Lessons From The Great Financial Crisis | "Clear Takes" with Vincent Daniel and Danny Moses

Summary

Danny Moses and Vincent Daniel describe their approach as selectively contrarian, not automatically bearish. They believe passive investing, momentum strategies, and volatility-targeting funds favor liquid assets and can leave less-followed securities mispriced. Their process is to conduct fundamental, balance-sheet-focused research, be patient, and look for a catalyst—especially for long positions. They caution that undervaluation alone is not a reason to buy when capital flows are absent.

They see the current AI investment boom as combining features of the dot-com bubble and the Global Financial Crisis (GFC). Like the dot-com era, they worry that a compelling long-term trend could lead to excessive infrastructure investment. Like the GFC, leverage and financing structures could amplify a reversal. Their central concern is whether AI projects can keep attracting capital at affordable rates and ultimately generate enough revenue and cash flow to justify the spending. They see substantial risk outside traditional banks, particularly in private equity and other financing for AI infrastructure, but do not characterize current banking risk as systemic.

Their questions about the AI sector include data-center power availability, GPU useful life and depreciation, customers’ ability to pay for AI services, and whether AI businesses can develop pricing power. They also question how much AI-company growth is supported by customers that themselves have negative free cash flow. They acknowledge the strong demand case for AI while stressing that demand alone does not establish attractive economics.

Investment Approach and Positioning

• Start with market structure: Account for passive flows, momentum, and volatility-targeting strategies, which the speakers say can make markets persistently favor liquid, popular assets.
• Do deep, bottom-up work: Examine company fundamentals, balance sheets, debt, and where a company’s debt trades—not just its stock.
• Be selective and patient: Avoid buying every apparently cheap asset; allow time for a thesis to work and look for a catalyst.
• Manage trading risk: They said they used hard stops on a CoreWeave short and closed the position after seeing that negative sentiment had become crowded. They viewed this as a trading decision, not evidence that the business model had changed.
• Favor select opportunities over broad bets: They described energy and gold as areas of interest and discussed selective opportunities in housing and consumer cyclicals. They said they were not fully invested.

Market, Macro, and Risk Discussion

• AI capex and financing: The speakers cited roughly $500 billion of data-center and GPU securitization activity involving major financial and technology firms as an example of the blend of innovation and financing risk. They questioned how many data centers will ultimately be funded, how long GPUs will remain economically useful, and whether financing will remain available at reasonable costs.
• Private capital and banks: They said banks’ exposure may be more concentrated in transaction activity and revenues, as well as lending to investment funds, than in direct holdings of AI assets. They see potential write-downs or defaults among private-equity funds financing AI projects and warned that tighter credit terms or reduced bank credit lines could create a chain reaction. They said credit problems did not currently look systemic to them, though commission and transaction revenues could be vulnerable.
• Sovereign debt and liquidity: Moses argued that global debt of roughly $350–400 trillion must be refinanced on an approximate five-to-seven-year cycle. He was concerned that capital needs for AI investment and government deficits could coincide with tighter Fed policy. He also warned that quantitative easing may not support markets as effectively as in the past, given accumulated public debt.
• Market concentration and downside risk: They described household wealth and economic activity as increasingly tied to equities, with a small number of companies accounting for much of S&P 500 earnings and growth. A major market decline, whatever its trigger, could hurt affluent consumers and contribute to a slowdown. Their principal macro downside concern was stagflation—weak or negative growth alongside persistent inflation.
• Hedge-fund performance and credit signals: A participant cited an estimate that the average or median hedge fund was up about 1% for the year. They also cited the $42 billion Paramount deal, saying its credit instruments had fallen about five points for some allocated investors over the preceding weeks.
• Midterms and seasonality: Moses cited historical figures suggesting U.S. stocks were positive in the 12 months after 20 of 20 midterm elections since 1980, and that there had been only two negative 12-month post-midterm periods dating back to the 1920s. He said he did not know the reason for the pattern and cautioned that it did not remove other risks. He also mentioned low hedge-fund net exposure and seasonally favorable conditions, which could argue against excessive pessimism. He anticipated possible market volatility around the elections and a potential government shutdown.
• Employment and politics: Daniel argued that AI-driven productivity gains could bring job cuts, making employment a more significant political issue in 2027–28. The transcript also refers to a Microsoft announcement of 60,000 layoffs, which the speakers discussed as a possible warning sign.
• Gold, oil, and the dollar: Moses said the group had bought gold below $1,000 during the GFC and argued it could be an alternative to betting against the S&P 500 amid sovereign-debt and currency concerns. He referenced gold trading near $4,000 and having approached $5,000, and said a hint of Fed rate cuts could lift it further. He noted that gold’s share of central-bank reserves had exceeded that of U.S. Treasuries, while the dollar’s reserve-currency share had fallen from the low 70% range to the mid-to-high 50% range. On oil, he cited roughly $65 as a level at which energy stocks had looked attractive in the past and around $100 as a level that could make them attractive now, while cautioning that high oil prices can eventually suppress demand.

Companies, Assets, and Instruments Mentioned

• AI and technology: CoreWeave, Anthropic, OpenAI, Meta, Alphabet, Amazon, Oracle, Nvidia, Broadcom, GPUs, data centers, and AI-related IPOs and financing. “Tropics” is also mentioned in the opening, though the name may be mistranscribed.
• Energy and industrials: Transocean, Valaris, Chevron, Exxon, oil, and drilling rigs. Daniel said Transocean’s planned Valaris acquisition was roughly one to two months from completion at the time discussed; he liked the consolidation and limited future rig supply.
• Housing and consumer: Home Depot, housing stocks, and consumer cyclicals.
• Financials and credit: MBIA, Wells Fargo, Countrywide, private equity, private credit, bank credit lines, securitizations, CDOs, mortgage-backed securities, and CDS.
• Other market instruments and assets: Gold, Bitcoin, 10-year bonds and Treasuries, the U.S. dollar, the S&P 500, oil, and sovereign debt.
• Historical technology comparisons: Lucent, Nortel, and Pets.com were invoked as dot-com-era examples.

Fed Comments and Key Figures

In the opening remarks, Moses said the Fed had left rates unchanged, with Beth Hammack and Lisa Cook dissenting. He quoted the Fed as waiting for “additional information to accumulate” and noted that the statement described the economy as “strong” rather than “robust.” The opening said the implied probability of a December hike had risen to over 80%.

Later in the discussion, the speakers cited a 75–80% probability of a December hike and discussed market pricing of three or four hikes by the following December. Their own views were more restrained: Moses expected at most a limited amount of tightening, while Daniel also said he expected fewer than three or four hikes. They noted that the Treasury Secretary could influence the outlook and discussed the possibility that longer-term yields had already tightened financial conditions.

Moses also cited a projected U.S. debt-limit figure of $41.6 trillion in the first six weeks of 2027, with debt then approaching $40.5 trillion. He said difficult earnings comparisons could emerge in 2027, and both speakers flagged 2027–28 as a period to watch.

Disclosures and Cautions

The speakers’ comments were presented as market views, not specific trade instructions. They emphasized that trading against crowded positioning can be risky, that AI-related businesses may require hard risk limits, and that their views could change as financing, earnings, and market conditions evolve.

The episode was sponsored by Clear Street. The program’s standard disclaimer states that opinions are for informational purposes, are not intended to influence specific trading or investment behavior, past performance does not guarantee future outcomes, and no particular results or profits are assured.

Presenters and Sources

Presenters: Danny Moses, Vincent (“Vinny”) Daniel, and Puneet Pardasani of Clear Street. Porter Collins was mentioned but did not participate in this call.

Sources and references mentioned: Mike Howell and CrossBorder Capital; The Wall Street Journal; Bloomberg; Kalshi; and Seawolf’s Substack and Discord research. Clear Street was the episode sponsor.