Strategist reveals an 'asymmetric AI' trade: 'A lot of upside without much downside'

Schroders' Dorian Carrell argued that convertible bonds — corporate debt instruments that can convert to equity — offer an asymmetric, lower-risk way to gain AI exposure, noting the asset class is up roughly 15% year-to-date. He highlighted increased tech issuance (Alphabet, Meta, Amazon, Oracle) to fund AI efforts, cautioned about cash-flow and leverage risks, and recommended geographic and asset-class diversification, citing Europe’s ~2% rates/inflation and cheaper AI ‘picks and shovels’ opportunities in Asia including Japan.

Analysis

Market structure: Rising AI-driven debt issuance and convertible performance (≈+15% YTD) make convertibles the likely near-term winners (convertible funds/issuers, AI infrastructure suppliers in APAC/Japan), while high-debt, tight-FCF names (Oracle/ORCL flagged) and long-only equity holders face dilution and credit-pressure risk. Increased supply of corporate debt shifts demand into hybrid instruments—convertibles absorb equity upside while capping downside, flattening equity-vol term-structure and modestly tightening credit spreads in the short run.

Risk assessment: Key tail risks are a sudden 100–150bp widening in corporate spreads, a >100bp move up in 10y yields, or an AI regulatory shock that collapses tech multiples; any of these would rapidly reprice convertible bond floors. Immediate (days) sensitivity is to CPI/Fed comments and big-tech earnings; short-term (weeks) to funding/credit-curve moves; long-term (quarters) to real cash-flow delivery on AI capex. Hidden dependency: convertibles’ convexity relies on implied vol staying elevated; falling vol reduces option value disproportionately.

Trade implications: Tactical allocation (2–3% portfolio) to active convertible strategies (e.g., Schroders Global Convertibles or broad ETF CWB) for 3–12 months; pair trade long convertibles vs short ORCL equity (or buy ORCL 3‑month 10% OTM puts) sized 0.5–1% PV. Use 6–9 month call-spreads on GOOGL/META sized 1–2% for upside capture while buying convertibles as downside hedge. Rotate 5–8% into Asia/Japan AI supply-chain names over 6–12 months where capex is cheaper.

Contrarian angles: The market underestimates convertibles’ asymmetric payoff—if tech earnings beat, convertibles rerate faster than equities due to convexity. Conversely, the consensus may be underpricing credit risk: a 100bp+ BBB widening would hurt convertibles’ bond-floor and make the trade overdone. Historical parallels (post-2016 tech cycles) show convertibles outperform early in re-rating but are vulnerable if rates re-accelerate; set explicit stop-loss thresholds.

Key Decisions for Investors

• Establish a 2–3% portfolio position in diversified convertible exposure over the next 30 days: prefer active managers (Schroders Global Convertibles Fund) or ETF CWB as tactical 3–12 month hedge against AI equity drawdowns.
• Enter a pair trade: long CWB (or Schroders fund) 2% vs short ORCL equity 1% (or buy ORCL 3‑month 10% OTM puts, size to 0.5–1% premium). Trim/close if ORCL free cash flow surprises positively or ORCL outperforms by >15% in 6 weeks.
• Buy 6–9 month call spreads on GOOGL and/or META (allocate 1–2% each) to capture upside with defined risk; finance by selling nearer-dated calls on non-AI cyclical tech positions to keep net delta modest.
• Rotate 5–8% of equity risk into Asian/Japanese AI 'picks-and-shovels' (semiconductor equipment and data-center suppliers) over 6–12 months; re-evaluate if 10y UST >4.0% or BBB spreads widen >100bps, at which point reduce convertible exposure by 50%.