method without a written, falsifiable judgment.
You want a research or investment role but have no testable work. “Following DeFi for three years” on a résumé means little without something others can open, reproduce, and challenge.
This track cannot supply an answer to what you should buy now.
Prerequisites
Complete nine Part A chapters and genuinely perform two of their Labs first.
| Required chapter | Role here | Consequence of omission |
|---|---|---|
| Chapter 15 · Liquidity | Distinguish market capitalization from recoverable money | Valuation has no realizable exit |
| Chapter 17 · Sources of Yield | Ask who pays the return | Subsidies become a business model |
| Chapter 19 · Token Value | Supply, unlock, and dilution calculations | Undefined valuation metrics |
| Chapter 20 · Tokenomics | Check the value-capture connection | Model a token that captures no value |
| Chapter 21 · Cycles | Liquidity, leverage, and narrative feedback | Confuse cycle position with project quality |
| Chapter 23 · Onchain Data | Six measurement questions and cross-checking | Every figure is secondhand |
| Chapter 24 · Collapse and Attacks | Ten risks and explicit nonassessment | Generic risk disclaimers |
| Chapter 25 · A Thirty-Minute Project Review | Thirteen cells and source grades | Inconsistent comparisons |
| Chapter 26 · Protocol Revenue | Five income-statement lines and realized take rates | Confuse scale with profitability |
Why these nine? Each report section rests on one: data on 23, valuation on 19–20, business models on 17 and 26, risks on 24, comparisons on 25, and macro context on 15 and 21. Skipping a chapter leaves a section without foundations.
Also read Chapter 31 · AI and Crypto Research. Its evidence grades and conclusion cards provide this track's recording method.
Actually complete two Labs: Chapter 23's first onchain data chart and Chapter 25's thirteen-cell review. Without them, later work remains theoretical.
What You Will Learn
Nine topics form six groups. The order is deliberate: examine the water, then the ship, and only then price.
- Macro and liquidity
- Fundamentals and onchain data
- Valuation
- Sectors
- Portfolio and risk
- Research writing
| Group | Topics | Problem addressed | Connecting chapters |
|---|---|---|---|
| Macro and liquidity | Macro, Liquidity | Current environment and capital flows | 13, 15, 21 |
| Fundamentals and onchain data | Fundamental Research, Onchain Research | Actual protocol activity and independent verification | 23, 26 |
| Valuation | Token Valuation | Value under assumptions and their weakest link | 19, 20 |
| Sectors | Sector Research | Why this project among peers | 25 |
| Portfolio and risk | Portfolio, Risk Management | Avoid failure for unrelated reasons despite a correct judgment | 18, 24 |
| Research writing | Research Writing | Work others can verify, challenge, and use | 31 |
Three Difficult Tasks
Valuation's goal extends beyond calculating a number. Combine cash flows, take rates, supply changes, and discount assumptions into an adjustable table. Develop three scenarios, identify the most sensitive input, and specify when the entire model fails. A useful model identifies the most costly mistaken assumption, rather than merely outputting one number.
Portfolio risk moves from individual assets to the whole portfolio. Chapter 24's ten categories concern protocols. Here ask whether several holdings fail in one event, whether mistakes have bounded losses, and what exit depth remains in the worst market. Chapter 15's liquidity question becomes personal.
Research writing is underestimated. Chapter 31 excludes unchecked model claims from conclusions. Extend that into delivery rules: a card for every conclusion, a definition for every figure, and explicit uncertainty at each section's end. The report becomes shorter and usable by others for the first time.
Define a sector's boundaries and survival conditions, then compare projects in a common table using consistent measurements, rather than describing them individually.
Single-project research asks whether a project is good; sector research asks what survives within the structure. The first judgment fluctuates with sentiment; the second changes when structure changes.
Defining boundaries takes the most time. State excluded projects and the reasons in the report. Exclusion criteria demonstrate your understanding of the sector.
Grade the basis of every conclusion. Chapter 31 defines verified, consistent secondary evidence, model output, and inference.
This is a required delivery rule. An ungraded report is worth only its weakest claim, which readers cannot identify.
Shorter reports are a sign that grading works.
Final Deliverable
A sector research report ready to submit to a fund, including a valuation model and risk list.
“Ready to submit” means a professional reader can assess its reliability without asking you questions.
Distinguish it from the Capstone Project. Crypto Research OS studies one project vertically to connect the 36 core chapters. This track studies one sector horizontally to demonstrate professional direction, adding comparable measurements, adjustable valuation, and conclusions addressed to critical readers.
Five testable acceptance criteria:
Define sector boundaries and publish exclusions.
State inclusions, exclusions, and reasons for investigated projects you omitted. An inclusion list alone cannot explain sample selection.
Every number includes definition, collection date, and source; onchain work includes reproducible queries.
Put Chapter 23's six measurement answers in an appendix. Running your query should reproduce the magnitude. Remove unreproducible figures or attribute them explicitly to a source.
Provide an adjustable valuation table, three scenarios, and a sensitivity conclusion.
Include pessimistic, base, and optimistic assumptions; identify the most influential parameter. Separately state when the entire model is inapplicable.
Assess all ten risk categories; mark unassessed items.
Use the Analysis Toolkit's common risk model. Explain why any category cannot be assessed. Ten reassuring answers usually indicate superficial assessment.
Grade every conclusion and exclude unchecked content from conclusions.
End with a self-check: what remains after removing unchecked content? If little remains, you performed editing rather than research.
The report must contain no trading instructions. Deliver structured judgment and evidence, rather than an action.
Assess Your Readiness
Answer each yourself before opening the comparison.
Ready: identify the exact chapter, page, or contract address, collection date, and rerun the query immediately.
Not ready: only another secondary article, a remembered conclusion without source, or a link that lacks the claim.
This is the track's foundation. Chapter 31's second researcher had many correct conclusions without knowing which. Even highly accurate research is unusable when claims cannot be located.
Ready: name the parameter immediately, its directional effect, and the range where conclusions reverse.
Not ready: a result without knowledge of its sensitivity. The problem is unjustified certainty, rather than insufficient precision.
Change each parameter by 10% and inspect rankings. If every small change reverses rankings, your conclusion reports parameter choices rather than research.
Ready: labels concentrate around money and permissions, distinguishing pending investigation, undisclosed information, and unresolved explanation as Chapter 25 taught.
Zero labels almost certainly means speculation filled gaps. Chapter 25's completely filled report did not necessarily falsify information; it failed to distinguish facts from inference, preventing readers from discounting claims.
Serious sector research often has many unknowns; their distribution is itself a conclusion.
Ready: identify a definition, assumption, omitted chain, or unverifiable offchain fact, and the sections affected.
This is the conclusion card's “If wrong” row. Some errors are minor; others invalidate the whole report. Treat them differently.
Inability to answer suggests you have not dissected your reasoning. A researcher unable to explain likely errors cannot explain why they are right.
No standard answer; check:
- Did you present the strongest opposing argument voluntarily?
- Have you graded the assumption opponents will attack, and at what level?
- Could sector boundaries look selected to force your conclusion? Are exclusions explained?
- Have you identified a risk that would invalidate your main conclusion?
- How many assumptions come from industry consensus without independent verification?
Ask an opposing reader only where they first disagree. That passage usually has your weakest evidence.
Common Mistakes
Mistake one: collecting conclusions as research.
Taking the intersection of ten reports produces consensus rather than fact. When consensus fails, everyone fails together and you have no independent judgment.
Use secondary material for identifying investigations and cross-checking independently collected data. Chapter 23 emphasizes explaining discrepancies rather than celebrating identical numbers.
Mistake two: valuing precision over usefulness.
Three-decimal discount rates and quarterly five-year cash flows refine assumptions without making them truer.
Simplify the model, expose assumptions, and focus on sensitivity. Five sourced parameters with ranges are more useful than thirty elaborate rows because others can substitute an assumption and recalculate.
Mistake three: treating risk management as asset selection.
“Good projects mean controlled risk” conflates whether you are right and what happens if you are wrong.
Research answers the first; risk management answers the second: loss limits, underestimated correlated mistakes from shared assumptions, and worst-market exit capacity. Liquidity disappears when needed most, making the last question crucial here.
Mistake four: studying survivors only.
Failed projects quickly vanish from discussion, biasing material and inferred success factors toward survivors.
Include failures deliberately. Explain sector deaths, mechanisms, and advance signals in at least one section. Apply Chapter 24's ten-risk retrospective method. This is often the most informative and least commonly written section.
Next Steps
Complete three foundations: Chapter 23's data Lab, Chapter 25's thirteen-cell Lab, and Chapter 31's conclusion cards.
Complete the Capstone Project's Crypto Research OS first. Deeply study one project before expanding across a sector; reversing this order sacrifices depth.
Maintain two ongoing practices: source ranking and monitoring from Staying Current, and model proposals with human verification from AI's Role in Crypto OS. Research expires; these practices keep it current.
Compatible tracks:
| Combination | Why it fits |
|---|---|
| N5 · Crypto Founder | Research ends by judging whether a project deserves to exist; founding begins there. Many questions overlap |
| N4 · Crypto BD & Ecosystem | Research maps resource networks; BD connects them. BD experience improves discounting of partnership announcements |
| N3 · Crypto Growth & Operations | Retention and subsidy-removal measurements test genuine demand in research |
For collecting data without third-party platforms, T3 · Onchain Data Engineer extends naturally: its replayable indexing pipeline repairs research's weakest link.