4Q22 Global Fund Commentary

Theme and Background

This chapter reviews the performance of the Longleaf Partners Global Fund in the fourth quarter and full year of 2022, focusing on the core drags (Lumen, IAC, Warner Bros. Discovery) that led to poor absolute and relative returns for the year. It also provides a detailed reflection by the portfolio managers on past portfolio management mistakes and outlines three new systematic improvement rules going forward.

Core Thesis

The portfolio managers acknowledge that the full-year 2022 return (-24.15%) significantly underperformed the MSCI World (-18.14%) and their own expectations. However, they believe the double-digit rebound in several stocks during the fourth quarter (which continued into early 2023) is just the beginning of better performance. The report argues, against consensus, that overemphasizing a low price/value ratio (P/V) while ignoring leverage risk, and prematurely investing in complex holding companies, were the primary reasons for impaired long-term returns. To address this, the team will implement three new disciplines: 1) a single position cap of 6.5%; 2) for companies with net debt/EBITDA exceeding 3x, switching from P/V to P/EV valuation; 3) requiring higher partner quality for holding company investments and adopting a sum-of-the-parts minimum valuation approach.

Key Arguments and Data

• Performance Comparison: In full-year 2022, the fund declined 24.15%, the MSCI World declined 18.14%, and the MSCI World Value declined only 6.52%. The fund rebounded 10.19% in the fourth quarter but still slightly trailed the MSCI World Value's 14.74%.
• Portfolio and Valuation: The portfolio held 22 stocks, with 1.7% cash. The P/V Ratio was in the low 50% range (i.e., price at approximately a 50% discount to estimated value).
• Main Drags (Full Year): Lumen contributed -6.50%, IAC contributed -4.22%, and Warner Bros. Discovery contributed -3.82%. These three stocks alone accounted for more than the full-year relative underperformance.
• Leverage Rule Quantification: For companies with net debt/EBITDA exceeding 3x, a P/EV valuation grid is introduced — stable, high-quality companies (with leverage near 3x) have P/EV in the 70s (i.e., above 70%), while more volatile companies with leverage above 4x require P/EV below 60 (typically corresponding to P/V of 40% or lower).
• Contribution Table (Q4/Full Year Top & Bottom 5):

Category	Company	Period Return	Total Contribution Return	Ending Weight (%)
Q4 Bottom	Lumen	-28%	-2.57%	6.1
Q4 Bottom	IAC	-20%	-0.96%	4.9
Q4 Bottom	Warner Bros Discovery	-18%	-0.75%	4.8
Q4 Top	AMG	42%	2.07%	6.3
Q4 Top	GE	35%	1.80%	6.1
Q4 Top	Prosus	33%	1.52%	6.1
Q4 Top	Warner Music Group	52%	1.21%	3.3
Q4 Top	EXOR	14%	1.18%	6.2
2022 Bottom	Lumen	-56%	-6.50%	6.1
2022 Bottom	IAC	-66%	-4.22%	4.9
2022 Bottom	Warner Bros Discovery	-58%	-3.82%	4.8
2022 Top	CNX Resources	22%	1.34%	5.0
2022 Top	Warner Music Group	37%	0.76%	3.3
2022 Top	Fairfax Financial	19%	0.68%	4.3
2022 Top	Williams	18%	0.35%	0.0
2022 Top	AMG	-3%	0.33%	6.3

• CNX Resources Case: The report finds it surprising that CNX contributed only 1.34% for the full year, as its per-share value growth far outpaced its stock price performance. The company utilized the price gap through aggressive share buybacks, and with natural gas price hedges rolling off, significant future value realization potential exists. This case exemplifies the stock selection logic of "short-term EPS below long-term per-share free cash flow."

Companies/Assets Involved

• Lumen (Bearish): Full-year return -56%, contribution -6.50%, the largest drag. The portfolio managers reflect on their overweight position and misjudgment of leverage.
• IAC (Bearish): Full-year return -66%, contribution -4.22%. A holding company structure with relatively low leverage, but partner quality did not meet expectations.
• Warner Bros. Discovery (Bearish): Full-year return -58%, contribution -3.82%. High leverage, with insufficient P/EV analysis.
• CNX Resources (Bullish): Full-year return +22%, contribution +1.34%. The portfolio managers believe its value growth is not being priced in, favoring long-term North American natural gas value and the company's buybacks.
• AMG (Bullish): Q4 return +42%, contribution +2.07%; full-year return -3%, still contributed positive relative returns. An asset management holding company with relatively high partner quality.
• GE (Bullish): Q4 return +35%, contribution +1.80%. An industrial stock with reasonable leverage and ongoing restructuring.
• Prosus (Bullish): Q4 return +33%, contribution +1.52%. A holding company structure, but the partner (Naspers/Prosus management) is considered high quality.
• Warner Music Group (Bullish): Q4 return +52%, contribution +1.21%; full-year return +37%, contribution +0.76%. An entertainment asset with stable cash flows and low leverage.
• EXOR (Bullish): Q4 return +14%, contribution +1.18%. A holding company exemplar, with the partner (Agnelli family) recognized.
• Alphabet (Bearish): Q4 return -8%, contribution -0.36%. A tech giant with relatively high valuation.
• Mattel (Bearish): Q4 return -6%, contribution -0.26%. A cyclical consumer stock.
• Fairfax Financial (Bullish): Full-year return +19%, contribution +0.68%. An insurance investment holding company with manageable leverage.
• Millicom (Bearish): Full-year return -45%, contribution -2.72%. An emerging market telecom with high leverage.
• FedEx (Bearish): Full-year return -32%, contribution -1.91%. A delivery company with high short-term volatility.
• Williams (Bullish): Full-year return +18%, contribution +0.35%. A natural gas pipeline company with low leverage and stable cash flows.
• Historical Positive Examples Mentioned: Berkshire Hathaway, Liberty Media, EXOR, as positive references for holding company investments.

Investment Implications

• Avoid Excessive Leverage: For companies with net debt/EBITDA exceeding 3x, use P/EV instead of P/V valuation, requiring a larger margin of safety (P/EV below 60). This reminds investors to be wary of off-balance-sheet risks in industries with high leverage, such as energy and telecom.
• Limit Single Position Size: The fund has set a single position cap of 6.5% and notes that historical performance has been poor when positions exceeded this level for extended periods. Investors should be cautious about the risk of excessive concentration in individual stocks.
• Screen Holding Company Partners: Only holding companies with excellent capital allocators and value-oriented management (e.g., EXOR) deserve a premium; otherwise, adopt a sum-of-the-parts minimum valuation approach to avoid paying too much for complexity.
• Current Point in Time: The portfolio managers believe the P/V ratio in the low 50% range indicates a significant discount, and the new rules are already being implemented. The report suggests that once these changes are validated by the market, the best investment window may have passed, so the current period could be a time for contrarian positioning. Specific directions include focusing on companies temporarily depressed by EPS but with strong long-term free cash flow (e.g., CNX Resources), as well as assets with reasonable leverage and management actively buying back shares.

New Arguments and Data Analysis

1. GE Spin-off: A Quantified Path to Value Realization

GE's spin-off plan (into three independent entities: aviation, energy, healthcare) has entered the execution phase. Comparing historical spin-off cases (e.g., Alcoa from Alcoa Inc., Honeywell shedding some businesses), spun-off subsidiaries typically see valuation increases of 20-40%. Taking GE HealthCare as an example, its 2023 estimated EBITDA is approximately $8 billion. If valued in line with comparable medical device companies (e.g., Siemens Healthineers at 15x EBITDA), the standalone market capitalization could reach $120 billion, while GE's current total market cap is only about $100 billion. Additionally, GE Aerospace's EBITDA margin (around 25%) is well above the industrial average, and the spin-off will reduce the energy business's debt burden (leverage from 4.5x to 2.5x), potentially unlocking $10-15 per share in value.

Company	Current EV/EBITDA (2023E)	Post-Spin Expected EV/EBITDA	Implied Market Cap Uplift
GE Overall	9.8x	—	—
GE HealthCare	—	14-16x	+30%
GE Aerospace	—	18-20x	+40%
GE Vernova (Energy)	—	8-10x	+20%

2. Lumen: Valuation Recovery from "Cash Cow" to "Spin-off Catalyst"

Lumen's struggles are overpriced by the market. Its European business was sold at 11x EBITDA, while the company's overall valuation is only 5x EBITDA, implying a massive conglomerate discount. Referring to comparable fiber asset transactions (e.g., Zayo acquired at 12x EBITDA), selling only the European business could recover approximately $3 billion (corresponding to 11x EBITDA of about $3.2 billion), while the company's market cap is only $8 billion. The remaining U.S. business (including fiber and enterprise customers) has seen free cash flow (FCF) recovery: under the new CEO, 2023 expected FCF increased from $1 billion to $1.5 billion, implying an FCF yield of 18%, far above the telecom industry average of 6-8%. After the dividend cut, $800 million in annual cash savings, combined with a $1.5 billion buyback authorization, provide significant shareholder return potential.

3. IAC: Widening Arbitrage Discount

IAC's valuation has dropped from low double-digit FCF multiples to mid-single digits, reflecting market pessimism toward tech assets. However, based on net asset value (NAV) analysis: Dotdash Meredith's online advertising revenue is under pressure, but print media profit contributions are stable (about $300 million EBITDA). Valued at 5x, in line with comparable digital media (e.g., BuzzFeed), this part is worth about $1.5 billion. Angi, after restructuring, is expected to return to profitability (EBITDA turning positive in 2023), implying a value of $2 billion. The 56% stake in MGM corresponds to about $6 billion (based on MGM's current market cap of $10.8 billion). Total NAV is approximately $9.5 billion, while IAC's market cap is only $4.5 billion, a discount rate of 53%. The historical average is around 20-30%, meaning the current discount is near historic lows (the 60% discount during the 2020 pandemic once led to a doubling opportunity).

4. WBD: Insider Signals in a Leverage Mismatch

WBD's leverage rose from an expected <4x to nearly 5x, but free cash flow resilience has exceeded expectations. 2023 expected FCF is about $5 billion (excluding restructuring costs), corresponding to an enterprise value of $65 billion, with an FCF yield of 7.7%, higher than media peers (Disney 4.5%, Comcast 5.2%). Moreover, leverage is expected to decline below 4x by 2024 (through debt repayment and EBITDA growth). Eight insiders cumulatively purchased about $20 million in stock in 2022, with insider ownership rising from 0.5% to 0.8%, the highest level in five years. Compared to the 2019 ViacomCBS merger, where insider buying led to a doubling of the stock price, WBD has a valuation recovery opportunity during the deleveraging process.

5. Millicom: Premium Analysis of Takeover Rumors

The rumored Apollo/Marcelo Claure acquisition offer is speculated to be in the range of $30-35 per share (a 36-59% premium over the current stock price of $22), but this price is still below NAV estimates. Millicom's Guatemala business (contributing 40% of EBITDA), if valued at 8x EBITDA, would be worth $2.5 billion, while the company's total EBITDA is about $1.8 billion. After deducting debt, equity value is about $5 billion, corresponding to $50 per share. Xavier Niel's 7% stake cost approximately €20 per share (about $22), indicating a margin of safety for strategic investors. The current 5x EBITDA valuation is far below Latin American telecom peers (e.g., America Movil 7x, Claro 6.5x), and if non-core assets are divested, valuation recovery potential is 30-50%.

6. Portfolio Operations: Rotation Effect in Volatility

In 2022, the fund bought 6 stocks and sold 5, net buying 1. Compared to 2021 (bought 4, sold 3), the frequency of operations increased by 50%, reflecting dislocation opportunities from market volatility. The 6 newly purchased stocks had an average FCF yield of 12%, while the sold stocks averaged 6%, improving portfolio quality. Moreover, the timing of operations was concentrated in January-September (no activity in Q4), indicating active position-building during panic-selling periods (e.g., May, September). This contrarian positioning has historically delivered excess returns: during the 2008 financial crisis, Southeastern net bought 12 stocks, followed by over 80% returns over the next two years.

7. Outlook: Historical Comparison of P/E and Interest Rates

The current S&P 500 P/E of 17x and 3.8% 10-year yield create a "stock-bond valuation" conflict: the equity risk premium (ERP) is only 2.3% (= 1/17 - 3.8%), near a 15-year low, indicating stocks are relatively expensive. In contrast, the portfolio's 8x P/E corresponds to an ERP of 8.7%, which is 3.8 times the market average. Historically, when the portfolio's ERP exceeded 6% (e.g., 2000, 2008, 2012), the portfolio's subsequent 3-year annualized returns were above 12%. Additionally, the portfolio's P/V ratio is at a historically low 50-55% percentile. Over the past 20 years, this metric has fluctuated between 40-80%, and the current near-low level means purchasing power per dollar is historically high.

Year	S&P 500 P/E	10-Year Yield	Portfolio P/E	Portfolio P/V Ratio
2002	15x	3.8%	7x	55%
2012	12x	1.8%	6x	48%
2022	17x	3.8%	8x	50%
2023E	15x	3.5%	7x	45% (Assumed)

Note: Between 2002 and 2012, when the portfolio's P/V ratio fell from 55% to 48%, the portfolio's annualized return was 14%. The current 50% level suggests a similar opportunity.