4% Inflation. Stretched Valuations. Why Is the Market Still Risk-On? | Tian Yang

Episode Summary

Executive Summary: The episode centers on Variant Perception’s macro framework, which emphasizes causal, adaptive leading indicators over static models. Cheon Yang argues the economy remains risk-on because growth is resilient, liquidity supportive, and inflation manageable, while AI capex and household dissaving are propping up activity. He is bullish on equities, cautious on recession calls, and skeptical that the Fed will hike this year.

Main Topics: Causal, first-principles macro analysis (Priority: 5/5): Yang explains that useful data is not just abundant but must be selected for causal, persistent relationships—favoring sequencing and real-world logic over black-box correlations. Leading indicators and model adaptability (Priority: 5/5): Variant Perception treats leading indicators as a mosaic that changes over time; the firm uses a causal-discovery process to dynamically weight inputs that are most predictive at recent turning points. Macro risk indicator and current regime (Priority: 5/5): Their 0-100 macro risk score combines growth, inflation, policy, and liquidity. It currently remains in a risk-on regime because growth is solid, policy is not very restrictive, and liquidity is supportive. AI capex, savings behavior, and growth resilience (Priority: 4/5): Yang sees AI-related corporate spending and low household savings as key supports for U.S. growth. He argues the main risk is if hyperscalers slow capex and begin saving more. Inflation, oil, and the Fed outlook (Priority: 4/5): He thinks inflation is above target but not a major underlying problem, with headline inflation driven by energy and core pressures subdued. He doubts the Fed will hike this year, citing politics and policy priorities. Market tops, crowding, and cash settlement risk (Priority: 4/5): Yang describes a checklist for market tops that includes excesses, tightening liquidity, slowing growth, and a real cash-settlement trigger. He says current conditions are only amber warning, not a full top. AI diffusion, labor market effects, and portfolio strategy (Priority: 4/5): He expects AI to create bifurcation in labor and capital markets, compress mediocre jobs, and favor complementary assets. This outlook informs the VPX ETF, which actively tilts sectors and stocks based on capital cycle, crowding, and macro signals.

Key Arguments: Static recession indicators have failed post-2020 because the macro regime shifted toward fiscal dominance, geopolitical intervention, and more active policy management. Leading indicators should be judged by causal sequencing, not just predictive fit; data must explain why it leads and why that relationship should persist. The macro environment remains risk-on because growth is resilient, policy is only slightly restrictive in some regions, and liquidity is not as tight as many investors assume. AI capex is a major economic support because one person’s spending is another’s income; corporate dissaving plus low household savings keeps demand and incomes circulating. Recession risk is better captured by rising precautionary savings than by falling savings rates; if everyone saves more at once, income falls and the negative loop intensifies. Headline inflation may stay elevated due to energy, but core inflation looks contained because housing, wages, and small business pricing power are not showing major upside pressure. The Fed is unlikely to hike because political and sovereign considerations matter more than a small incremental tightening when reforms and credibility are the bigger priority. A true market top requires more than valuations and enthusiasm; it usually needs a mechanism such as tightening liquidity, slowed growth, and forced cash settlement. AI may not instantly destroy jobs, but it will compress the value of middling labor and deepen bifurcation between high performers and everyone else. Investing success requires both fundamental analysis and understanding market participants, positioning, and flows; playing the game matters as much as being right on the thesis.

Data Points: Macro risk indicator scale: 0 to 100 - Yang describes the firm’s composite macro score as a smooth scale used to dial risk exposure up or down. Number of macro dimensions: 4 - The macro risk indicator is built from growth, inflation, policy, and liquidity. Manually curated global inputs: about 1,000 - The firm says it tracks roughly a thousand theoretically leading, low-revision inputs globally. Fed hike expectation: 0 hikes expected this year - Yang says he does not think the Fed will hike in the current year. Inflation LEI forecast: 4% - He says their headline inflation leading indicators point to around 4% inflation. Typical top-to-policy lag: 6 to 9 months - He notes major market tops often require roughly six to nine months of monetary tightening before stress becomes visible. Somme rule threshold idea: 50 bps rise from a 3-year low - He explains the Somme rule as a recession proxy based on unemployment rising quickly from recent lows. Current number of U.S. states meeting the Somme rule: 9 - A chart discussed in the interview shows about nine states currently meeting the state-level Somme rule criterion. Consumer finance-labor divergence period: post-COVID - He says personal finances and job availability measures diverged sharply after the pandemic, reflecting price-level resets. Relative role of households vs corporates: both low savings rate / high dissaving - He argues U.S. households and corporates are simultaneously spending more and saving less, supporting growth.

Pivotal Quotes: "Data is easier to access and more available than ever before. The key now is how creatively you use these inputs, and most importantly, what you choose to leave out." — Jack Foorhand quoting the discussion prompt: Introduces the episode’s core theme: filtering noise and focusing on meaningful inputs. "What are the causal reasons that data is statistically meaningful?" — Chian Yang: Explains Variant Perception’s first-principles approach to model-building and indicator selection. "We think a lot about causal relationships, which actually just means sequencing. What moves first to then cause something else to move?" — Chian Yang: Defines how the firm distinguishes true leading indicators from coincidental data.

Implications: Listeners should expect a slower-moving but still resilient macro regime, with AI and policy coordination keeping equities supported. The biggest risks are a capex slowdown, cash-settlement events, or a genuine policy/liquidity squeeze rather than classic recession signals.