The Bulls Are Trapped | TCAF 266

Episode Overview

• Alex Altmann discusses with Josh Brown and Michael Batnick how tactical strategists combine quantitative positioning data with qualitative sentiment to navigate short-term market movements, retail dominance, multi-manager crowding, and the sustainability of the AI trade.
• The conversation explores the structural forces fueling the equity market, specifically how retail investors, automated 401(k) inflows, corporate buybacks, and priced-out homebuyers are creating a relentless bid for stocks.
• The hosts and guest examine how the tripling of multi-manager hedge fund assets alongside a shrinking number of public companies has distorted stock correlations and amplified single-stock volatility.
• The discussion concludes with an analysis of the massive capital expenditure and debt issuance required to fund artificial intelligence, highlighting why widening credit spreads and a looming revenue gap point toward a potential tactical market pullback.

Key Concepts

• Speaker Context, Industry Ties, and Pre-Show Observations: The episode is presented by Grayscale, a crypto-focused investment platform. Josh Brown and Michael Batnick are co-hosts at Ritholtz, and their guest is Alex Altmann (known as Alti, a nickname he adopted at work in 2007 because his London high school class had 5 Alexes, 2 Alastairs, 1 Alec, and 1 Ali). Altmann is Managing Director and Global Head of Equities Tactical Strategies (BETS) at Barclays, having started his career as a single-stock analyst in 2004 at JPMorgan, pivoted to macro in 2012, served as head of equity trading strategy at Citi for a decade, and worked as a macro portfolio manager at Millennium. Altmann's BETS team comprises 11 people globally (3 in New York, 3 in London, and 5 in Asia) who generate tactical trading ideas and research for Barclays' institutional clients. During the show, Altmann makes a shameless plug for his team's research and references proprietary Barclays models like the Barclays Equity Timing Indicator (BETI). In pre-show banter, Brown notes that office space on a decently high floor in Hudson Yards costs $250 a square foot and Manhattan requires signing 10-year leases, while Altmann mentions the park on the West Side of Manhattan has been built all the way south to 59th Street on the water and claims it is very plausible the U.S. sees a recession in the next 18 months before the next election.
• Tactical Time Horizons and Baseline Market Odds: While traditional research outfits focus on 6- to 12-month horizons, Barclays' tactical strategy focuses on 1, 2, 4, or 8 weeks ahead (typically 4 to 8 weeks out). Understanding tactical setups requires knowing baseline market odds: if an investor buys the S&P 500 on any random day and holds it for 42 trading days (about 2 months), the historical odds of making money are about 73%, with an average return just shy of 200 basis points (190 basis points). To beat this baseline, the BETS team looks for setups with 80% to 85%, and in some cases 90%, hit rates based on backtesting across long in-sample and out-of-sample periods using their BETI framework. Altmann hedges that while BETI is pretty good, it is not perfect, noting that if it were perfect, he would not be sitting on a podcast or talking to customers about it, but would run it internally from a beach. Furthermore, backtesting positioning data reveals that heavy crowding in an asset is not always a contrarian short signal; sometimes crowding validates that forward returns are statistically very strong.
• Retail Dominance and the Structural Bid for Equities: Retail investors are the largest owners of stocks, followed by mutual funds, and the vast majority of retail portfolios have outperformed institutional portfolios. Altmann explains that retail investors have superior holding power because they do not face tight institutional drawdown limits (such as risk committees forcing sales after a 400-basis-point drawdown); the only things that cause retail to stop out and sell are margin calls or losing their jobs. Barclays tracks retail positioning data going back to 1995, and Altmann identifies October 2019 as the true demarcation for retail trading due to the introduction of zero-commission trading and gamified apps like Robinhood. Brown adds that 40 million new brokerage accounts were opened in 18 months during the COVID era, and both hosts note that retail trading still accounts for roughly 25% of overall market volume, having survived 2022 when Nvidia and Meta both fell 75%. Today, approximately two-thirds of all American homes own stocks, which is higher than during the dot-com bubble. This ownership is supported by a massive structural bid: 401(k) contributions bring an estimated $800 billion to $900 billion a year into equities (calculated by taking the number of U.S. homes, applying a haircut for those without 401(k)s, assuming a 12% paycheck contribution, and using a median household income of $88,000), while corporate stock buybacks add $1 trillion a year in authorizations, or about $750 billion after haircut adjustments since actual executions are typically 60% to 70% of announced authorizations. Additionally, because housing is unaffordable, Gen Y and Gen Z savings that would typically go into new mortgage originations and down payments, estimated at $200 billion to $300 billion a year, are flowing into fractional shares of stocks instead (Altmann guesses the median U.S. home price is between $300,000 and $400,000, requiring a $50,000 to $60,000 deposit, hedging that he does not know the exact number off the top of his head). Combined, these channels create well over $1 trillion in consistent annual equity inflows.
• The Financialized Economy and the Real-Time Wealth Effect: Currently, 35% of U.S. household net worth is in equities, the highest level ever, while the second-highest wealth channel is home equity. The percentage gap between stocks and home equity in household wealth is the widest in history going back to the 1960s, especially as Brown points out that home prices have appreciated 0% over the last 3 years due to interest rates while the stock market has doubled. Because the economy is so heavily financialized around the stock market, equity drawdowns immediately impact the real economy. For example, when the S&P 500 fell 20% last year during the tariff scare, there was a visible drop in consumer spending over the subsequent two months due to a real-time negative wealth effect. Policymakers are acutely aware of this dynamic: Brown notes that in December 2018, the stock market entered a 21% statistical bear market on Christmas Eve, prompting the Federal Reserve to reverse its hawkish stance and cut rates a month later. While there was only 1 Bloomberg terminal at the Federal Reserve in Washington ten years before the Great Financial Crisis (GFC), Altmann confirms that both the Treasury Department and the New York Fed have Bloomberg terminals today because they message him on them.
• Multi-Manager Pod Shops, De-Equitization, and Stock Dispersion: Traditional prime brokerage data showing hedge fund net exposure is largely uninformative because net exposure fluctuates in a tight range from 48% to 55% and back to 48%, driven by relative value (RV) pod shops running market-neutral strategies. Multi-manager (pod shop) assets under management (AUM) tripled from around $300 billion in 2018 and 2019 (pre-COVID) to about $1 trillion today. In a high interest rate environment, market-neutral RV strategies are highly attractive because a manager puts up 15% collateral per side (30% total) for prime brokerage requirements and earns 5% or more in interest on the remaining 70% cash while isolating idiosyncratic alpha. However, this tripling of AUM has collided with de-equitization: Brown notes there are roughly 3,000 tradable stocks today, down from about 10,000 when he started in the business. Because more money is chasing fewer stocks while neutralizing sector and factor exposures, crowding is pushed into other portfolio areas, causing low index volatility alongside massive single-stock and factor volatility. On earnings days, the majority of single-stock flow comes from discretionary investors, whereas on non-earnings days, the vast majority comes from quant and retail. This dynamic is visible in Société Générale (SocGen) Chart 15 (260-day correlation, average since 1995 vs. Latest), which shows that out of 124,750 possible pair-wise stock combinations within the S&P 500, 40% are currently negatively correlated. On the day of recording, this dispersion is evident as the equal-weight S&P 500 is up 98 basis points while the cap-weighted S&P 500 is down 53 basis points, ServiceNow is down 27%, and in after-hours trading, Intel is down 25%, Amazon is down 12%, semiconductors are down 4%, and SanDisk is down 9% before flipping to green later in the show.
• AI CapEx, Cross-Asset Issuance, and Widening Credit Spreads: The S&P 500 has become heavily reliant on artificial intelligence: for next year (2027), 75% of consensus S&P 500 earnings growth is expected to come from just semiconductors and hyperscalers. Batnick notes that Meta is spending 70% of its top-line revenue on CapEx, and cites Chart 8 from The Compound Media showing Microsoft Cloud quarterly revenue grew 30% year-over-year to $63.9 billion in Fiscal Q1 2027 (accelerating to a growth rate not seen since Q3 2022 when revenue was $23.4 billion at 32% growth). Yet on the day of recording, Microsoft announced a $25 billion debt offering and the stock traded down despite strong cloud growth. Altmann explains that funding this AI buildout is shifting capital markets: 2026 is projected to be a positive net equity issuance year for the first time in years, a multi-hundred-billion-dollar swing from historical net negative issuance of -$400 billion to -$600 billion a year (Barclays Chart 16 shows gross equity issuance exploding in 2026 while buybacks remain flat, though Altmann hedges that this relies on a guess that buybacks remain flat year-over-year, which might be too generous given massive CapEx commitments). Simultaneously, net Investment Grade (IG) and High Yield (HY) corporate debt issuance in 2026 is projected to reach $1.2 trillion. Barclays Chart 12 (tracking data back to 1995 with an $R^2$ of 0.4712) plots combined cross-asset net issuance (equities, corporate debt, and government debt as a percentage of market size) against 2-year Treasury yields, showing 2026E as an extreme outlier with roughly 13% issuance colliding with a roughly 4.8% 2-year yield. This supply is straining credit markets: Barclays Chart 11 shows the spread on an equal-weighted basket of long-dated IG data center bonds (average duration around 17 years) against the 20-year U.S. Treasury has widened from roughly 100 to 140 basis points out to 222.3 basis points. Despite this, a Barclays survey of 200 Wall Street investors released on the day of recording found that only 16% of respondents thought there was an AI financing problem.
• Balance Sheet Quality, Credit Warnings, and Index Concentration: Credit stress is increasingly dictating equity performance. Barclays Chart 13 demonstrates that a long/short equity basket of strong versus weak balance sheet stocks is closely tracking the widening of CCC versus BB credit spreads. Furthermore, Batnick cites Chart 14 from Blake Millard's Daily Chartbook, showing the spread between B and CCC credit has reached nearly 8%, an event seen only 4 other times since 1996 (October 19, 1998; November 10, 2000; January 18, 2008; and March 23, 2015). On every prior occasion, the S&P 500 was already at least 11.6% (or 12%) below its 52-week high, and twice it was down around 30%, whereas today the index sits just 2% below its highs. In discussing potential shakeouts among neoclouds like CoreWeave, which Batnick notes has a $33 billion market cap and Altmann notes is still trading above its IPO price, Altmann hedges on whether Nvidia CEO Jensen Huang would step in to buy distressed neocloud assets, stating he does not know what Jensen is thinking and does not want to speak for him, though it would be logical to hoover up assets in a shakeout. Altmann also explicitly hedges when comparing hyperscalers and neoclouds to G-SIBs and subprime during the 2008 financial crisis, stating it is not the right analogy and he does not want to draw too strong a parallel to the GFC or imply hyperscalers are going to zero. He also admits that despite his team's efforts, there is currently no way to quantitatively map the AI factor because betas and correlations are shifting too rapidly. Meanwhile, index concentration creates structural constraints: Nvidia trades at 15 times forward earnings and represents 8% to 9% of the S&P 500, creating holding limits for mutual funds capped at 5% per position; if Nvidia traded at 40 times forward earnings, it would be a $13 trillion company and the S&P 500 would sit at 12,000.
• The AI Revenue Gap and Tactical Market Outlook: To justify current AI CapEx commitments of $1.2 trillion next year and $1.4 trillion the year after, Altmann estimates the AI ecosystem must generate $1 trillion in revenues by the end of 2028, yet the current run rate for that timeframe is only around $400 billion. To put $1 trillion of new revenue into perspective, total S&P 500 top-line revenue is $17 trillion (with about $5.5 trillion in profits), and the cumulative inflation-adjusted revenue of Apple and Microsoft combined from 1990 to today is $9.8 trillion (while SpaceX claims a $30 trillion total addressable market). Because of these dynamics, Barclays' BETI model has been signaling various degrees of risk-off since the end of May, following a bullish stance during the Iran conflict. Altmann notes that in the summer of 2024, an AI return-on-investment debate combined with a Bank of Japan (BOJ) move and a weak Non-Farm Payrolls (NFP) report caused a 12% to 14% dent in the S&P 500, whereas the largest pullback so far in 2026 was 8.5% around the Iran situation. Consequently, Altmann's tactical target for a healthy market pullback on the S&P 500 is 7,100 to 7,200, which aligns with the 200-day moving average and represents about 17.5 times forward next-twelve-months (NTM) price-to-earnings.

Quotes

• At 0:07:32 - "Positioning can be instructive because obviously flow tends to dictate price, right? And price dictates narrative." - Alex Altmann explaining why tactical strategists prioritize tracking capital flows and positioning over market stories.
• At 0:10:18 - "So if you were to buy the S&P on any given random day, and you were to hold that position for 42 trading days, that's equate to about two months, the odds of you making money, right, is about 73%, and your average return would be around about, just shy of 200 basis points, 190 basis points." - Alex Altmann establishing the baseline historical probability of positive equity returns over a two-month tactical window.
• At 0:17:30 - "Only 16% of people that responded thought that we had a financing problem. I'm like, how can that be? We have a financing problem because we've just gone through one over the past two weeks, so that number should have been 100%." - Alex Altmann reacting to a Barclays survey of 200 Wall Street investors that ignored the recent widening of data center credit spreads.
• At 0:20:17 - "People always ask me, oh, what gets retail to sell? I'm like, lose your job. Or when your neighbor loses their job." - Alex Altmann explaining why retail investors have stronger holding power during market drawdowns than institutional managers.
• At 0:23:27 - "35% of US household net worth is now in equities, the highest ever. The next highest wealth channel is, unsurprisingly, home equity. But the gap between stocks and home equity in terms of your wealth percentage for households is the widest it's ever been in history going back to the 1960s." - Alex Altmann highlighting the unprecedented financialization of U.S. household balance sheets.
• At 0:28:37 - "What keeps you awake at night? Is it like levered ETFs, or is it Iran, or is it AI? I'm like, no, it's retail. It's retail is the scary one in the sense that because we've got such a financialized economy, right?" - Alex Altmann discussing how a sustained drop in equities could trigger a rapid negative wealth effect in consumer spending.
• At 0:35:09 - "If you think about multi-manager AUM in 2018, 2019, before COVID, was around about $300 billion, and today it's like a trillion, right? So it's basically tripled since COVID." - Alex Altmann detailing the rapid growth of market-neutral pod shops.
• At 0:37:42 - "No, all you've done is push the crowding into other aspects of your portfolio that you are either aware of or unaware of, right? And this really matters because when you've ballooned the size of the number of strategies, so to speak, you're creating more of these vicious, or the propensity for these vicious unwinds." - Alex Altmann explaining the unintended consequences of multi-manager funds neutralizing sector and factor risks in a shrinking stock market.
• At 0:43:30 - "Next year, 2027, consensus S&P earnings, 75% of that is expected to come from semiconductors and hyperscalers." - Alex Altmann illustrating the extreme earnings concentration driving the S&P 500.
• At 0:43:55 - "Meta is spending 70% of their top line on CapEx. Of their top line, of their fucking revenue, just on CapEx. It's unbelievable." - Michael Batnick emphasizing the scale of capital expenditure commitments among major technology companies.
• At 1:02:06 - "We need a trillion dollars of revenues, basically, by the end of 2028 to pay for all of this CapEx. And right now, our run rate for then is going to be somewhere around about 400, right?" - Alex Altmann quantifying the $600 billion revenue shortfall facing the artificial intelligence ecosystem.

Takeaways

• Benchmark tactical trading strategies against the S&P 500's historical baseline odds of a 73% win rate and a 190-basis-point average return over any random 42-trading-day period, deploying capital only when backtested hit rates reach 80% to 90%.
• Stop relying on prime brokerage net exposure reports as a directional sentiment gauge, as market-neutral multi-manager pod shops keep net exposure locked in a tight 48% to 55% band regardless of macro conditions.
• Retail investors are not bound by institutional drawdown limits and only liquidate when facing job losses or margin calls.
• Track cross-asset net issuance and long-dated data center credit spreads against U.S. Treasuries as early warning indicators for the AI trade, recognizing that widening credit spreads signal capital strain even when equity investors remain complacent.
• Use high-yield credit spreads, specifically the spread between B and CCC rated debt and BB versus CCC debt, to guide equity factor selection toward strong balance sheets and away from weak balance sheets during periods of heavy corporate debt issuance.
• Account for hidden factor crowding and violent single-stock unwinds on non-earnings days, when quant and retail flows dominate trading across a shrinking pool of roughly 3,000 public equities.
• Prepare for a potential tactical pullback in the S&P 500 toward the 7,100 to 7,200 range (aligning with the 200-day moving average and a 17.5x forward P/E multiple) as the market digests the gap between massive AI CapEx commitments and actual revenue run rates.