Active Since Jan 2025
Long/short value investing across US and China equities. Top-down macro framework, concentrated high-conviction ideas, focused on value/growth discrepancy. Currently long-biased by design — shorting chip-cycle strength into a momentum tape is a poor risk/reward.
The AI-driven data center buildout requires massive power generation and backup systems — Caterpillar's Energy & Transportation segment is the backbone supplier of diesel generators, gas turbines, and power management systems for hyperscalers. Simultaneously, US infrastructure reshoring and the IRA spending wave are driving construction equipment demand. CAT is a world-class compounder trading at a reasonable multiple relative to earnings power. Consensus underestimates Energy & Transport as a multi-year growth driver.
MasTec is the picks-and-shovels play on US grid modernisation and renewable energy transmission. As the #1 utility-scale solar and wind EPC contractor in the US, it is directly exposed to the IRA-driven renewable buildout. The company is also a leading contractor for 5G wireless infrastructure and power grid upgrades. Margins are recovering after a difficult 2023 transition year, and the backlog is at record levels — visibility is high. The stock significantly lags the fundamental re-rating that has begun.
The market is pricing Microsoft as if AI is a cost centre and competitive threat, not an accelerant. In reality, Copilot is embedding AI across 400M+ Office seats, Azure's AI workloads are growing north of 50% YoY, and enterprise switching costs are near-infinite. The AI disruption narrative actually strengthens Microsoft's moat — it controls both the productivity layer (Office) and the infrastructure layer (Azure/OpenAI). The stock is a high-quality compounder with a widening moat being sold at a multiple that underestimates the AI monetisation cycle.
Take-Two is mis-valued on near-term losses while the market ignores the magnitude of GTA VI. The game is the most anticipated software release in history — pre-launch monetisation from in-game purchases, GTA+ subscriptions, and Shark Cards creates a multi-year cash flow compounding engine. The market prices TTWO on current-year GAAP losses; I value it on the normalised free cash flow engine post-GTA VI launch. AI actually helps Take-Two — generative AI slashes content development costs while enriching game world depth. Fear of disruption is precisely backwards here.
The market fears autonomous vehicles will destroy Uber. I believe the opposite — Uber will be the primary distribution layer for robotaxi fleets. Waymo and Tesla don't want to manage driver supply, insurance, customer support, or demand aggregation at scale. Uber does. Its network effect (platform with 150M+ monthly users) is the distribution moat that no AV player can replicate cheaply. Meanwhile the core rideshare business is printing free cash flow. Uber is being priced as if it will be disrupted; it is actually the disruption aggregator.
Our oil bear thesis is unchanged and well-supported: the IEA projects global supply rising 2.5mb/d in 2026 against demand growth of only 930kb/d. Reuters consensus from 34 economists has Brent at $63.85 and WTI at $60.38 for the full year. JPMorgan's base case has Brent spending extended periods below $60. The current $18–20/bbl Hormuz risk premium creates a large-magnitude deflation trade once geopolitical tensions normalise. We own USO Jul 17 puts (~60% at $100 strike, ~40% at $90 strike), giving us defined downside risk and exposure to the eventual mean reversion — with no volatility decay.
LATAM is positioned asymmetrically — while I am short oil as a theme, LATAM benefits directly from lower jet fuel costs (fuel is ~30% of COGS) while serving a high-growth Latin American travel market. Post-bankruptcy restructuring has created a leaner cost base. The stock is early in its re-rating as institutional investors who avoided the bankruptcy emerge are now comfortable re-entering. A falling oil price environment dramatically expands margins and free cash flow. This is the long side of the oil thesis — own the beneficiary, short the commodity.
Petrobras operates the pre-salt Santos Basin — among the world's most prolific and lowest-cost offshore oil fields. Lifting costs of ~$5/barrel and a full-cycle breakeven near $25/bbl mean the business remains highly cash-generative even if Brent trades at $50–55. In a world where I am short oil at the margin, PBR is the long that survives: it makes money at prices that would destroy higher-cost producers. The bear case on oil is not a bear case on PBR's economics.
PBR trades at ~3–4x EV/EBITDA and yields ~15% — valuations typically reserved for businesses in structural decline. Petrobras's underlying production growth is among the strongest of any major: output is targeting 2.3mb/d by 2028 from the Buzios and Sépia fields. The market is applying a blanket EM discount and political risk premium that materially overstates the actual governance deterioration under Lula. Free cash flow generation is structurally robust; the dividend policy, even in its revised form, implies a yield that global institutional capital will eventually chase.
The Lula-era governance overhang — fears of fuel price caps, dividend interruption, and state-directed capex — is largely already priced in. Actual execution since 2023 has surprised positively: the revised dividend policy retains meaningful shareholder distributions, capex discipline has held, and the strategic plan continues to prioritise the pre-salt core. The delta between feared governance outcomes and actual outcomes is the investment opportunity. As Brazilian macro stabilises and the political cycle matures, re-rating toward EM oil peer multiples implies 60–80% upside from current levels.
Trading at 19x P/E vs. 36x peer median despite 56% YoY revenue growth. Discount anchored to legacy fraud concerns — no longer warranted. Re-rating to peer median EV/Revenue of 3.8× implies 145% upside, excluding earnings growth.
Delivery price war ends earlier than consensus (Q4 2025 vs H1 2026). Opex falls from ~53% to ~47%, adding ¥20mn+ to net income. Gross margin expands 2–5% as competition rationalises.
Proprietary data-driven R&D (app DAU 5.5M vs Starbucks 1.1M). 8-month store payback vs 2.5–4 years for Starbucks. ~31k stores vs 60k modelled capacity — halfway through expansion cycle.
Cloud Intelligence Group revenue +36% YoY in Q3 FY2026 — fourth consecutive quarter of re-acceleration. Qwen AI model surpasses 1B downloads on Hugging Face. AI-related revenue growing at triple-digit rates for 7 consecutive quarters.
Trading at ~9x EV/EBITDA — steep discount to global peers. $11.9B in buybacks in FY2025 demonstrates management conviction. Near-term margin compression is a deliberate investment cycle choice, not structural deterioration.
Qwen app ecosystem connects Taobao, Instant Commerce, Amap, Fliggy and Alipay into a unified agentic platform. 300M+ MAU. By Feb 2026, ~140M users had their first AI-driven shopping experience. T-Head chip production strengthens supply independence.
The Athene-Apollo perpetual capital flywheel — $315B+ of insurance policyholder liabilities continuously recycled into proprietary credit origination — is the most durable competitive moat in alternative asset management. $3.8B annual spread-related earnings (SRE) is deeply underappreciated.
At $109, APO trades at 5.5x EV/EBITDA vs Blackstone at 22x, KKR at 18x. Litigation overhang is pricing in near-terminal risk that is wildly disproportionate to realistic outcomes. Combined FRE + SRE per share of $19.20E by FY2027 implies massive undervaluation.
Apollo is the highest-conviction platform play on private credit's structural growth. AUM target of $1T+ by FY2027 (from $785B). FRE growing from $3.2B (FY2023) to $10B+ target — doubling in 3 years. The secular trend toward alternatives is in its early innings.
Different asset classes are pricing the Iran war shock with dramatically different levels of pessimism. US Treasuries are the most fearful — yields have collapsed as capital fled to safety, implying a severe growth slowdown or recession. Commodities sit in the middle — pricing elevated geopolitical risk but not catastrophe. Equities are the most complacent — still pricing a soft landing and normalisation. This pricing divergence creates a clear arbitrage: if bonds are correct, gold and copper — which have not fully repriced to the bond market's level of pessimism — still have substantial upside. Gold is the purest expression of this trade, and Barrick is the highest-quality equity vehicle to own it.
Gold spot has rallied sharply, repricing toward the pessimism embedded in bond markets. Barrick's equity, however, has badly lagged the gold price move — a classic gold miner underperformance gap driven by cost inflation fears and retail selling. This creates a direct arbitrage: own the miner at a discount to the commodity it extracts. Historically, gold miners exhibit 1.5–2× leverage to the gold price in trending bull markets. At current spreads, Barrick offers a leveraged catch-up trade on top of an already underpriced commodity.
Barrick has been technically oversold relative to both the gold price and its own historical trading range. RSI, moving average divergence, and miner-to-metal ratio signals all point to a mean-reversion setup with a high probability of sharp recovery once momentum rotates back into the sector. This is not a thesis built on technical analysis alone — fundamentals are sound — but the technical oversell creates an asymmetric entry point where the risk/reward is skewed materially to the upside. The combination of a fundamental catalyst (gold re-rating) and a technical reset is a high-conviction entry signal.
Consensus is pricing China copper demand through the lens of property sector weakness — missing that the demand mix has fundamentally shifted. State Grid's grid upgrade capex, EV penetration (China now ~45% of global EV sales), and the domestic renewable buildout are absorbing copper volumes that the construction sector previously monopolised. Chinese copper imports and refined copper consumption data remain robust even as property starts decline. The bear case on China copper demand is based on an outdated demand model that has not updated for the energy transition composition shift.
FCX's share price has disconnected materially from copper spot. LME copper has held near multi-year highs while FCX equity has been punished by tariff fears, China growth concerns, and general risk-off selling. This spread — equity vs underlying commodity — is at historically wide levels. This is the same arbitrage identified across the Iran war cross-market thesis: the equity market is more complacent (or more fearful, in this case) than the commodity market. Either copper falls sharply, or FCX catches up. Given the supply fundamentals, the latter is far more probable.
Sulfuric acid is the critical processing input for DRC copper smelters — and it is sourced almost entirely as a by-product of oil refinery operations in the region. As the Iran war disrupts regional oil flows and refinery throughput, sulfur by-product supply is being squeezed. Without sulfuric acid, DRC copper refineries cannot process copper concentrate into finished cathode at full capacity. This is not yet in consensus supply models. The DRC is the world's largest copper producer — a processing constraint here creates a hard supply ceiling that tightens the global market faster than any demand story alone.
Primary monitoring of labour conditions in Chile and Peru — the world's #1 and #2 copper producers — is showing early-stage strike mobilisation signals: union negotiation breakdowns, worker safety grievances, and community blockades that typically precede formal work stoppages by 4–8 weeks. Chile's Codelco and several privately operated mines are in contract negotiation cycles that have historically been contentious. A simultaneous or sequential strike across multiple South American operations would remove a meaningful percentage of global mined supply at exactly the moment DRC processing capacity is being squeezed. This asymmetric supply shock scenario is entirely absent from current copper price models.
I am a long-only value investor focused on US and China equity markets, with a deep conviction that superior returns come from disciplined top-down macro analysis combined with rigorous bottom-up fundamental research.
My professional experience spans institutional equity research and asset management across Asia. At Nomura Securities, I covered Asia consumer staples as an equity research analyst, initiating coverage on the global tobacco industry and producing research delivered to 200+ institutional clients. At Manulife Investment Management, I supported a fixed-income and A-shares portfolio that achieved a top ranking (1st out of 3,000 funds) in China A-Shares. I am currently an incoming BNP Paribas Fixed Income Summer Analyst (Jun–Aug 2026) and hold the CFA Level 1 designation.
This portfolio documents my live investment thinking. My ambition is to build a career as an equity research analyst and eventually portfolio manager — with the same rigour and passion I bring to every position here.