Skip to main content

Market Starting To Crack Under High Bond Yields? | Michael Lebowitz

Adam Taggart | Thoughtful Money® Sep 26, 2026 1:04:15

Summary

Michael Lebowitz argues a calm S&P 500 hides real strain, with rate-sensitive sectors and the equal-weight index already well behind.

Companies that issued bonds in 2021 at low rates now refinance near 6%, forcing layoffs, store closures and cut capital budgets.

Lebowitz calls an AI capital spending slowdown the biggest market risk, since data centers supply about 4 points of GDP growth.

Measuring real yields against breakeven inflation rates (what investors expect prices to rise) shows policy is more restrictive than the headline yield suggests.

Lebowitz's team trimmed tech for value and moved from the 30-year into the 5-year, though oil near $120 could undo the rotation.

Adam Taggart's takeaways

  • Top takeaway His top point is the return and timing on AI investment: he frames the current earnings boom as largely balance-sheet money that hyperscalers are converting into compute, chips and data centers instead of cash flow, and asks when revenue actually shows up elsewhere in corporate America.
  • On whether the Fed's tightening or the administration's growth push wins, he argues rates eventually prevail because very few companies are earning meaningful AI revenue yet, so non-AI companies absorb the pain through layoffs.
  • He faults the Fed's projections as incoherent, saying rates were marked higher while every economic forecast was left alone, which amounts to assuming higher rates do no harm.
  • He is skeptical that AI answers inside search can be monetized, citing his own experience getting the same answers from free models, and expects users to ration usage toward cheaper, smaller models to cut token costs.
  • He closes by urging viewers to consider raising their credit allocation now and to plan against a goal rather than trying to beat the market year after year, pushing back on the idea that more equity risk is needed to hit a target.

Michael Lebowitz's takeaways

  • Top takeaway His top point is that calm index levels are misleading: a very small group of AI tech stocks is masking significant churn underneath, with the equal-weighted S&P down 4 percent while rate-sensitive sectors are worse (financials -5, short small cap -6.5, real estate -7, utilities -9), and his dispersion index sitting at the extreme readings that previously preceded index declines.
  • He calls an intentional hyperscaler capex pullback the biggest market risk over the next year or two, on the idea that they decide they are spending too much and build five data centers instead of seven, which he says could easily send the market into a 25 percent correction.
  • On rate transmission he stresses delay rather than absence of impact: bonds issued in 2021 come due now at rates repriced from about 2.5 percent to 6 percent, forcing cost cuts, and by his breakeven math real yields near 2.7 percent are very restrictive. He adds that the pain is concentrated in the belly of the curve, with two-year yields up nearly half a percent over the last month while the 30-year rose only 6 basis points, so it lands first on floating-rate and shorter-duration borrowers.
  • He criticizes the Fed's latest projections as internally odd, saying the 2027 year-end funds rate median was raised about half a point from 3.6 to 4.1 percent with no change to the growth, inflation or unemployment forecasts, which implies higher rates do no harm.
  • On positioning he says the firm sold some longer bonds to buy the five-year at about 5 percent, trimmed the large-cap tech position, added value and a little utilities, and he argues expected ten-year equity returns at today's valuations look negligible next to a return locked in at roughly 5 percent.

Darius Dale's takeaways

  • Top takeaway Taggart recaps that Dale read the Fed's move as signaling to the bond market, with Warsh potentially hiking once or twice to establish his hawkish bona fides and calm the long end.
  • He also relays Dale's call for one or two hikes followed by rate cuts, a path Taggart said he broadly agreed with at the time.

Kevin Warsh's takeaways

  • Top takeaway Lebowitz's main criticism is aimed at the Fed projections under Warsh: the 2027 year-end funds rate median was raised about half a point to 4.1 percent from 3.6 percent while growth, inflation and unemployment forecasts stayed unchanged, which he calls a strange forecast given that higher rates weigh on the economy.
  • Taggart says Warsh dropping forward guidance is a big deal, and that the resulting uncertainty about the Fed's path is one of the main reasons rates are rising across the curve.
  • A thesis Taggart relays from earlier episodes is that Warsh may hike once or twice to establish his hawkish credentials and settle the bond market, then be positioned to cut in 2027, with Lebowitz questioning how much those eventual cuts will do economically.
  • Lebowitz says if bond yields keep rising, Warsh is forced to chase them with hikes, whereas settling yields would give him more room to cut, which is why he watches the long end so closely.
  • He agrees with Warsh that core measures are the right lens, putting core CPI around 2.4 percent and core PCE around 3.4 percent as of the recording, still above the 2 percent target.

Elon Musk's takeaways

  • Top takeaway Lebowitz names Musk's xAI, which Taggart cannot recall the name of on air, alongside Anthropic and OpenAI as one of the model labs most at risk if their models are commoditized against much cheaper free ones, given what was spent building them and their debt loads.