Brent Donnelly on the Fed, Inflation, and Why 2% No Longer Matters

Episode Summary

Executive Summary: The episode explores Brent Donnelly’s short-term macro trading framework: stay agnostic, use stop-losses, and structure trades around regime shifts rather than fixed forecasts. He discusses labor-supply versus demand shocks, gold/TLT positioning, FX as a policy exhaust valve, Canada’s weakening economy, Fed politicization and inflation risk, and how LLMs are becoming practical tools for trade discovery, sizing, and headline interpretation.

Main Topics: Agnostic macro trading and stop-loss discipline (Priority: 5/5): Donnelly argues that the best trading edge is flexibility: form a view, size it with risk limits, and exit quickly when data or price contradicts the thesis. He rejects perma-bear or ideological investing in favor of reassessment triggers. The failed August reacceleration thesis (Priority: 5/5): He explains how he briefly framed weak labor data as a labor-supply shock that could allow a growth reacceleration, then abandoned it when multiple labor market numbers stayed weak. The trade was designed as a hedge across scenarios using GLD calls and TLT puts. Fed independence, politicization, and inflation targeting (Priority: 5/5): Donnelly argues the Fed is moving toward less independence and that inflation targeting is effectively dead or hollowed out. The key risk is not slower growth but rising inflation while the Fed cuts, which could trigger a bond vigilante response. FX as an exhaust valve for U.S. tariff and capital-flow dynamics (Priority: 4/5): He revisits the idea that FX absorbs policy shocks, especially tariff-related uncertainty and foreign hedging flows. He says the dollar’s next leg lower likely needs a cyclical U.S. slowdown because prior structural dollar-selling has already run its course. Canada’s leverage, housing stress, and dollar-CAD trade (Priority: 4/5): He sees Canada as more vulnerable than the U.S. due to variable-rate mortgages, high household leverage, and rising unemployment. He uses CAD weakness as a proxy for broader macro deterioration, with CIBC as a poster child for housing-linked stress. LLMs as trading assistants (Priority: 4/5): Donnelly describes using LLMs for headline digestion, event interpretation, trade structuring, payoff analysis, and quick data-mining. He stresses they are idea-generation tools, not truth machines, but are already useful for experienced traders. Avoiding narrative traps and using imagination (Priority: 3/5): He emphasizes that consensus forecasts are often too narrow, and that good trading requires imagining regime changes larger than the median forecast. Technicals and price levels help define when a thesis is invalidated.

Key Arguments: Short-term trading works best when the trader is highly Bayesian: new contradictory evidence should quickly reduce conviction. The August labor-market thesis was plausible as a labor-supply shock, but repeated weak labor prints made a reacceleration harder to sustain. A gold-plus-bond-bearish structure fit the macro setup because a dovish Fed cutting into stronger growth would be inflationary and bullish for gold. The Fed is becoming more politically influenced, but the truly dangerous scenario is not disinflation; it is inflation rising while the Fed still cuts. FX can act as a release valve for global policy shocks, but the structural dollar-selling from Liberation Day has largely run its course. Canada’s economy looks more fragile than the U.S. because leverage is higher, mortgages reprice faster, and unemployment is rising. LLMs are useful because they scale an analyst’s pattern recognition across headlines and datasets, especially in obscure or event-driven situations. Consensus forecasts and media sentiment are structurally too bearish or too clustered, so traders need imagination plus explicit reassessment triggers.

Data Points: Fed above-target inflation streak: 53 months - He says U.S. inflation has been above target for 53 consecutive months, suggesting inflation targeting has lost credibility. Labor market numbers in one week: 4 weak prints in a row - Used to explain why he abandoned the labor-supply/reacceleration thesis. Gold trade return: 5x realized; 9x peak - He bought gold calls during the reacceleration setup and sold them at 5x, noting they eventually reached 9x. TLT put outcome: Went to zero - His bearish-bond hedge via TLT puts expired worthless. Inflation target: 2% - The official target he says markets and policymakers no longer treat as binding. Canadian unemployment rate change: 5.0% to 7.1% - He cites this rise as evidence of worsening Canadian labor-market stress. Canadian housing example: $900K pre-builds trading at $620K - Illustrative example of price pressure in Toronto condos. Dollar-hedge cost for foreign investors: ~4% per year - He says being short dollars while long U.S. equities can cost about 4% annually. MSCI World U.S. weight: 72% USA - He uses this to explain foreign portfolio managers’ career-risk constraints. Ontario teachers’ FX sales: $50 billion - Example of large post-Liberation Day dollar selling via hedging. Fed hike prediction dinner: 17 hikes occurred; one guest predicted 6 - He recounts a 2021 dinner where the hawkish outlier was still far too low. Typical Fed hike size: 25 basis points - Referenced in the 2021 forecasting anecdote. Tail risk estimate for inflation-plus-cut regime: 15% chance - His rough probability for U.S. inflation rising while the Fed cuts. Volatility example: September 12 expiry - He described buying gold and TLT options around an August 30 trade with multiple data releases before expiry. Short-call-spread idea size: Risk $20,000 - He described using LLMs to structure option trades with defined downside. LLM example horizon: 7 days to expiry - Used in the example prompt for call spreads on a speculative short idea.

Pivotal Quotes: "I think honestly, like inflation targeting in the U.S. is kind of dead." — Brent Donnelly: His view that the Fed no longer truly adheres to a strict 2% inflation regime. "I will never be a perma bear because that is the absolute biggest leak in finance." — Brent Donnelly: He explains why flexible, stop-loss-driven thinking beats a fixed bearish worldview. "LLMs are idea generation machines, but not calculation or right answer machines." — Brent Donnelly: His practical description of how he uses AI tools in trading.

Implications: Listeners should expect more regime-driven, less rules-based policy and market behavior. For traders, the edge is agility: use price/data triggers, respect FX and rates as transmission channels, and treat LLMs as fast research assistants, not substitutes for judgment.