I only buy a company if it's creating a brand new category, growing fast, hard to replace, and small enough that it can still 10x.
Most investors pick stocks based on price movements, tips, or vibes. My framework rejects about 87% of companies that look interesting on the surface. When something passes all four filters — which happens maybe once or twice a year — I invest with conviction and hold for a decade.
The goal isn't to be right often. It's to be very right when it matters.
Imagine you're back in 2005 and someone tells you about this small company called Salesforce. Before Salesforce, companies tracked their customers in spreadsheets and sticky notes. Salesforce didn't just make a better spreadsheet — they invented an entirely new category called CRM software, put it in the cloud, and made themselves the only name anyone thought of when someone said "customer relationship management."
I ask four questions about every company, and all four have to pass:
Not "are they a good company in an existing market." Did they create the problem definition itself? Rubrik didn't just make better backup software — they invented "Zero Trust Data Security" as a concept. They own the language. That's worth a lot.
Revenue growing 30%+ per year, margins above 60%, path to making money. This filters out story stocks with no substance behind them.
If a company's product is deeply embedded in how a business runs — years of data stored there, employees trained on it, everything connected to it — customers don't leave. That stickiness is the moat that protects the investment for a decade.
This is the brutal math filter. Great company does not equal great investment. If a company is already worth $80 billion, it needs to reach $800 billion to 10x. Very few companies ever get there. So I only invest when the company is still small enough that the math is actually possible.
A true opportunity that passes all four filters appears roughly once or twice a year. That's why this portfolio is concentrated, not diversified.
On the radar:
| Position | Entry Date | Return |
|---|---|---|
| DDOG Datadog · 53 shares |
Feb 9, 2026 | +136.73% |
| RBRK Rubrik · 132 shares |
Nov 24, 2025 | +48.89% |
| Total | — | +88.18% |
What happened: DDOG was purchased on February 9, 2026 — the day before Q4 2025 earnings. The stock moved +16% intraday the following day on a strong print. This looks good in hindsight. It was still wrong by the framework.
Why it was wrong: The framework has no earnings-timing component. Buying ahead of a known catalyst is event-driven speculation layered on top of a thesis — it introduces a short-term price dependency into a long-term conviction. If DDOG belongs in the portfolio, the earnings date is irrelevant to that decision. The correct process is to buy when the framework and valuation align, then hold through all earnings events without reacting to them.
The lesson: A track record that only documents clean decisions isn't honest documentation. This deviation is logged here permanently because the framework worked — the entry was right — but the process was wrong. Those are not the same thing.
Positions are held through price volatility. Positions are sold when the investment thesis breaks. Those are different things.
Position drops 60% from entry price. Capital preservation rule. No debate, no re-evaluation — sell immediately.
Position drops 30% from entry. Full four-filter framework re-score required within 48 hours. If any single filter scores below 7.0, sell. If score holds above 8.0 and thesis is intact, hold.
Company loses category leadership. Revenue growth falls below 20% for two consecutive quarters. A competitor displaces them as the default name in the category. Any of these triggers a sell regardless of price.
Price volatility alone. Market-wide selloffs. Short-term underperformance versus the S&P 500. Analyst downgrades unaccompanied by thesis change.
How returns are calculated. Time-weighted return (TWR), which removes the distortion caused by adding new capital mid-period and allows apples-to-apples comparison against the S&P 500 benchmark. Benchmark is SPY total return with dividends reinvested over the identical period. All figures are derived from personal Charles Schwab brokerage statements. Pre-tax returns shown. Monthly statements are maintained and available upon reasonable request.
Path to independent verification. Currently self-reported (2025–2027). A CPA verification letter confirming return calculations against brokerage statements is planned for 2028, estimated cost ~$2,000. Full GIPS-compliant audit to follow if and when seeking to manage capital for others — required for Series 65 / RIA registration.
I'm a Software Product Manager with 10+ years building enterprise software — the same category of products this portfolio invests in. When a company claims to be "creating a new category," I've been through enough vendor evaluations and product cycles to tell the difference between a genuine category shift and a good marketing story.
This is a personal account, not a fund. The long-term goal is to manage capital for others — with a Series 65 license and an RIA practice — and this public, time-stamped track record is the foundation for that. The uncomfortable months stay on the record. The mistakes stay on the record. That's the point.