I Correctly Predicted Nvidia Would Overtake Apple in Stock Buybacks and Dividends. Here's the Better Buy Now.

Nvidia authorized a $150 billion increase to its repurchase program, taking total buyback authorization to $235 billion, above Apple's newly authorized $100 billion program. Nvidia returned $19.7 billion through buybacks and $6 billion in dividends in its latest quarter, and management said it returned 60% of first-half fiscal 2027 free cash flow to shareholders versus a 50% target. The article argues Nvidia offers the more compelling AI-driven value proposition, trading at 24.6x forward earnings versus Apple's 38.3x, while guiding for 70% fiscal 2028 revenue growth.

Analysis

The capital-return framing is directionally supportive for NVDA but is not itself a fundamental catalyst: repurchases create value only when executed below intrinsic value and can mask a maturing reinvestment opportunity. The more consequential signal is that management sees enough internally generated cash to fund both the Rubin transition and shareholder returns; this lowers balance-sheet risk versus AI infrastructure buyers, but does not eliminate the risk that customer capex normalizes before those systems are absorbed. Near-term, the announcement may support NVDA on retail flows; over 1-3 months, Blackwell/Rubin delivery cadence, networking attach rates, and hyperscaler capex commentary remain the actual earnings drivers.

The proposed NVDA-versus-AAPL valuation comparison is likely based on inconsistent or stale forward estimates and should not drive positioning without checking consensus fiscal-year alignment. AAPL's multiple reflects a lower-volatility installed-base and services cash-flow franchise, while NVDA's earnings are exposed to concentrated customers and a semiconductor supply cycle. The better relative expression is not simply long NVDA/short AAPL: Apple can monetize AI through replacement cycles and services without carrying data-center depreciation, whereas NVDA faces the risk that customers redirect incremental AI spend toward proprietary silicon.

Second-order pressure remains on AMZN, MSFT and GOOG: AI capex depresses reported FCF and raises depreciation over the next 6-18 months, creating a valuation dispersion opportunity if AI revenue ramps lag capacity deployment. Conversely, any evidence of inference monetization or easing capex intensity would compress the NVDA premium and favor hyperscalers. Falsify the NVDA constructive thesis on a material cut to customer capex plans, a decline in networking attach, or gross-margin guidance below expectations during the next platform ramp.

Key Decisions for Investors

• Maintain an overweight NVDA versus SOXX over the next 1-3 months only if hyperscaler earnings continue to show accelerating AI capex and no delivery bottlenecks. Use a 10-12% downside stop or exit on a gross-margin guide-down; upside is 15-20% if Rubin revenue and networking mix exceed consensus.
• Initiate a modest long NVDA / short AAPL pair only after validating comparable forward EPS vintages and if the relative valuation spread remains near the article's claimed extreme. Size for a 2:1 upside/downside profile over 3-6 months; cover the AAPL short if iPhone upgrade data or Apple Intelligence adoption materially improves.
• Watch for a 6-18 month long basket in AMZN, MSFT and GOOG against a reduced NVDA overweight once AI revenue growth begins catching up with depreciation growth. Trigger on two consecutive quarters of stable-to-lower capex-to-revenue ratios or explicit management commentary that incremental AI workloads are profitable.
• Do not add exposure solely on the repurchase authorization. Confirm actual quarterly share-count reduction versus stock-based compensation and strategic-investment outlays; a large authorization with limited net dilution reduction is not an EPS catalyst.