Mar 21, 2024 · 1h 26m · bg2-pod
Ep5. Stock Comp, AI Cold War, Valuations for LLM | BG2 with Bill Gurley, Brad Gerstner, & Bob Mylod · Bg2 Pod
gold bands on the timeline = statements, start to end. Hover to read, click to jump. CC turns on captions
In Episode 5 of the BG2 Podcast, co-hosts Brad Gerstner and Bill Gurley, along with guest Bob Mylod, analyze the mechanics and governance pitfalls of Stock-Based Compensation before evaluating major technology developments, including Apple-Google AI rumors, TikTok legislation, and LLM market valuations.
How this conversation actually went
Every chapter scored 0–10 on four independent dynamics. Hover any point for the reasoning behind the score. Brad and Bill hold 74.1% of the talking time here. How this is scored →
speaking balance: gold is Brad and Bill, purple is the guest (3 minute bins)
Mylod forcefully dismisses the conventional boardroom focus on dilution percentages, arguing that dilution percentages do not appear on W-2s and obscure true compensation costs.
Hardest push from Brad and Bill ▶ 13:57 Gurley challenges the notion that stock options lack intrinsic valueGurley directly interrupts the framing that strike-price options have no value by citing how synthetic option creation in Black-Scholes proves at-the-money options trade with real value.
Biggest teaching moment ▶ 11:15 Mylod explains how FAS 123 drove companies toward RSUs and pro-forma distortionsMylod provides a first-hand historical account from the dot-com crash of how accounting mandates forced public CFOs to abandon options and begin pro-forming out equity costs.
Brad and Bill hold their own ▶ 1:19:32 Gerstner reframes generative AI revenues from ARR to ERRGerstner demonstrates proprietary market expertise by banning the term ARR for LLM business models and coining ERR to reflect zero switching costs and low revenue durability.
the scores for every segment, with the reasoning behind each
| Chapter | Topic | Brad and Bill as informed peer | Guest teaching | Guest disagreement | Brad and Bill pushing back | Why |
|---|---|---|---|---|---|---|
| The Historical Context and Evolution of Stock-Based Compensation | 8 | 2 | 1 | 1 | Bill Gurley provides deep historical framing on Silicon Valley stock options, the 1999 dot-com fallout, ISS governance, and the mathematical transition to RSUs under Black-Scholes models. | |
| Priceline's Dot-Com Collapse, FAS 123, and the Accounting Shift to RSUs | 7 | 7 | 2 | 3 | Bob Mylod educates the hosts on the unintended consequences of FAS 123 and Sarbanes-Oxley during Priceline's restructuring, while Gurley pushes back to clarify that at-the-money options do hold traded market value. | |
| Pro-Forma Earnings Manipulation and the AI Compensation Arms Race | 8 | 6 | 3 | 2 | Mylod exposes how tech companies treat SBC as zero-cost in pro-forma earnings while inflating engineering compensation, contrasting this with Booking's share count reduction strategy. | |
| Evaluating SBC Metrics, Board Governance, and Market Standards | 8 | 6 | 2 | 2 | Gurley and Gerstner dissect the upward ratcheting effect of comp consultants and peer benchmarking, while Mylod demonstrates how annual dilution compounds against real free cash flow. | |
| Startup Equity Practices, Compensation Rules, and Guest Wrap-Up | 7 | 4 | 1 | 1 | The discussion turns to startup compensation rules from Series A to late stage, citing Gary Steele's dramatic reduction of equity participation at Splunk. | |
| BG² Podcast Interlude Graphic Bumper | 8 | 0 | 0 | 2 | Gerstner and Gurley analyze rumors of Apple partnering with Google Gemini, arguing Apple will maintain its own SLM for personal assistant functionality on handset devices. | |
| The TikTok US Ban Legislation and the US-China AI Cold War | 8 | 0 | 0 | 4 | Gerstner and Gurley debate the TikTok divestiture bill and sovereign AI dynamics; Gurley challenges ByteDance's data assurances and points out structural asymmetry in China market access. | |
| Microsoft's Inflection AI Deal, LLM Switching Costs, and ERR vs. ARR | 9 | 0 | 0 | 2 | The hosts analyze Microsoft's Inflection deal as a yellow flag for AI valuations, introducing the concept of Experimental Run Rate (ERR) vs. ARR and arguing that LLM switching costs approach zero. |