Goldman's Hatzius and Snider on the Outlook for 2026

Episode Summary

Executive Summary: Goldman’s Jan Hatzius and Ben Snider argued that 2026 should still feature solid U.S. growth, easing inflation, and moderate Fed cuts, while markets remain driven more by earnings than AI hype. They pushed back on claims that AI has meaningfully boosted GDP, emphasized broad-based corporate profit strength, and said market concentration reflects earnings concentration rather than pure speculation.

Main Topics: 2026 macro outlook: growth, unemployment, and productivity (Priority: 5/5): Hatzius forecast solid GDP growth alongside only modestly higher unemployment, arguing that faster productivity growth can allow the economy to expand without sharply lowering joblessness. He expects AI to matter more in coming years than it has so far. Equity market resilience and earnings strength (Priority: 5/5): Snider said 2025’s strong market performance was underpinned by robust earnings, not just multiple expansion or AI enthusiasm. He highlighted that even excluding mega-caps, broad U.S. stocks delivered strong returns and earnings growth. AI’s real economic impact vs. market narrative (Priority: 5/5): Both guests distinguished between AI capex as a market trade and AI as a measured GDP driver. Hatzius argued AI investment added very little to measured GDP because much of it is imported or counted as intermediate inputs. Inflation, tariffs, and pricing pass-through (Priority: 4/5): Hatzius said tariffs contributed about 50 basis points to core PCE inflation in 2025 and should behave like a price-level effect that fades in 2026. He also said the latest CPI was somewhat understated due to shelter and timing distortions. Valuations, concentration, and bubble risk (Priority: 4/5): Snider said valuations are elevated but justified in part by better earnings expectations and healthier growth. He argued the current market is less speculative than dot-com-era bubbles, though he watches jobless claims, Fed tightening, and narrative excess for warning signs. China, housing, and sector-specific slowdowns (Priority: 3/5): Hatzius noted China’s exports remain resilient despite tariffs, but domestic housing is still a major drag. In the U.S., housing is relatively unimportant to the 2026 outlook, with prices expected to move mostly sideways. Fed path and data uncertainty (Priority: 3/5): The guests discussed a likely Fed easing path toward neutral, but noted uncertainty from delayed data due to shutdown effects. Hatzius said a new Fed chair should not materially alter the near-term forecast.

Key Arguments: U.S. growth can remain strong even with unemployment around 4.5% because productivity growth has accelerated to about 2% underlying trend since the pandemic. AI has not been a major measured contributor to GDP so far; much of the spending shows up as imports or intermediate inputs rather than domestic investment. 2025 equity gains were supported by real earnings growth, including 12% S&P 500 earnings growth in Q3 and about 10% median earnings growth excluding mega-caps. The broader market, not just mega-cap tech, has performed well; the S&P 490/493 has returned about 15% annually for the last several years. Tariffs have so far been more of a price-level shock than a sustained inflation spiral, with about 50 basis points of pass-through into core PCE. Margins held up better than feared because companies offset tariff pressure through pricing, supplier negotiations, supply-chain changes, and cost cuts. Valuations are high but not automatically a sell signal; they are better viewed as potential energy that needs a catalyst. The current AI rally differs from the dot-com era because investors are focusing on present earnings rather than paying purely for distant productivity dreams. Productivity gains from AI could raise frictional unemployment in the short run, but history suggests technology does not permanently raise aggregate unemployment over long horizons. China’s external growth remains supported by competitive manufacturing and export re-routing, even as domestic property weakness continues to drag on GDP.

Data Points: Recording date: December 19, 2025 - Used to frame the discussion of late-2025 CPI, Fed policy, and 2026 forecasts. Goldman U.S. real GDP forecast for 2026: 2.6% - Hatzius said this forecast likely overstates underlying trend somewhat because shutdown effects distort quarterly timing. Unemployment rate forecast: 4.5% - Hatzius said unemployment is expected to stay roughly flat rather than fall despite solid GDP growth. Underlying productivity growth since pandemic: ~2.0% - Hatzius said the last five years have seen about 2% underlying trend productivity growth. Prior-cycle productivity growth: ~1.5% - Referenced as the pre-pandemic comparison point. S&P 500 Q3 2025 earnings growth: 12% - Snider cited aggregate third-quarter earnings growth for the S&P 500. Median S&P stock earnings growth excluding mega-caps: ~10% - Used to show that profit growth was broad-based, not only concentrated in mega-cap tech. Top 10 stocks market cap share: Over 40% - Snider said the top 10 names now account for more than 40% of S&P market cap. S&P 493 annual returns: About 15% per year - Snider said the broad non-mega-cap market has returned around 15% for each of the last three years. AI contribution to measured GDP growth: About 20 bps over 3-4 years; near zero in the last year - Hatzius argued AI capex has had little effect on measured GDP after accounting for imports and input classification. AI contribution to 2025 GDP growth (misconception): Some estimates claimed 50% of growth - Hatzius pushed back on viral charts overstating AI’s macro role. Tariff pass-through to core PCE inflation: About 50 bps - Hatzius estimated 2025 tariff effects added roughly half a percentage point to core PCE inflation. Initial expected tariff pass-through: About 100 bps - Goldman had initially thought the pass-through could be roughly double the realized amount. Core CPI for November report: 2.6% y/y - Discussed in relation to the late-2025 CPI release and shelter data distortion. S&P 500 target for 2026: 7,600 - Snider disclosed Goldman’s official 2026 S&P 500 target. Expected AI productivity boost in 2026 earnings model: Under 0.5% - Snider said Goldman is modeling a small but growing AI productivity effect for next year. Top 10 stocks share of earnings: About one-third - Snider noted earnings concentration has risen to roughly match market-cap concentration. Fed cuts forecast: Two cuts in 2026 - Hatzius expects two rate cuts, penciled in for March and June, bringing policy to neutral. Neutral Fed funds range: 3.0% to 3.25% - Hatzius’s estimate of where policy rates should settle. China export growth: Up 5% to 10% - Hatzius said overall Chinese exports remain healthy despite weaker shipments to the U.S. China exports to U.S.: Down 25% to 30% y/y - Trade diversion and tariffs reduced direct exports to the U.S. China property drag on GDP in 2025: -2.0 percentage points - Goldman’s China team estimate of direct and indirect property effects on 2025 GDP. China property drag on GDP in 2026: -1.5 percentage points - Expected continued but smaller property-related drag next year. S&P 500 market concentration study horizon: 100 years - Snider said Goldman examined a century of U.S. market concentration and found no close modern precedent. Probability of January Fed cut: Less than 20% - Mentioned as current market pricing at the time of recording.

Pivotal Quotes: "AI shouldn't eliminate them, it should elevate them." — Podcast sponsor copy / Palantir: Opening ad message framing AI as worker-enhancing rather than labor-replacing. "The market was correctly looking forward back then and saying earnings are going to grow." — Ben Snider: Explaining why rising concentration in the S&P 500 can be rational if earnings follow market cap leadership. "What you really need to look at is the imports as well." — Jan Hatzius: Why AI-related investment has contributed little to measured GDP despite massive spending headlines.

Implications: For listeners, the message is that 2026 may still reward economically solid, earnings-backed assets rather than pure AI stories. The biggest risks are inflation surprises, labor-market deterioration, and valuation excess—not necessarily AI itself.