2026: Second Quarter Review

Portfolio Update 
- Uranium (URNJ): Exited at the end of April – a couple of months later than ideal, as flagged last quarter. OM is now more actively using technical signals to help time position sizing; having ignored the first warning in January, he acted on the second in April.

- Crypto/China/India: Exited the remaining stub positions mid-quarter.

Together, these exits pushed cash above 25% by quarter-end. Part of this is a timing mismatch – capital left Crypto, India and China before new ideas (Brazil) were sized up. That, plus a likely re-investment in Uranium, should bring cash down in the third quarter.

Performance and Review
Our Man’s portfolio rose +4.3% during the second quarter, substantially lagging both the S&P 500 TR (+15.2%) and the MSCI World (+14.0%). The portfolio underperformed in every month of the quarter – rising in April (+6.5%), well behind the index rally, then falling outright in May (-0.2%) and June (-1.8%) while markets continued higher.

For the year, the portfolio (+8.5%) sits a little behind the S&P 500 TR (+10.2%) and MSCI World (+10.4%).

The medium-term history of the portfolio is below: 

 

Second Quarter Attribution

 
There are few surprises behind the underperformance: an average of ~20% cash during the quarter, no direct AI exposure, and weak commodity performance. The cash drag isn’t as bad as it looks – about one-third of it came from exiting URNJ at a small profit, which helped avoid a 20%+ loss over the rest of the quarter.

OM was late to recognize AI’s demand pull through the supply chain and missed the initial move, choosing instead to hold indirect exposure – through the US Reindustrialization theme (data centers, grid buildout, etc.) and Commodities (metals used in chips and power). That indirect exposure worked well over the last year but got hurt in Q2 as capital rotated hard into direct AI/growth names. Nothing changed in the theses on Uranium (-261bps) and Commodities (-40bps) – capital simply rotated out, and that will happen intermittently.

Not investing in HBM/DRAM manufacturers was a major error of omission. The setup mirrors Shipping/Tankers, Tin and Uranium: supply is fixed, demand jumps, and price is the variable that adjusts.  AI data centers need HBM (high-bandwidth memory) to feed GPU clusters, and it draws roughly 3x the wafer capacity of standard DRAM per bit; manufacturers are shifting production toward HBM’s better margins, tightening conventional DRAM supply and pushing DRAM prices up 90%+ per quarter. OM has known this was coming for almost a year but wasn’t invested.  Before April 2026 this was because there was no effective way to own it; the manufacturers are mostly single non-US companies (Samsung, SK Hynix, Kioxia), there was no dedicated ETF, and the names were too small a slice of broad semiconductor ETFs to matter.  This changed with the April 2026 launch of a Roundhill DRAM Memory ETF. The sin was knowing this ETF was launching but not acting. The excuses are the usual ones – “up too much already,” “wait for a pullback,” “don’t want to buy the top” – and none of them really matter.
This is a recurring lesson for OM, tied to his personality and approach rather than a rule for everyone, and a reminder of why investing is hard.  It is better to be early and small than not there at all, and if you can't be early but do have conviction then at least be small. It's easier to size up an existing position than start a new one.

The rest of the portfolio held up well:
-    European/UK Financials (+200bps) began to re-rate on strong results – particularly returns on tangible equity – aggressive buybacks, and analyst upgrades.
-    US Reindustrialization (+184bps) benefited from capital rotating toward direct AI beneficiaries.
-    Tin (+155bps) prices remained strong – Alphamin Resources returned ~10% of its market cap as a dividend, though the stocks remain underappreciated by the market.
-    Carbon Credit Allowances (+101bps) strengthened after CARB removed the policy overhang and began finalizing steeper cap-tightening trajectories (10-14% per annum vs. 4% previously).
-    Argentina (+74bps) banks benefited from continued pro-market reform, accelerating private credit growth, and healthy net interest income.
 

Shipping/Tankers (+4bps) and Idiosyncratic Equities (-0bps) were flat.  The exited stub positions added small gains: India (+5bps), China (+5bps) and Blockchain/Crypto (+6bps).

Portfolio (as at 6/30/26 - all delta and leverage adjusted, as appropriate)
Dislocations: 33.7%
14.1% - Uranium (URNM)
6.9% - Argentina (BMA, GGAL, SUPV)
6.6% - Tin (AFMJF, MLXEF and SBWFF)
6.1% - Commodities/Mining (LUNMF, TLOFF, PMCOF and IVPAF)

Thematic: 37.4%
10.4% - US Reindustrialization (AIRR)
10.2% - European/UK Financials (BCS, LYG, NWG)
9.2% - Shipping/Tankers (STNG, INSW, TNK, DHT and FRO)
7.6% - Carbon Credit Allowances (KCCA)

Idiosyncratic: 2.5%
2.5% - Equities (JOE)

Shorts/Hedges: 0.0%

Cash: 26.3%

Disclaimer: Nothing above should be considered investment advice or a recommendation to buy or sell any security. While Our Man is invested in all of the securities mentioned, that alone is a terrible reason for anyone else to be. Our Man also holds some cash and a few other positions (of negligible value). Investors should always do their own work and make decisions based on their own circumstances, objectives, and risk tolerance—and not because Our Man happened to mention something here.