Bloomberg Intelligence: Nvidia to Invest in MediaTek (Podcast)

Nvidia plans to invest $3.5B in MediaTek, leveraging NVLink Fusion and NVHBM to improve chip-to-chip communications for data centers and reinforce its AI hardware platform standards. On deal activity, Eli Lilly is set to buy Merida Biosciences for up to $2.88B in cash to expand autoimmune/allergy pipelines, while KKR is expected to earn a $3.3B windfall from selling USI Insurance Services to Aon for $17B, generating about $2B of adjusted net income and a 3.4x return on invested balance-sheet capital.

Analysis

This is less about headline M&A and more about control points. NVIDIA is effectively buying deeper penetration into the server stack, which should widen switching costs around its interconnect standards and make it harder for hyperscalers to architect around the platform over the next 6-18 months. The near-term read-through is strongest for NVDA; the second-order losers are alternative accelerator and custom-silicon ecosystems that depend on open interconnects, especially if this becomes a template for broader OEM adoption.

For Lilly, the strategic value is duration, not this year’s EPS. Management is using obesity-generated cash flow to buy optionality in autoimmune/allergy, which can support a premium multiple if it keeps compounding outside the obesity franchise. The market may underappreciate that the real bull case is not one asset but a repeatable capital-allocation engine that extends Lilly’s growth runway beyond the current consensus window.

AON’s transaction matters most as a competitive signal: scale is being rewarded in middle-market brokerage, and that can slowly pressure smaller intermediaries on pricing and retention. KKR’s realized gain is a clean positive for private-markets sentiment, but it is a one-off unless exit markets stay open; the better read-through is to BX/APO/CG on whether monetization is improving broadly. The Anthropic IPO angle is a liquidity event risk for late-stage private tech, but it is more a 1-3 month capital-allocation headwind than a fundamental threat to public AI leaders.

Key Decisions for Investors

• Long NVDA / short AMD over 1-3 months: express the view that this is a standards-lock-in story, not just incremental revenue. Falsify if hyperscaler capex shifts to open/custom architectures or AMD books a material server-design win in the next earnings cycle.
• Buy NVDA on any post-news pullback; treat the investment as a signal that the platform moat is widening rather than a one-quarter financial event. Best risk/reward if the stock gives back the first reaction and holds SOXX relative strength.
• Accumulate LLY on weakness for a 6-18 month horizon; Merida is an option on pipeline duration, not an immediate earnings driver. Falsify if obesity growth decelerates faster than expected or early autoimmune data disappoints.
• Long AON on pullbacks, or pair long AON / short a smaller-brokerage basket over 3-6 months if you want a relative-value expression. Thesis breaks if integration costs or retention issues dilute the expected middle-market leverage.
• Do not chase KKR on the print; use the realized gain as a watch item for BX/APO/CG sentiment. If exit markets stay open for another quarter, revisit a sector-long; otherwise it is mostly a one-off earnings pop.