Why this $2.7-billion money manager recently bought Enerflex and trimmed Warner Music

Portfolio manager Jeff Mo said he is becoming more cautious amid what he calls "schizophrenic optimism," citing fragile economic conditions, Middle East war risks, energy-market disruption and uncertainty around the upcoming CUSMA review. He highlighted continued buys in Enerflex, CACI International and Lumentum, which represent 3.5%, 3.5% and 1.5% of the portfolio, respectively, while trimming Warner Music Group to about 1% because of AI-driven music disruption concerns. Performance data showed Mawer U.S. Mid Cap Equity Fund up 17.3% over the past year and Mawer New Canada Fund up 20.5%.

Analysis

The portfolio skew is revealing: the closest thing to a common factor across the adds is not “AI” per se, but hard-asset optionality plus mission-critical spending. That matters because in a late-cycle tape, the market often pays up for duration via AI narratives while underpricing operators with real pricing power and replacement-cost barriers. EFXT and CACI both look like beneficiaries of a world where buyers care less about perfect GDP and more about resiliency, redundancy, and defense spend that can be justified politically even if growth softens.

The second-order setup is that the same macro anxieties that hurt broad cyclicals can actually tighten demand for these names: more geopolitical stress pushes governments toward electronic warfare, more data-center buildout stresses power infrastructure, and more supply-chain uncertainty improves the value of onshore or North American equipment. The risk is that these are crowded “good business in a bad world” themes, so the entries matter: if rates back up or growth decelerates abruptly, multiples can compress faster than fundamentals re-rate upward. The market is likely underestimating how much of the near-term upside in EFXT and CACI is already in execution, not just in TAM expansion.

LITE is the cleanest expression of AI capex scarcity, but also the most fragile on a sentiment basis because it trades like a long-duration growth asset with a premium multiple. The contrarian point is that investors may be too fixated on end-market demand and not enough on customer concentration and procurement cadence: if AI capex pauses for even one budget cycle, the stock can de-rate hard despite intact secular demand. WMG is the clearest tell that management is becoming more sensitive to AI cannibalization risk; the market may still be underestimating how quickly generative music can pressure licensing economics and bargaining leverage, especially if labels are forced into lower-margin rights-sharing structures.

The broader takeaway: this is not a blanket risk-off memo, but a preference for cash-generative niche leaders with identifiable demand catalysts over IP-heavy businesses facing synthetic substitution. The next 3-6 months are about whether macro fragility translates into capex discipline and whether AI becomes a demand accelerator or an economic substitute in each vertical. That bifurcation should widen dispersion, which favors pair trades over outright beta.

Key Decisions for Investors

• Long EFXT vs short industrial beta: initiate a relative-value long in EFXT against XLI or a mid-cap industrial basket for a 3-6 month horizon. Thesis: resilient energy infrastructure and data-center backup demand should outperform when growth uncertainty rises; risk is if gas capex slows or valuation compresses with rates.
• Long CACI on dips over the next 2-4 weeks; use weakness tied to headline risk as entry. Risk/reward is favorable because defense-electronic warfare spend is less rate-sensitive than software multiples, and any escalation in geopolitical stress can extend the rerating window for months.
• Buy LITE only via call spreads rather than common stock for the next 6-9 months. The setup benefits from AI capex strength, but the high multiple makes it vulnerable to any pause in data-center orders; defined-risk convexity is preferable to outright exposure.
• Trim or hedge WMG against AI-content disruption over a 6-12 month horizon. Consider a partial short or put spread if the market starts pricing generative music monetization more aggressively; downside can accelerate once rights holders realize substitution risk is structural, not cyclical.
• Pair long CACI / short WMG as a quality-vs-substitution trade for 6 months. The spread favors businesses selling mission-critical, budget-protected capabilities over media assets facing IP commoditization and margin pressure.