140 – Craig Torres on Fed Transparency, Automation, and the Bear Stearns Bailout

Episode Summary

Executive Summary: Craig Torres and David Beckworth discuss the Fed’s pre-crisis blind spots, the Bear Stearns rescue and ensuing transparency battles, Jay Powell’s push for clearer communication, debates over interest-rate rules and yield-curve inversion, plus how automation and public R&D reshape labor and innovation. The conversation centers on humility, institutional openness, and how policy is explained to the public.

Main Topics: Fed blind spots before the Great Recession (Priority: 5/5): Torres says community-affairs staff and credit activists warned about predatory lending, but the Fed’s siloed structure kept that information from monetary and regulatory decision-making. He also notes Richmond Fed officials argued policy was too loose, but low inflation obscured the warning signs. Bear Stearns bailout and Fed transparency litigation (Priority: 5/5): Torres recounts Bloomberg’s FOIA lawsuit over the Bear Stearns portfolio and discount-window borrowing. The eventual disclosure showed some assets were not investment-grade, underscoring that the Fed took taxpayer credit risk and that transparency can be enforced externally. The Fed’s evolution in communication and openness (Priority: 5/5): He contrasts the opaque 1980s Fed with today’s frequent statements, press conferences, reports, and congressional engagement. He argues Jay Powell has accelerated transparency and made the institution more accessible to the public and stakeholders. Powell’s leadership style and public-facing strategy (Priority: 4/5): Torres portrays Powell as pragmatic, plain-spoken, and more willing to embrace debate, uncertainty, and basic communication than prior technocratic Fed leaders. He sees this as helpful amid Trump-era criticism and recurring calls for audits. Monetary policy uncertainty, neutral rates, and yield curves (Priority: 5/5): The discussion covers rate hikes, inflation running below expectations, uncertainty about the neutral rate, and debate over whether yield-curve inversion is a meaningful warning signal. Torres is skeptical of overreading the curve and wants more humility about unobserved variables. Automation, AI, and labor reorganization (Priority: 4/5): Torres argues automation is often overstated as a job destroyer. In factories and retail, it tends to reorganize work, create complementary roles, and raise demand for technical skills, data work, and human oversight rather than eliminating labor outright. Government-backed innovation and DARPA-style research (Priority: 4/5): He defends public investment in high-risk, high-return R&D, citing DARPA, the Human Genome Project, and startup ecosystems. He argues private capital often won’t fund speculative projects, while government support can catalyze major breakthroughs and local clustering.

Key Arguments: The Fed ignored important warnings before the crisis because its community-affairs, regulatory, and monetary-policy channels were too siloed to influence decisions. The prevailing belief in securitization and risk dispersion made policymakers overconfident that risky lending was being safely distributed to sophisticated investors. Bloomberg’s litigation showed the Fed’s Bear Stearns interventions were not as clean as officials claimed, since the portfolio included non-investment-grade securities. Fed transparency has improved enormously over decades, but much of the progress came only after pressure, court orders, and media scrutiny. Jay Powell’s communication style matters because technocratic messaging alone has not won public trust or insulated the Fed from political attacks. Policy debates should be explained in plain language; Torres suggests the Fed could frame policy as stabilizing nominal income growth rather than using abstract targets. He is skeptical that the yield curve alone should drive policy, suggesting bond demand may reflect low inflation and global savings rather than a reliable recession signal. Automation usually changes the composition of work rather than simply reducing headcount; firms often create new technical and supervisory roles around machines. Companies can help solve skill shortages by investing in training and community colleges instead of assuming labor supply is fixed. Public R&D is essential for speculative projects with uncertain commercialization, because private investors often cannot justify the risk or see a clear exit path.

Data Points: Congressional meetings by Powell: more than 70 - Torres says Powell has had intensive contact with Congress since becoming chair. Press conferences: 8 per year - Powell said he would hold a press conference after every FOMC meeting rather than only quarterly. Technical worker pay at precision machining facilities: $48,000 starting; up to $70,000 - Torres cites community-college-trained technicians in advanced manufacturing. Fed R&D outlays trend: Peaked in the 1960s - Torres refers to a chart showing federal R&D spending was highest in the 1960s and later declined. Baltimore/Wall Street-style transparency example: Fed repo signals via "three-day system" - He recalls late-1980s market participants inferring policy from repo operations before modern statements existed. Yield curve policy forecast: Three rate hikes in 2019 - Torres says markets were skeptical of the Fed’s projections at the time of the interview. Employment effect at BMW/automation: More humans near the end of the production line - He describes robots handling welding while humans do quality control and sensory inspection. Cancer therapy success rate example: 60% - Torres mentions a startup testing chemo against patient-derived cancer cells and notes that a 60% survival rate typically reflects treatment efficacy roughly 60% of the time.

Pivotal Quotes: ""Here's our goals. Here's our instrument. Here's what we're worried about. Here's how we respond."" — Craig Torres: He proposes a simple, classroom-friendly way for the Fed to explain its mission and tools to the public. ""I think the Fed should be much more basic in its communication"" — Craig Torres: Torres argues the Fed’s technocratic style has not built enough public trust and should be simplified. ""The machine only knows what it knows."" — Craig Torres: He uses this idea to explain why human workers are still needed alongside automation to detect nonstandard events.

Implications: The Fed’s credibility depends not just on policy, but on transparency and plain-language explanation. Automation will likely reshape rather than erase jobs, and public R&D remains crucial for breakthroughs private markets avoid.