Kallum Pickering Warns of AI Boom Bust Risks

Peel Hunt economist Kallum Pickering warns that high interest rates and heavy AI investment are concentrating capital and increasing boom-bust risks. He also discusses Fed policy, U.S. debt and the AI race with China, while expecting inflation to move decisively lower next year.

Analysis

The investable tension is between disinflation that could support duration and a structurally higher term premium from fiscal borrowing and AI-related capital demand. These forces need not resolve together: softer inflation may bring expected policy rates down while Treasury supply keeps long yields elevated, limiting the benefit to long-duration equities and bonds. AI spending also has a two-sided effect—productivity could lower unit costs over time, but near-term competition for power, data-center capacity, and advanced chips can sustain bottlenecks and increase the cost of deploying capital. That makes the AI trade vulnerable to a shift from growth narratives to returns-on-invested-capital scrutiny. China competition and policy restrictions could reinforce duplicated supply chains rather than deliver near-term cost relief.

Over days to weeks, rate-sensitive assets remain exposed to inflation surprises and Treasury auctions. Over 1–3 months, the key test is whether core inflation and wage data validate disinflation without a material rise in long-end yields. Over 6–18 months, productivity gains could broaden beyond a narrow set of AI beneficiaries; failure to translate investment into measurable efficiency would raise downside risk for concentrated equity leadership. The interview’s inflation outlook is a forecast, not evidence of a realized trend. With no market levels, valuation data, or positioning supplied, conviction is insufficient for an outright duration or AI-sector trade.

Key Decisions for Investors

• No immediate directional macro trade: avoid adding unhedged long-duration exposure solely on the disinflation call. Reassess after successive core-inflation readings confirm cooling and Treasury supply is absorbed without a persistent rise in long-end yields.
• Use a relative-value watchlist rather than a broad AI short: consider reducing cap-weighted US growth concentration or pairing it with broader/equal-weight equity exposure if AI leaders weaken while earnings breadth improves. Falsifier: continued earnings upgrades and broadening returns on AI investment.
• Track real yields, the 10-year term premium, Treasury auction demand, and power/data-center constraints. A rise in long yields despite softer inflation would argue fiscal and capital-demand effects are dominating and would undermine the duration thesis.
• Treat AI capex as a potential source of both future productivity and near-term bottleneck inflation. Look for evidence in company guidance and productivity data that spending is translating into lower costs; absent that evidence, avoid extrapolating investment announcements into durable profit growth.