← Analysis page  ·  Todd Sohn hub  ·  Research hub

Actionable insights — The Market's Biggest Warning Signs Right Now

The repeatable analysis behind the read: not what he'd buy, but how he reads the tape — chart tells and ETF-flow signals written so the process can be rerun later on different names.
2026-JUN-29 · The Real Eisman Playbook (Ep 66) · guest Todd Sohn (Strategas) · ▶ Watch · full analysis · transcript
How to read this page: each insight is a method — the tell that puts him onto a view, the steps to turn it into a stance, and the signal to watch when re-running it. The boxed line shows how it played out in this episode. Timestamps deep-link into the video.

9:38 1. "Print 100 charts" — read the market's message before you read any story

The repeatable method
  1. Periodically print (or pull up) ~100 charts spanning the big-cap leaders, sectors, factors and assets — the breadth, not a watchlist of favorites.
  2. Flip through them fast and let the aggregate picture surface: which groups are trending, which are messy, where the leadership and the divergences are.
  3. Treat the exercise as a falsification tool, not confirmation — charts "tell you when you're wrong or right," so look for the names that contradict your current view.
Here: the 100-chart sweep yields the whole thesis in minutes — semis & power leading (SOXX, GEV), software/Meta/Microsoft messy, one theme dominating. Practitioners cited: Chris Verrone and Strategas's Jason Trennert.
Watch for

7:09 2. The 200-day slope is the trend tell — not just price-above-or-below

The repeatable method
  1. For any name, look at the direction of the 200-day moving average, not merely whether price sits above or below it. A rising slope = healthy trend; flattening = trend in question; rolling over (down) = trend changing for the worse.
  2. Combine the slope read with a new-high/new-low check (insight 4) for a two-factor confirmation.
  3. Downgrade a name when the slope flattens-to-down even if price is still elevated — the change of slope leads the change of price.
Here: META "looks more like a short" — 200-day flattening to downward; ORCL's 200-day "rolling over, downward to flat"; MSFT already retesting its spring lows. Contrast GEV / SOXX, where the slope is still up (overbought, but trending).
Watch for

1:39 3. RSI divergences — a new high that's less overbought is a warning

The repeatable method
  1. Track momentum with RSI (relative strength index) — how fast price is rising/falling and how overbought/oversold it is.
  2. Look for divergence against price: a fresh price high reached at a lower RSI than the prior high = weakening momentum = warning; a fresh price low at a higher (less oversold) RSI = potential change of trend / bottoming.
  3. Use it to time trims and flag exhaustion, not as a standalone buy/sell.
Here: the leaders (SOXX, GEV, GOOGL) are "super overbought" — trim-and-revisit candidates rather than fresh buys; the less-oversold tells underpin his "maybe bottoming" reads on GLD and IBIT.
Watch for

7:09 4. The new-high confirmation check — a leader that stops confirming the market

The repeatable method
  1. When the broad market (S&P, equal-weight S&P, small caps) makes new highs, list which leaders are not joining — names that haven't made a new high in months.
  2. Treat that non-confirmation as relative weakness: the market is fine, but the laggard is being left behind and is the candidate to fade/short.
  3. Invert it for the broad-tape read: when most groups confirm, the trend is healthy; when leaders peel off one by one, breadth is deteriorating.
Here: META hasn't made a new high in months while the market does → "looks more like a short." MSFT retesting spring lows is the same tell, more advanced.
Watch for

21:01 5. ETF flows as an investor-behavior barometer — the one-game test

The repeatable method
  1. Read cumulative sector-ETF flows from a clear anchor date (e.g. a market low) as a direct map of where investors are actually putting money — flows are behavior, not opinion.
  2. Compute the concentration: one sector's inflow vs the sum of all others. A single sector taking essentially all the money ("one game") is both the engine and the single point of failure.
  3. Pair extreme concentration with a top-check (insight 8): concentration alone isn't a sell, but it tells you where an unwind would do the most damage.
Here: since the March-30 low, ~$27B into tech ETFs vs −$4.4B for every other sector combined — "one game." Within tech, strip out software (a problem) → it's semis + hardware. "If there's an unwind in tech, it's going to be ugly."
Watch for

19:08 6. "Look under the hood" — index, quality and factor ETFs are disguised tech bets

The repeatable method
  1. For any "diversified" or factor fund you own, pull its actual top holdings and sector weights — most are market-cap weighted, so the biggest stocks dominate.
  2. Check correlation/R-squared to the S&P: if a "quality," "momentum" or growth ETF moves nearly one-for-one with the index, it isn't diversifying — it's the same bet at a higher fee.
  3. Audit the whole sleeve for hidden double-counting: owning the S&P + a growth fund + a quality fund + a thematic fund can be four ways of owning the same mega-cap tech names.
Here: tech ~40% of the S&P (>50% with Google + Amazon); a ~$50B "quality" ETF has near-perfect R-squared to the S&P because the index "has morphed into a quality index" — "quality is wasting space in your portfolio." Owning an index is "not diversified."
Watch for

27:37 7. The thematic / leveraged-ETF caution — drawdown, Sharpe and daily rebalancing

The repeatable method
  1. Before buying a thematic ETF, look at the category's historical drawdown and risk-adjusted return, not its recent performance — the typical thematic fund is high-vol with a poor Sharpe ratio (return per unit of risk).
  2. Read a surge of new thematic/leveraged launches as a herding/performance-chasing tell — froth, not opportunity.
  3. Understand the mechanics of leveraged ETFs: they reset exposure daily, so they buy more into up days — adding volatility — and decay point-to-point from slippage + fees. They are trading vehicles, never buy-and-hold.
Here: thematic ETFs average a ~32% three-year drawdown and 70% have a Sharpe below 1 (cannabis ~−80%). Leveraged + levered-single-stock ETFs are a ~$200B record-usage pile (e.g. a "2x SpaceX" launching) — "kind of like a lotto ticket… things are getting frothy."
Watch for

30:23 8. The >15%-sector-weight road map — history bounds how far a theme can run

The repeatable method
  1. When one group's S&P weight gets extreme, compare it to the historical club of groups that have crossed ~15% (a long, rare list).
  2. Use prior peaks to bound the upside (how far it ran) rather than to call the exact top — "a road map, not a prediction."
  3. Study how those episodes resolved for the group and the market; let the base rate temper position sizing as the weight climbs.
Here: the >15%-since-1990 club is tech hardware (peaked ~30% in the bubble), energy (mid-2000s) and software (late last decade) — two of the three ended badly. Tech is ~40% now; staples, by contrast, fell from ~19% (early '90s) to 4.5%.
Watch for

41:39 9. Pay attention when an asset doesn't behave as it should — plus the outflow bottoming-watch

The repeatable method
  1. Form the textbook expectation first (gold "should" rally on war/inflation; a defensive should hold in a scare), then watch whether the asset actually does it.
  2. Flag a failure-to-behave as a regime tell — "pay attention when something supposed to behave one way doesn't," even if you can't yet explain why.
  3. Cross it with fund flows: heavy ETF outflows from a beaten, out-of-favor asset are the contrarian setup that can mark a bottom ("the bar's low").
Here: GLD didn't rally on war/inflation and is bleeding ETF money — a "metal mania" blow-off that "maybe starts to bottom out." IBIT is "rough," same setup — the degens left for Kalshi — with money leaving Bitcoin ETFs.
Watch for

Methods distilled from the public YouTube video (transcript in transcript.txt) for personal study. Not investment advice. © The Real Eisman Playbook / Steve Eisman for source material.