Weeks of full-time research per company.
Every company you choose.
Measuring whether every dollar a company reinvests generates more than the last. Testing whether competitive advantages are widening or narrowing. Assuming the investment failed three years from now — then writing the narrative of how. Each company takes full weeks of dedicated research to examine this way. That's the barrier between knowing what good analysis looks like and actually having it.
Assay Thesis
The analysis builds the strongest case against before it builds the case for — examines filings, verifies competitive claims through independent sources, and renders judgment on what the evidence supports and what it doesn’t. Any company. Any time.
· 3 deep dives per month — any company you choose
· Research library — grows with every subscriber, searchable
· Curated findings — editorially reviewed, published monthly
· Tracking Ledger — 12–19 companies tracked with current assessment, monthly
· Bellwethers — Magnificent 7 as systemic market indicators
· Thesis trajectory — how assessments evolve over time
Assay Conviction
Company-level depth, plus your portfolio as a whole. The analysis examines how your holdings interact — whether too much depends on one bet, whether multiple holdings share the same vulnerability, and the gap between your thesis and the evidence. It surfaces which positions warrant deeper investigation — and which new opportunities survive cross-examination against what you already own.
· Full analysis of every holding
· System-level portfolio analysis — shared vulnerabilities, sizing mismatches, evidence gaps
· Priority mapping — where to focus deeper investigation
· Cross-examination — new opportunities weighed against what you own
· Portfolio trajectory — how your thesis evolves over time
· 8 deep dives per month — dedicated research on any company
Deeper investigation when the thesis deserves it.
1 credit: $9 · 5 credits: $39 · 10 credits: $69
Available to subscribers of either tier.
One lazy conviction on a $100,000 portfolio costs more than an entire year of Thesis. The analysis doesn't need to be right about everything. It needs to catch one thing you wouldn't have.
Portfolio Analysis
MELI at 2.8% with rank #1 conviction is the single largest misallocation in the portfolio. A position that would not move the needle if it doubled — a company with the strongest flywheel verification, the longest demonstrated growth record, and a PEG of 0.99 — absorbing less than 3% of capital. At current weight, even a 50% gain adds only ~$4,000 to a $289K portfolio. This is the clearest example of a position too small to matter if correct. The frameworks indicate a non-speculative holding of this conviction quality warrants 15-20% allocation.
← Conviction-sizing analysis — detecting when capital allocation contradicts the conviction hierarchy
AMZN at 2.1% with rank #4 conviction presents the same structural problem. A triple-acceleration business with the most proven flywheel in corporate history, trading 20% below its peak during a sentiment-driven selloff, occupying a portfolio weight that makes even a 50% gain immaterial (~$3,000). The Q4 results showed the best quarter in years; the frameworks indicate a Core Compounder of this quality warrants 12-18%.
Combined, the five highest-conviction holdings (MELI, TSM, NTRA, AMZN, TEVA) represent only 63.8% of the portfolio, while the two medium-conviction holdings with the weakest forward returns (INSM and WWD) represent 30.9%. The portfolio’s capital is concentrated in its fifth and seventh-ranked holdings by conviction rather than in its first and fourth-ranked.
← Portfolio-level finding — the pattern individual analysis would never surface
From a portfolio analysis — unedited.
ServiceNow (NOW) — Deep Dive
Bear argument 1: AI agents will reduce seat count, destroying per-seat revenue. Partially right, but misapplied to ServiceNow. The seat-reduction thesis has force for companies like Salesforce (where AI agents might replace sales reps, reducing CRM seats) and Adobe (where AI generates content, reducing creative tool seats). That’s the self-destruct paradox confronting SaaS: the more successful your AI is, the less valuable your per-seat pricing model becomes. But ServiceNow’s value proposition is workflow orchestration, not human productivity within a workflow. If AI agents handle more tickets, they still need the platform to route, track, escalate, log, and ensure compliance. More autonomous agents means more workflow transactions, not fewer. The pricing transition to consumption/hybrid models is specifically designed to capture this dynamic. Verdict: wrong as applied to ServiceNow, partially right for broader SaaS.
← Bear-case-first — the strongest argument against is built and tested before the case for
This is not a pre-revenue startup requiring belief in undemonstrated capabilities. ServiceNow’s AI products are already selling, bookings are accelerating, and the people with the most inside information are putting personal capital at risk. The evidence exists — the market has chosen to ignore it in favor of a sector-wide narrative. This pattern — the market applying the wrong category to observable data — is precisely what the historical patterns framework identifies at inflection points for generational compounders.
← Evidence over narrative — the market applying the wrong category to observable data
Excerpt from a deep dive — unedited.
Portfolio Cross-Examination
COP — Sunk cost and loss aversion. The position at 1.3% has existed long enough to create familiarity bias (“it’s a fine company, oil prices could spike”). The ~$650 tax cost is negligible relative to portfolio size, yet it may create a psychological anchor — the feeling that paying any tax to exit means “losing.” This is a classic sunk cost dynamic. The question is not whether COP is a bad company but whether 1.3% of capital earns better forward returns elsewhere. Verdict: mild sunk cost bias detected — the analysis below adjusts by evaluating COP strictly on forward merit without reference to purchase history.
← Disposition effect check — behavioral biases identified before they can distort the analysis
The cost of NOT initiating NOW is harder to quantify due to lower confidence but potentially much larger. If the base case plays out (24.8% CAGR), not allocating 2-3% of the portfolio costs approximately 50-75 basis points annually versus the deployed alternative. Over 3 years, this compounds to 1.5-2.3% of total portfolio value — material.
← Opportunity cost of inaction — the gap between what you hold and what you’re missing, quantified
Excerpt from a portfolio cross-examination — unedited.
What this cannot do
Predict stock prices. Guarantee returns. Replace your judgment. Tell you what to buy.
What this can do
The analytical depth that takes weeks per company — examining financial filings for what they reveal and what they obscure, testing whether each dollar reinvested earns more than the last, measuring whether competitive advantages are widening or narrowing, and assuming the investment failed three years from now to find what's being overlooked — applied to every company you choose.
Every analysis ever run becomes part of the research library — searchable, growing with every subscriber. The depth compounds.
We’re onboarding founding members now.
Assay provides analytical research for informational and educational purposes only. It does not constitute investment advice. Analysis is generated using artificial intelligence, which may produce errors, inaccuracies, or incomplete conclusions. All analytical output should be independently verified. You are solely responsible for your own investment decisions.