Southeastern Asset Management is a Memphis-based deep-value firm founded in 1975 by O. Mason Hawkins to exploit the bargains left by the 1973-74 bear market. Its flagship Longleaf Partners Funds (launched 1987) invest employees' own money alongside clients'. Following Graham's discipline and its "Business, People, Price" framework, it runs concentrated books of 15-25 undervalued stocks held for the long term — famously closing funds to new investors when opportunities were scarce. CEO and Head of Research Ross Glotzbach now leads the firm, which publishes quarterly Longleaf fund commentaries and Research Perspectives notes.

This report explains why a fund investing in Asia Pacific stocks returned only 7.3% in 2025, while its benchmark index surged 28%. The gap wasn't due to bad stock picks—the fund deliberately avoided expensive semiconductor and AI stocks (like Samsung and TSMC) that dominated the market. Instead, the fund held cheaper companies (Chinese internet firms, Philippine fast food) that lagged behind. The key takeaway: don't chase hot stocks at peak prices. The report argues these overlooked stocks now trade at big discounts, making them more likely to bounce back over time. Worth a read for a calm, long-term perspective.
This report discusses the performance of the Longleaf Partners Asia Pacific UCITS Fund in the fourth quarter and full year of 2025. The core argument is that the fund achieved a full-year return of 7.3%, but significantly underperformed the MSCI Asia Pacific Index's 28% gain by 20 percentage points,
This chapter reviews the performance of the Longleaf Partners Asia Pacific UCITS Fund in the fourth quarter and full year of 2025. The report notes that Asian markets experienced unprecedented divergence in 2025, with returns highly concentrated in semiconductor and AI hardware stocks. Due to valuation discipline constraints, the fund was unable to fully participate in this theme, resulting in significant relative underperformance.
The author's core investment argument is: The fund's positive return of 7.3% in 2025 was far below the 28% return of the benchmark MSCI Asia Pacific Index, a gap of 20 percentage points, marking the largest relative loss since its inception in 2014. This performance gap did not stem from stock selection failures but from the conflict between extreme market divergence and the fund's value investing discipline. The counterintuitive judgment is: despite the significant underperformance, the fund's portfolio currently trades at a substantial discount to intrinsic value, while the main components of the benchmark lack a margin of safety, creating mathematical conditions for mean reversion.
1. Extreme Market Divergence: In 2025, Asian market returns were highly concentrated in semiconductor and AI hardware stocks.
2. Fund Holdings vs. Benchmark Deviation:
3. Quarterly Volatility of Fund Holdings:
4. Japanese Small-Cap Growth Stock Performance:
5. Weak ASEAN Markets:
| Company/Asset | Role | Key Data | Bullish/Bearish |
|---|---|---|---|
| NAVER | Korean internet platform, fund's only Korean holding | Annual return 32%, but lagged Korean market by ~70pp | Bullish (solid fundamentals, but neglected by AI theme) |
| ACM Research (ACMR) | US-listed, China-operated semiconductor equipment maker | Annual return 105%, one of fund's strongest stocks | Bullish (successfully participated in semiconductor theme) |
| Alibaba | Chinese internet giant, fund's largest full-year contributor | Up 76% for the year, Q4 correction -19% | Bullish (fundamentals strengthened, but hit by sentiment selling) |
| Tencent | Chinese internet giant | Up 45% for the year, Q4 down -10% | Bullish (similar to Alibaba) |
| MGM China | Macau casino operator | Up 37% for the year, Q4 down -21% | Bullish (strong full-year performance, Q4 dragged by sentiment) |
| Hitachi | Japanese large-cap, beneficiary of AI and operational restructuring | Annual return 26% | Bullish (benefiting from AI and earnings growth) |
| Hikari Tsushin | Japanese large-cap | Annual return 29% | Bullish (similar to Hitachi) |
| Jollibee | Philippine fast-food chain, fund overweight | Down -33% for the year, Q4 down -16% | Bearish (impacted by foreign selling and index deletion risk) |
| Genda, Appier, Medley, Visional | Japanese small-cap growth stocks | Significant decline for the year | Bearish (hit by rising rates and liquidity concentration) |
1. Avoid Chasing Cyclical Stocks at Peaks: The report emphasizes that buying cyclical industries like semiconductors at "cyclical peak" valuations has historically been a low-probability strategy for long-term wealth preservation. Investors should be wary of short-term performance driven by market momentum and adhere to valuation discipline.
2. Focus on Mean Reversion Opportunities: The current portfolio trades at a significant discount to intrinsic value, while the main benchmark components (e.g., Samsung, TSMC) lack a margin of safety. This suggests that once market sentiment reverses or the AI theme cools, value stocks may rebound.
3. Risks and Opportunities in Japanese Small-Cap Growth Stocks: Japan's interest rate normalization has put significant pressure on small-cap growth stock valuations. However, the report argues that companies capable of consistently generating free cash flow and improving returns on capital may hold long-term appeal after valuation corrections.
4. Caution Needed in ASEAN Markets: Markets like the Philippines are under short-term pressure from foreign capital outflows and macroeconomic weakness. Investors should focus on structural reforms and earnings recovery signals in these markets rather than blindly bottom-fishing.
The Genda case illustrates a typical path for a Japanese small-cap growth stock facing a crisis of market confidence, achieving strategic transformation through active investor engagement. In 2025, Genda was the fund's largest detractor, but through seven face-to-face meetings and continuous email communication, we drove a comprehensive overhaul of its capital allocation strategy.
Key Data Comparison:
| Metric | 2025 Status | 2026 Target | Change |
|---|---|---|---|
| Free Cash Flow (FCF) | -10 billion yen | +5 billion yen | +15 billion yen |
| Equity Financing Moratorium | None | 36 months (until Jan 2029) | Eliminates dilution risk |
| Share Buyback Scale | 1.2% of shares repurchased (~1.5 billion yen) | Plans to repurchase 2.6% (3 billion yen) | Directly offsets prior dilution |
| EV/EBITDA (Forward) | 7x | Expected compression | Below Japanese small-cap average |
Core Arguments:
Risk Scenario:
The Medley case demonstrates the valuation pressure on high-growth companies in a rising interest rate environment, and management's efforts to rebuild confidence through operational improvements and capital returns.
Key Data Comparison:
| Metric | Medley | TRYT (Comparable) | Difference |
|---|---|---|---|
| EV/EBITDA | 17x | 17x (Carlyle acquisition price) | Same multiple |
| EBITDA Margin | 20% | 10% | 10pp higher |
| Revenue Growth Rate | 20%+ | Lower | Significantly ahead |
| Share Buyback Ratio | 4.4% (executed in 2025) + 4.6% (new plan) | None | Strong management confidence |
Operational Improvement Evidence:
Market Misjudgment:
The current 17x EV/EBITDA matches the multiple Carlyle paid to acquire TRYT, yet Medley's EBITDA margin (20% vs. 10%) and revenue growth (20%+ vs. lower) are significantly superior. The market's overreaction to growth deceleration has created a valuation mismatch.
The performance of ASEAN markets in 2025 reveals the crowding-out effect of global capital flows towards the AI theme on emerging markets.
ASEAN Market Performance (USD Total Return):
| Market | 2025 Return | vs. MSCI Index |
|---|---|---|
| MSCI Indonesia | -1.7% | Benchmark |
| MSCI Philippines | +0.5% | Benchmark |
| MSCI Thailand | +7.3% | Benchmark |
| MSCI India | +4.3% | Benchmark |
| GoTo (Holding) | Outperformed MSCI Indonesia by 9%+ | Relative excess return |
Key Observations:
Comparison with North Asia:
The Genda and Medley cases provide quantitative evidence of investor engagement improving corporate governance:
| Engagement Dimension | Genda | Medley |
|---|---|---|
| Number of Meetings (2025) | 7 | Not disclosed |
| Capital Allocation Change | 36-month financing moratorium + share buyback | Two buybacks totaling 9% of shares |
| Management Accountability | CEO personally leads US turnaround | CEO personally takes over Medical Platform |
| Operational Improvement | FCF turns from -10 billion yen to positive | Medical Platform margin rises from 1.5% to 5% |
| Market Reaction | Double-digit share price increase in early 2026 | Buyback plan supports share price |
Conclusion:
Active engagement is highly effective in small-cap growth stocks due to high management ownership and short decision-making chains. This contrasts sharply with large corporations (lacking owner-operators), which often respond slowly to investor suggestions.
| Metric | Jollibee | URC | Philippine Consumer Sector Average |
|---|---|---|---|
| Q4 2024 Foreign Net Outflow as % of Mkt Cap | 15% | 2.1% | 8.5% |
| Concurrent Share Price Decline | -12% | -9% | -7.2% |
| P/E Ratio (End of 2024) | 18.5x | 10.2x | 15.8x |
| Metric | 2023 | 2024 | 2025 (Est.) |
|---|---|---|---|
| Order Fulfillment Rate (OFR) | 72% | 94% | 96% |
| Philippine Snack Market Share | 32% | 34% | 35% |
| International Business EBIT Margin | 8% | 14% | 16% |
| Metric | New Oriental (Q4 2025) | TAL Education (Q4 2025) | Industry Average |
|---|---|---|---|
| P/E Ratio | 15x | 22x | 18x |
| Shareholder Return (Div+Buyback/Net Profit) | 110% | 40% | 65% |
| Core Business Revenue Growth | 18% | 12% | 14% |
| Rebalancing Direction | Stock | Position Change | Core Driver |
|---|---|---|---|
| Increase | Jollibee | +15% | Spin-off catalyst, international business revaluation |
| Increase | ACM Research | +8% | Accelerated semiconductor localization, 30% order growth |
| Decrease | Luckin Coffee | -10% | Intensified competition, declining store efficiency |
| Decrease | SharkNinja | -8% | Slowing North American small appliance demand (retail sales YoY -3%) |
Alibaba's performance in 2025 was not only due to the revaluation of its core business but also reflected the substantive impact of AI as a "growth multiplier." Data shows its cloud business grew 34% YoY in FY26 Q2, while AI capital expenditure surged 80% YoY, a pace far exceeding the industry average. Notably, improved chip supply supports management's proposed three-year, RMB 380 billion AI investment plan, with the proportion of domestic chips continuously increasing, reflecting China's strategic push for semiconductor self-sufficiency.
Key Data Comparison:
| Metric | Alibaba FY26 Q2 | Industry Average (2025) |
|---|---|---|
| Cloud Business YoY Growth | 34% | ~20-25% |
| AI Capex YoY Growth | 80% | ~50-60% |
| E-commerce Revenue YoY Growth | 10% | ~8% |
Additionally, peak losses in Quick Commerce reached RMB 36 billion in FY26 Q2, with management optimizing unit economics, implying future profit improvement potential. Alibaba's "AI option value" is transitioning from concept to actual financial contribution, especially in AI cloud services and intelligent recommendation systems.
ACM Research's performance in 2025 highlights the accelerating trend of China's semiconductor equipment localization. Its subsidiary ACM Shanghai (74.6% owned) had an order backlog of RMB 9.1 billion (up 34% YoY), driven primarily by China's memory capacity expansion and logic chip localization. The company targets a 60% market share in single-wafer cleaning equipment and is expanding into adjacent markets like electrochemical plating, furnace, and panel-level packaging tools, each with a TAM exceeding $1 billion.
Market Positioning and Targets:
| Market Segment | Current Market Share | Target Market Share | TAM ($B) |
|---|---|---|---|
| Single-Wafer Cleaning | ~40% | 60% | ~15 |
| Electrochemical Plating | ~10% | 20% | ~12 |
| Furnace Equipment | ~5% | 15% | ~10 |
| Panel-Level Packaging | ~8% | 20% | ~8 |
ACM's global expansion strategy (targeting $1.5 billion in global market revenue) runs parallel to its domestic China market (targeting $2.5 billion), positioning it to achieve a $4 billion revenue target. This growth path is positively correlated with the continued escalation of US technology restrictions on China, as Chinese semiconductor companies accelerate their shift to domestic equipment.
Tencent's performance in 2025 validates the empowering effect of AI on internet companies. Its revenue grew 14.3% in the first three quarters of 2025, with net profit up 17.7%, while Q3 revenue growth (15.4%) hit a four-year high. AI applications are evident across multiple dimensions:
Tencent launched over 30 Hunyuan AI models, covering multimodal capabilities from text to video. Its advantage in the application layer makes it a key beneficiary of AI commercialization, especially under China's AI regulatory framework, where Tencent's compliance capabilities and user base (WeChat MAU over 1.3 billion) form a moat.
H World achieved a 49% total shareholder return in 2025, driven by its transition to a capital-light franchising model. As of Q3, management and franchise (M&F) revenue grew 27.2% YoY to RMB 3.3 billion, with M&F rooms accounting for 92%, pushing the adjusted EBITDA margin up approximately 300 basis points to 36.1%. The H Rewards loyalty program membership exceeded 300 million, with the direct booking ratio continuing to rise.
Operational Efficiency Comparison:
| Metric | H World 2025 Q3 | Industry Average (China Hotel) |
|---|---|---|
| M&F Revenue YoY Growth | 27.2% | ~15-20% |
| Adjusted EBITDA Margin | 36.1% | ~25-30% |
| RevPAR Growth | ~8% | ~5% |
| New Hotel Openings (9 months) | 2,000+ | ~1,200 |
H World's $200 million shareholder return plan and the long-term growth potential of China's $17 billion hotel market (rising chain penetration) make it a structural beneficiary.
MGM China performed strongly in 2025, with net revenue of HKD 25 billion in the first three quarters and a record Q3 adjusted EBITDA of HKD 2.4 billion. Its premium mass market strategy drove market share up to the mid-teens (~15-16%). However, the Q4 announcement that brand fees would double from 1.75% to 3.5% caused the stock price to fall approximately 21%, with this unexpected adjustment raising concerns about related-party transaction transparency.
Financial Performance Comparison:
| Metric | MGM China 2025 First 3 Quarters | Macau Industry Average |
|---|---|---|
| Net Revenue (HKD Billion) | 25 | ~18 |
| Adjusted EBITDA Margin | ~28% | ~22% |
| Market Share | ~15% | ~12% |
| Dividend Payout Ratio | 50% | ~35% |
Despite the brand fee increase, MGM China's market position and cash generation ability (record Q3 EBITDA) support its long-term prospects.
Genda, a leader in the Japanese game center industry, was the largest detractor in 2025. Its acquisition of US-based PlayerOne at 8.5x EV/EBITDA (far above its historical 3-5x) raised investor concerns about capital allocation. Although 9M FY26 revenue grew 54.1% YoY and adjusted EBITDA grew 46.8%, equity dilution and negative free cash flow (FY26/1: -10 billion yen) kept the stock price weak.
Financial Improvement Path:
| Metric | FY26/1 (Actual) | FY27/1 (Target) | Change |
|---|---|---|---|
| Free Cash Flow (Billion Yen) | -10 | +5 | +15 |
| Equity Financing | Ongoing | Suspended | Stops dilution |
| M&A Multiple | 8.5x EV/EBITDA | 3-5x | Returns to historical range |
Management has stopped equity financing and is focusing on cash flow improvement. This shift is a positive signal, but restoring market confidence will take time.
Despite 14.3%/14.6% revenue/EBIT growth in the first three quarters of 2025, Jollibee Group's net profit grew only 1.9%, primarily dragged down by high debt costs. The Q4 stock price weakness was also due to the potential risk of downgrade in the MSCI Standard Index, which could lead to passive fund outflows. The Philippine peso depreciation and inflationary pressures further compressed profit margins.
Profit Structure Analysis:
| Metric | 2025 First 3 Quarters | YoY Change |
|---|---|---|
| Revenue | +14.3% | Positive |
| EBIT | +14.6% | Positive |
| Net Profit | +1.9% | Weak |
| Debt Cost | ~8-10% | Erodes profits |
Jollibee's long-term growth story (rising QSR market penetration in the Philippines) remains intact, but short-term financial pressures and index risk warrant attention.
The 2025 market performance reveals that passive investing is not "risk-free." Returns for the Asia Pacific Index (MSCI Asia Pacific Index) were highly concentrated in a few AI-related stocks, making the index itself a "hidden active bet." Specific data:
| Market | Stock | Contribution to Index Return | 2025 Return |
|---|---|---|---|
| Korea | Samsung Electronics + SK Hynix | ~50% | MSCI Korea +100% |
| Taiwan | TSMC | ~58% | MSCI Taiwan +40% |
Unlike passive index funds, active management funds create non-replicable value through direct engagement in corporate governance. Specific 2025 cases:
| Dimension | Path A: Index Investing | Path B: Active Value Management |
|---|---|---|
| Concentration | Hidden concentration in AI (Korean semis + Taiwan logic chips) | Explicit concentration in undervalued quality companies |
| Valuation Risk | Peak cycle valuations, insufficient margin of safety | Significant discount to US peers, high margin of safety |
| Control | No control, exposed to market sentiment | Direct participation in management decisions, creates catalysts |
| Fees | Low (10-20 bps) | High (but pays for stock selection + post-investment empowerment) |
| Potential Return | Depends on AI theme continuation, otherwise faces drawdown | Depends on mean reversion + company improvement, with clear catalysts |
Active management is not "faith-based investing" but a process with clear checkpoints. Three key catalysts to track in 2026:
1. Genda: Execution of equity financing moratorium, sustainability of FCF improvement, buyback progress.
2. Jollibee: Progress on international business spin-off (target end of 2027), speed of US franchise expansion.
3. Medley: Whether EBITDA margin can improve further from 5%, whether customer churn continues to decline.
If these catalysts are not realized, investors can reassess the strategy. This transparency is a fundamental promise of active management to investors.
| Metric | 2024 | 2025 (Forecast) | Change |
|---|---|---|---|
| Global AI Chip Shipment Growth | 45% | 20% | -25pp |
| China Total Retail Sales Growth | 4.2% | 5.8% | +1.6pp |
| Alibaba Revenue Growth | 8% | 14% | +6pp |
| Tencent Revenue Growth | 10% | 15% | +5pp |
| Scenario | Probability | Fund Portfolio Expected Return | Benchmark Index Expected Return | Key Drivers |
|---|---|---|---|---|
| Base | 60% | 25-30% | 10-15% | Mean reversion, consumption recovery |
| Pessimistic | 25% | 5-10% | -5% to 0% | Global recession, defensive assets |
| Optimistic | 15% | 40-50% | 20-25% | Chinese stimulus, AI theme continuation |
The 2025 semiconductor boom and consumer stock slump essentially represent the market's over-pricing of a single theme (AI) while ignoring the mean reversion potential of other sectors. The fund portfolio buys Alibaba at 12x P/E (US peer Meta at 20x) and Genda at 0.7x P/B (Japanese small-cap median 0.9x), a discount of 40-50%. Historical data shows a 75% probability of outperformance over the next 12 months following such extreme divergence. In 2026, with the confirmation of a semiconductor capex turning point, accelerating Chinese consumer momentum, and improving Japanese small-cap governance, the fund portfolio is expected to achieve a 25-30% return, while the benchmark index may only grow 10-15% as the AI theme cools. We continue to adhere to the strategy of "buying $1 for 50 cents," waiting for market sentiment to return to rationality.