I've been building a concentrated growth portfolio around a simple thesis: bet on civilization's infrastructure, buy quality in pairs.
The "two by two" approach means every investment theme gets two positions — one for redundancy, one for diversification within the thesis. If I'm right about the theme but wrong about the stock, I've got a backup. If I'm right about both, even better.
| Theme | Pair | Weights | Total |
|---|---|---|---|
| Semis Design | NVDA + SNPS | 4.7% + 4.7% | 9.5% |
| Aerospace | GE + TDG | 4.7% + 4.7% | 9.5% |
| Big Tech | GOOG + AMZN | 4.7% + 4.7% | 9.5% |
| Insurance | FRFHF + UNH | 4.7% + 4.7% | 9.5% |
| Energy/Commodities | Sogo Shosha + XOM | 4.7% + 4.7% | 9.5% |
| Semis Infrastructure | TSM + ASML | 4.7% + 4.7% | 9.5% |
| Payment Rails | V + MA | 4.7% + 4.7% | 9.5% |
| Crypto/Fintech | HOOD + COIN | 4.7% + 4.7% | 9.5% |
| EV | TSLA + BYDDY | 4.7% + 4.7% | 9.5% |
| Subscription Consumer | NFLX + DUOL | 4.7% + 1.4% | 6.2% |
| Sports/Entertainment | MSGS + GENI | 1.4% + 1.4% | 2.9% |
| Hardware | AAPL + DELL | 1.4% + 1.4% | 2.9% |
| Credit Ratings | SPGI + MCO | 1.4% + 1.4% | 2.9% |
Full weight (19 positions): 4.7% each = 90% total
Reduced weight (7 positions): 1.4% each = 10% total
That's 30 positions across 13 themes. The Sogo Shosha (MITSY, MSBHF, ITOCY, MARUY, SSUMY) trade as a single basket — Japan's commodity trading houses that Buffett has been accumulating.
I ran mean-variance optimization, but ultimately went with a modified equal-weight approach. The math wanted extreme concentration — 10% in NVDA, 10% in GE, minimize everything else. That's fine for a spreadsheet, but I want to keep my conviction bets.
My compromise: equal weight the themes I believe in, reduce the ones I'm less sure about.
The reduced positions (Sports/Entertainment, Hardware, Credit Ratings, and DUOL) get 1.4% each. These are either speculative holds (GENI), slower growers (SPGI, MCO), or positions where I have less conviction. The other 19 positions get equal 4.7% weights.
This beats pure equal-weighting on Sharpe ratio (0.45 vs 0.41) while keeping the portfolio simple to manage.
This is a growth portfolio, not an all-weather portfolio. Some honest numbers:
Tech/AI dominates the risk budget. When correlations spike in a crisis, expect pain. My hedges are FRFHF (Fairfax Financial — Prem Watsa's insurance float machine) and the Sogo Shosha basket. Both stay relatively uncorrelated when tech sells off.
Pure equal-weighting gives the same vote to NVDA (a compounder) and GENI (a drag). By reducing the weaker themes to 10% total, I tilt toward conviction without over-concentrating.
This portfolio makes a bet: civilization keeps building, technology keeps compounding, and quality businesses keep winning.
This portfolio bets on continued progress — on builders shipping products, engineers solving problems, and global trade connecting markets. If you believe in a world like this, these are the companies making it happen.
Long civilization. Two by two.
This is not financial advice. I hold positions in the securities mentioned. Do your own research.