Goldman ISSUES Warning for the Total Stock Market | 2027 will SUCK.
Overall Thesis
Goldman Sachs warns that S&P 500 earnings growth is artificially inflated by low depreciation drag in 2026, but 2027 will see significant headwinds as depreciation write-offs from AI capex investments offset roughly half of expected earnings growth.
Narratives
Kevin relays Goldman Sachs' warning that AI-related depreciation drag and fading 'other income' from equity gains could slow S&P 500 earnings growth starting in 2027, creating headwinds over the next 12 months. Despite this caution, he remains bullish on the index for the next six months, citing low investor leverage and absence of euphoria as bullish signals.
Key Arguments
- Goldman warns hyperscaler depreciation will offset nearly half of the capex-driven earnings boost starting in 2027
- Over $150 billion in 'other income' from equity investment gains lifted S&P 500 earnings in Q2 2026 but may diminish in 2027
- Goldman Private Bank clients are at only the 34th percentile for leverage, suggesting no euphoria and room for further gains
- Current earnings strength depends on semiconductor margin expansion, AI capex, and private investment gains that may not be sustainable
Predictions (1)
Kevin argues Nvidia is part of a 'hardware 2.0' rally thesis, believing chip and hardware stocks have lagged behind the profitability implied by massive capex flowing from companies like Google and SpaceX. He views Nvidia's valuation as cheap relative to fundamentals, though he flags eventual risk when hardware earnings growth turns negative.
Key Arguments
- Capital raised by Google and SpaceX is being funneled into hardware purchases from Nvidia and ASIC makers
- Hardware stocks have lagged behind near-term EPS estimate trajectories, implying an undervaluation
- Chip investments (e.g., his own company's GPUs) have doubled in value over the past year
Predictions (1)
Kevin notes AMD is benefiting from AI infrastructure spending flowing from hyperscalers and views AMD as part of the broader undervalued hardware trade, though he notes it's not as central to Nvidia's exclusive partnerships.
Key Arguments
- Capex raised by companies like Google and SpaceX partly flows to AMD as an ASIC/GPU supplier
- AMD's stock has skyrocketed, benefiting Meta's warrant position, reflecting strong hardware demand
Predictions (1)
Kevin highlights Meta's call options (warrants) on AMD as a source of significant 'other income' boosting earnings, and considers Meta undervalued on a PEG basis despite a recent rally.
Key Arguments
- Meta holds call options on AMD exercisable at a penny, generating large other-income gains as AMD's stock rose
- Meta's valuation is still viewed as cheap on a PEG basis even after a recent rally
Predictions (1)
Kevin includes Broadcom among hardware names he believes are undervalued relative to profitability as part of his 'hardware 2.0' thesis.
Key Arguments
- Broadcom is grouped with Meta and Nvidia as trading under a 1.0 PEG ratio, implying undervaluation
Predictions (1)
Kevin briefly suggests Oracle has likely bottomed as part of the broader hardware/AI infrastructure recovery narrative.
Key Arguments
- Oracle is cited as an example of a hardware-adjacent stock that has 'probably bottomed'
Predictions (1)
Kevin discusses Lennar's four consecutive years of negative EPS growth and the resulting steep stock decline, expressing uncertainty about whether it has bottomed despite a same-day 4% gain.
Key Arguments
- Lennar has posted four straight years of negative EPS growth, correlating with a steep stock decline since 2024
- The stock was up 4% on the day discussed, prompting speculation about a possible bottom
Hedges & Caveats
- Goldman Sachs framing these as 'headwinds' rather than bearish calls
- Analysis based on forward-looking depreciation estimates that may not materialize as predicted
- Host acknowledges earnings valuations are 'in line with 10-year average' on forward P/E basis
- Discussion of GPU price appreciation as unsustainable suggests speculative asset valuations
- Commentary focuses on 2027 specifically, implying near-term (12-month) caution rather than long-term bearishness