Massif Capital Sees Volatility Double as Geopolitics Hits Energy and Mining

Massif Capital's Real Assets Strategy lost 5.16% net of fees in the second quarter of 2026, trimming year-to-date net returns to 10.7%. April and May were strong, the firm said in a letter to investors. June was not. Losses concentrated in the European energy book, and significant selling hit single names across the materials sleeve, moves the managers attributed to a rotation out of materials and a general risk-off mood.

Portfolio volatility has roughly doubled. Annualized volatility from January 2019 through the end of 2025 ran 16.2%. Over the first six months of 2026 it hit 32.8%, higher than any full calendar year including 2020. The trailing twelve-month figure, spanning both 2025 and 2026, came in at 25.9%, also above every prior full year. Four of the first six months of 2026 moved more than 8% in either direction. No prior year except 2020 had three such months across twelve.

The managers traced the volatility to rising commodity beta at the portfolio level. Until late 2025, the fund's commodity sensitivity was low, with most returns explained by idiosyncratic factors. Over the fourth quarter of 2025, the portfolio's variance sensitivity to the rate of change in commodity prices increased. Toward the end of the first quarter of 2026, that convexity sensitivity began to fade while sensitivity to underlying commodity price movement rose. The transition corresponded with a stall in positive appreciation of positions.

The pattern repeated across individual names. Lundin Mining (OTCPK:LUNMF) fell 12.6% on June 8 with no stated catalyst, then another 7% on June 23, all while news flow ran the other way. The company closed an additional purchase of 5% of Caserones and a 30.9% stake in Los Helados for $215 million, reported Q1 free cash flow of $380 million on 80,000 tonnes of copper production, and on June 22 filed NI 43-101 technical reports confirming a 2.3-billion-tonne combined reserve base. Argentina granted the Vicuña project fiscal-stability status under its RIGI regime. None of it moved the stock until June 30, when Lundin rose 2% on Commerce Secretary Lutnick's recommendation of a 15% refined copper tariff from 2027.

Equinox Gold (EQX) fell 34% in the second quarter and 22% in June alone. The underlying business improved on nearly every dimension. Q1 production came in at 197,628 ounces. Debt fell from $1.15 billion to $252 million on an asset sale. The company declared its inaugural dividend and announced a definitive all-stock merger with Orla Mining on May 13 that would create an approximately $18.5 billion North American senior producer targeting 1.1 million ounces of 2026 production. The stock fell 8% on June 10 as gold dropped 3.5% to $4,136 an ounce. On June 25, Equinox signed 20-year land-access agreements with all three communities surrounding its Los Filos mine, ending a 14-month suspension and clearing the way for a restart that will produce 280,000 ounces of gold a year for 15 years. Production guidance already excludes any contribution from it.

Gold fell 15% for the quarter. The GDX dropped 17%. Equinox fell twice as much. Within the firm's four-factor model, commodity price and convexity explained most of the day-to-day variance but only half of the negative return. Currency exposure and idiosyncratic factors accounted for the rest. With operations and financials only improving, the managers attributed much of the negative idiosyncratic return to negative sentiment.

The energy book had a genuine macro shock layered on top. Brent round-tripped from a $107 spike in April to a sub-$80 close in June on an alternating sequence of Iran-Hormuz escalation and ceasefire headlines. Var Energi, Aker BP, Equinor, and Harbour Energy, roughly 29% of the book across the quarter, fell five to nine percent in single sessions on multiple occasions, almost always on a headline about the negotiation itself rather than any change to reserves, contracts, or dividend capacity.

Equinor's June 12 decline of 7.8% came the same day OPEC issued its second consecutive demand-growth downgrade. Its 6.4% decline on June 16 coincided with a Capital Markets Day that doubled the 2026 buyback to $3 billion, raised its dividend-growth framework, and scrapped the 2030 renewables target. Var Energi fell 6.9% and Aker BP 5.7% on June 12. Both were down a further 5-8% on June 16. Harbour Energy tracked Brent almost tick for tick, falling 4.3% on June 12 and drifting lower through month-end as crude settled near $72-75.

None of it touched what the businesses did. Var Energi posted record Q1 production of 406 kboepd, up 51% year over year, sanctioned $300 million dividends for both Q1 and Q2, took final investment decision on Balder Next New Wells and the NOK 14 billion Gjøa subsea project, and sold stakes in Goliat and Fenja to PKN Orlen for up to $350 million. Aker BP's Q1 EBITDA came in at $2.66 billion on 97% operational efficiency, with Symra field reaching first oil nine months ahead of schedule. Harbour Energy raised full-year production guidance to 480,000-500,000 boepd, won a tax dispute over the Waldorf acquisition, and closed a $215 million sale of its Indonesian assets in late May.

The managers published two white papers during the quarter. One argued that the cleared price of commodities has migrated from a geology-first regime to a geography-first regime, where chokepoints, processing nodes, export licenses, and sovereign policy set the price. The other extended the same logic to interest rates, arguing the long end of the sovereign curve carries a persistent geopolitical premium.

They offered two case studies. In the first, the Commerce Department directed Anthropic to suspend access to its Fable 5 and Mythos 5 models for all non-US persons on June 12, invoking export-control authority due to the models' cyber capabilities. Access was restored on July 1 with more restrictive guardrails. The episode lasted under three weeks. The managers argued the signal will last considerably longer. AWS Bedrock, Google Cloud, and Microsoft Foundry were all affected at once through a compliance obligation imposed on their supplier. At the G7 summit, the French president warned that no one would buy a model that could be switched off by a foreign government. A Japanese lab released a model marketed on the explicit promise of frontier capability without export-control risk. Even Palantir (PLTR), a US defense contractor, advised its customers to pursue corporate "sovereignty" through open-weight models.

The second case involved large power transformers. Imports account for roughly 80% of US power-transformer supply and nearly 90% of large-power-transformer demand in 2024. Lead times that ran from one to one-and-a-half years before the pandemic now run three to four years. Prices are 60% to 90% higher than in 2020. The transformer core requires grain-oriented electrical steel. The United States has a single domestic producer, meeting roughly one-fifth of demand. The remaining 80-plus percent is imported from South Korea, Japan, and Germany. As of April 2026, nearly half of US data centers planned for the year were expected to be delayed or cancelled, with shortages of transformers, switchgear, and batteries among the primary causes, including projects tied to Microsoft and OpenAI.

The managers concluded that geopolitics has become an ever-present endogenous variable. They argued the market is still pricing the first regime, where political risk is an occasional exogenous shock, while the portfolio is operating in the second. The gap between the S&P 500's forward earnings yield and the 10-year Treasury sits near 49 basis points against a long-run average of 300 to 400 basis points. A further 30 to 40 basis points on the 10-year, absent a commensurate move in the earnings yield, returns the premium to zero.

Existing investors in the fund were told to expect an invite to a Zoom call on August 19. The firm also expects to launch two SPVs during the third quarter.