Portfolio Construction: Building a Software Portfolio
Knowing how to value individual software stocks is necessary but not sufficient. The way you combine these positions into a coherent portfolio—managing concentration, correlation, and risk—determines whether your returns are exceptional or merely average.
The Barbell Strategy
The most effective framework for software investing is the Barbell Strategy: construct a portfolio that combines a concentrated core of high-conviction compounders with a diversified set of asymmetric bets.
The Core (60-70% of Capital) — Compounders: Allocate the majority of your capital to 5-8 positions in companies with proven generational compounder characteristics (NRR >120%, Rule of 40 >50, category leadership). These positions are sized at 8-15% each. You are buying these to hold for 3-5+ years, through volatility.
The key discipline: Do not sell a compounder because it "got expensive." If CrowdStrike trades at 15x revenue and the NRR is 130%, the "expensive" valuation is the market's rational acknowledgment of the compounding engine. Selling because of multiple anxiety and trying to buy back cheaper is a timing game you will lose more often than you win.
The Edge (20-30% of Capital) — Asymmetric Bets: Allocate a smaller portion to 3-5 positions in "emerging compounder" candidates—companies with strong early signals (accelerating growth, NRR inflecting upward, new product launches) but not yet proven at scale. These are sized at 3-6% each. The goal is to identify the next CrowdStrike or Datadog before the market consensus does.
Cash / Dry Powder (5-10%): Always maintain a cash reserve. Software stocks are volatile. The best buying opportunities—those beautiful fallen compounder moments—occur during market panics or sector-wide selloffs. If you are fully invested when the opportunity arises, you cannot capitalize on it. Discipline demands dry powder.
Position Sizing: The Conviction Framework
Not all positions deserve equal sizing. Use a conviction-based framework:
| Conviction Level | Position Size | Criteria | |---|---|---| | Tier 1 — Highest Conviction | 10-15% | Generational compounder with durable NRR >120%, best-in-class category leader, proven multi-product road map. | | Tier 2 — High Conviction | 6-10% | Strong compounder with NRR >110%, Rule of 40 >40, clear competitive moat, but may face moderate competitive risk. | | Tier 3 — Developing Thesis | 3-6% | Early-stage or emerging-compounder with accelerating fundamentals but limited track record. Higher risk, higher reward. | | Watchlist | 0% (Monitor Only) | Intriguing company that doesn't yet meet your quantitative criteria. Wait for catalysts before deploying capital. |
Correlation Management
One of the most common mistakes in software investing is building a portfolio of 10 stocks that all behave identically. If all your holdings are high-growth, high-multiple SaaS companies, a sector-wide multiple compression event (like 2022) will devastate your entire portfolio simultaneously.
Diversify across these dimensions:
- Growth Stage: Mix hyper-growth (>40% YoY) with mature cash generators (<15% YoY but 30%+ FCF margins). When multiples compress, the cash generators hold up far better.
- End Market: Don't over-concentrate in a single vertical. Spread across cybersecurity, data infrastructure, vertical SaaS (fintech, health-tech), developer tools, and horizontal platforms.
- Pricing Model: Mix subscription-heavy names with consumption-based names. Subscription revenue is more predictable (less downside), but consumption models capture more upside in strong macro environments.
- Market Cap: Include a mix of mega-cap platforms ($50B+), mid-cap growth leaders ($10-50B), and small-cap emerging compounders ($2-10B). Smaller companies have more room to grow but also more risk.
When to Sell
Selling is the hardest discipline in software investing. The natural instinct is to sell winners (locking in gains) and hold losers (hoping for recovery). You must do the opposite.
Sell When the Thesis Breaks, Not When the Price Drops:
- NRR drops below 100% for two consecutive quarters.
- The company loses its category leadership position (a competitor is consistently winning head-to-head deals).
- A technological paradigm shift (e.g., AI) structurally impairs the company's moat.
- Management credibility is destroyed by repeated guidance misses or accounting irregularities.
- The Rule of 40 score falls below 20 for more than two quarters.
Do NOT Sell For These Reasons:
- The stock is "up a lot" (compounders compound—that's the point).
- A single quarter missed estimates by 2% (noise, not signal).
- A macro panic causes a sector-wide selloff (add to positions, don't flee).
- A sell-side analyst downgrades the stock (sell-side analysts are lagging indicators).
The Power of Doing Nothing
The greatest returns in software investing come from buying a compounder at or below fair value and then doing absolutely nothing for years. Every trade you make is an opportunity to be wrong. Transaction costs, taxes, and behavioral biases all erode returns.
If you own a company with 125% NRR, a Rule of 40 score of 55, and a visionary CEO executing on a multi-product roadmap, your primary job is to sit on your hands and let the compounding work.
Time in the market, in the right stocks, beats timing the market every single time.