← Analysis page  ·  Liz Ann Sonders hub  ·  Research hub

Actionable insights — Beyond The Mag 7

The repeatable analysis behind the calls: not what she rates, but how she measures it — written so the tests can be rerun later on different data.
2026-SEP-01 · The Master Investor Podcast with Wilfred Frost · Liz Ann Sonders (Charles Schwab) · ▶ Watch · full analysis · transcript
How to read this page: each insight is a method — the test she actually runs, the numbers that make it decisive, and the signal to watch when re-running it. The boxed line shows how it played out in this appearance. Timestamps deep-link into the video. Sonders is a strategist, not a stock picker, so most of these are measurement disciplines: they tell you what a market statistic really means before you act on it.

5:44 1. The orderly-vs-disorderly yield test — three gauges, not one level

The repeatable method
  1. Stop asking "what yield breaks stocks." A level matters for psychology and how investors react, not as an economic turning point — so don't build a thesis on one number.
  2. First, sanity-check the level against fundamentals: track the 10-year against nominal GDP growth and against the level of inflation. If it sits at or below both, the move is normalization, not distress — and may have further to run.
  3. Then mark the round-number psychological thresholds investors actually watch (4.75%, then 5%) and treat a breach as a sentiment event, not a valuation event.
  4. Add the volatility gauge: the MOVE index is the bond market's VIX. A yield rise with a calm MOVE is orderly; a breach of the level plus a MOVE pickup is the combination that spills into equity volatility.
  5. Add speed as an independent variable — the same 20bp is benign over a month and disorderly over two days.
  6. Confirm the transmission is real by looking beneath the index: the interest-sensitive sectors should be the ones bleeding.
Here: 10-year at 4.80% (30-year ~5.3%), through the 4.75% marker with the MOVE "relatively calm" → still orderly; the damage shows only sub-index in XLU / XLRE, with the money in XLE and XLF (4:16).
Watch for

8:07 2. Diagnose the correlation regime before you design the portfolio

The repeatable method
  1. Measure the rolling correlation between bond yields and stock prices — this single relationship determines whether a stock/bond mix diversifies at all.
  2. Ask what yields are keying off. If yields move on growth expectations, yields and stock prices move together (rising yields = improving economy = good) — the great-moderation regime, late 1990s → the 2022 inflation spike, characterized by low inflation volatility, falling rates, and globalization (China into the WTO, 2001).
  3. If yields move on inflation, yields and stock prices move in opposite directions — meaning bond prices and stock prices move together, and the hedge fails exactly when you need it. That's the "temperamental era," mid-to-late 1960s → late 1990s: shorter cycles, more frequent recessions, higher inflation volatility.
  4. Classify the present regime, then set the diversification budget accordingly: 60/40 was a great-moderation artifact. In a temperamental regime you must source diversification from outside the two-asset mix.
  5. Treat this as the primary secular question and the yield level as the secondary tactical one — "not so much just how speedy the move is in the 10-year… it's that relationship."
Here: "we're now back in pretty deep negative correlation territory between bond yields and stock prices" → the great moderation is over; the offset is that access to non-correlated asset classes has been democratized, so individual investors are better equipped than 1970s investors were (10:55).
Watch for

11:41 3. Symptom vs disease — score any official intervention before you trade it

The repeatable method
  1. When an authority acts on a price (buybacks, issuance shifts, yield caps), first ask whether it addresses the cause or the symptom. Naming the disease is the whole exercise: here, fiscal profligacy, runaway deficits and debt, and investors demanding more compensation to finance them.
  2. Check whether the action is genuinely new or a change of degree — Treasury was already buying back the long end; Bessent doubled it. A change of degree rarely changes a trend.
  3. Check for cross-institution conflict: is another arm of policy pulling the other way? Treasury suppressing long yields cuts directly against a Fed that wants to shrink its balance sheet and let the long end do part of its tightening.
  4. Check who else is competing for the same buyer. A wave of AI-related investment-grade corporate issuance is "a shiny new object" pulling the marginal IG buyer away from Treasuries — supply the intervention can't offset.
  5. Finally, classify the inflation you're fighting: supply-side (energy, tariffs) is largely outside the Fed's reach; monetary policy only moves the needle properly on demand-side inflation.
Here: doubled long-end buybacks scored as symptom-treatment, in conflict with Warsh's Fed and swamped by AI corporate issuance → don't expect it to cap the long end; the term premium keeps doing its work.
Watch for

20:56 4. Escalator or elevator — price the speed of a Fed cycle, not its direction

The repeatable method
  1. Don't stop at "will they hike." Take the historical set of hiking cycles and measure equity performance over the subsequent one year from the first hike.
  2. Baseline: about +4.5% on average — positive, but below a normal annual return, so a hiking cycle alone is not a sell signal.
  3. Split the sample by speed. Fast cycles: about −4% over the following year. Slow cycles: more than +10%. The dispersion around the average is the entire signal.
  4. So the forecast to actually make is pace, not direction: "are they going to take the escalator or are they going to take the elevator."
  5. Sanity-check the institutional constraint separately: policy is set by a committee of seven-plus voters with a "cacophony" of public speakers — "the FOMC is a committee. It's not a chair" — so political pressure on the chair is a weak input compared with the speakers' consensus.
Here: best guess a 25bp September hike (what the market prices), with the burden of proof on inflation rather than labor → the equity call hangs on whether it is a slow, one-off-style tightening or the start of a fast run (19:58).
Watch for

28:30 5. Better-or-worse over good-or-bad — trade the second derivative, and use the whisper number

The repeatable method
  1. Rule of thumb from 40 years on the job: "better or worse often matters more than good or bad." A growth rate falling from 60% to 30% is still phenomenal and still a sell — the inflection point, the rate of change, the direction of travel is what stocks price.
  2. Identify the real bar. After a boom the published sell-side consensus is stale; the live bar is the buy-side whisper number. A print between the two is a beat on paper and a miss in practice.
  3. Expect the first cracks at the individual stock level — rising dispersion — before the aggregate group breaks. Don't wait for the index to confirm.
  4. Check index weight before sizing the consequence: if the disappointing name is a dominant weight, the stock event becomes a market event.
Here: Samsung beat sell-side consensus on top and bottom line, undershot the whisper number, and was routed — dragging the KOSPI ~40% down because Samsung and SK Hynix are such large weights. She calls it "the early poster child" of the dispersion phase to come (not necessarily as soon as Q3 reporting).
Watch for

29:57 6. Rank by contribution to return, not price performance

The repeatable method
  1. Compute each name's contribution to index return = price performance × cap size. Price performance alone tells you nothing about who actually moved the index.
  2. Rank the whole index on it and read the extremes — including where the members of your favorite thematic basket land.
  3. Use the spread within the basket as your dispersion gauge. If a nine-stock theme spans nearly the full 504-name ranking, the theme is not a single trade any more.
  4. Maintain the basket definition deliberately rather than inheriting a media label — she extended the "Mag 7" with Micron and Broadcom to build the Neural 9, and posts it daily.
  5. Cross-check where the money is going next: broad indices outperforming the mega-cap index says the theme has spread out.
Here: MU is the 3rd-best contributor to S&P returns this year while TSLA is the 503rd — "almost the best to the worst in an array of nine stocks"; meanwhile IWM has double the S&P's YTD return and has beaten it over two years.
Watch for

32:16 7. Measure concentration in earnings growth, not market cap

The repeatable method
  1. Take next calendar year's expected index EPS growth vs the prior year (part actuals, part consensus) and decompose it by company.
  2. Compute each name's share of total index earnings growth, then the cumulative share of the top 2 and the top 10.
  3. Read the fragility directly off those shares: a high-profile miss then hits twice — market psychology, and a mechanical ratcheting-down of index estimates.
  4. Look for the offset: how many of the 11 sectors show an improving earnings profile? Breadth of improvement is genuine good news — but test whether it is "meaty enough when you do the math for a cap weighted index." Usually it isn't.
  5. Keep the two concentrations separate in your head. Market-cap concentration can be easing while earnings-growth concentration worsens — the second is the dangerous one.
Here: NVDA 18% + MU 14% = 32% of all expected 2026 S&P earnings growth from two companies; the top 10 — which brings in CVX and XOM around #9/#10 — is two-thirds. Offset: 10 of 11 sectors improved in Q2, but not enough to matter in a cap-weighted index.
Watch for

39:18 8. Average-member drawdown — measure the correction the index is hiding

The repeatable method
  1. Take the index's own maximum drawdown for the period. Then compute the maximum drawdown of every single member and average those.
  2. The gap between the two is the amount of correction that has been absorbed by rotation rather than by the index falling.
  3. Run it on more than one index — the wider the gap, the more the "market" is really a sequence of single-stock bear markets.
  4. Use it to reframe the risk question: a large average-member drawdown means excesses (valuation, sentiment, an over-set earnings bar, a macro narrative change) are already being eased piecemeal, which is preferable to easing them all at once.
  5. Then ask which excess each rotation is discharging, and where the money is going — that names the next favored sector.
Here: S&P index max drawdown 9% and change (never even a 10% correction), but the average of all 504 members' individual drawdowns is −25.5%; Nasdaq −13% at index level vs −45% average member YTD. "We would all choose to have this experience versus the S&P in the aggregate dropping by 25% all at once."
Watch for

41:36 9. Factor overlay on sector allocation — sectors are too monolithic to act alone

The repeatable method
  1. Rate sectors on a deliberately soft favorable-to-unfavorable scale rather than overweight/underweight — the extra precision of the traditional labels is false in a rotating market.
  2. Then layer factors — "another word for characteristics" — over the sector call rather than instead of it. Three families: growth (rising forward estimates, margin stability or strength, positive earnings surprises), value (P/E, price/book, price/sales), balance sheet (strong free cash flow, high interest coverage).
  3. Justify the overlay empirically: factors have shown "more consistency in outperformance and underperformance" than sectors, which behave monolithically.
  4. Apply it within each favored sector, because dispersion inside sectors is now wide — the factor screen is what separates the better end from the worse end.
  5. Distinguish the reason for each sector rating so you know what would invalidate it — a cyclical rating breaks on the cycle; a valuation rating breaks on price.
Here: favorable on a cyclical basis — XLI, XLB, XLF (financials as steep-curve beneficiaries); favorable on valuation — XLV; less favorable — XLU, XLRE. XLE is the year's best performer but only ~3.5% of the index (38:19).
Watch for

50:40 10. Portfolio-based rebalancing — let the portfolio, not the calendar, set the trade date

The repeatable method
  1. Start from a strategic asset allocation that actually fits you: time horizon, risk tolerance, income needs, past experience — and note that financial risk tolerance and emotional risk tolerance "are two entirely different things."
  2. Set drift bands around each asset class (and, where warranted, around an individual stock or group of stocks).
  3. Rebalance when a band is breached, not when the calendar says so. Institutional programs trade the last week of a quarter, semiannually, or at year end — dates that have nothing to do with your portfolio.
  4. The rule cuts both ways: trim what has had outsized performance on the upside, and let the portfolio tell you when to add to underperforming areas.
  5. Pair it with the concentration discipline — diversify across and within asset classes, which is what keeps you participating without adding undue risk.
Here: her answer to "are you less constructive?" is no — "we have to be maybe a bit more mindful of the risks" — and this is the mechanism: stay invested, but let drift bands harvest the concentration a Neural-9-led market keeps building.
Watch for

52:35 11. The investing-vs-gambling test — participant or spectator?

The repeatable method
  1. Apply one question to any position: does it make you an owner of a stake in a business and its future cash flows — a participant in wealth creation — or a spectator who has placed a bet and can only hope?
  2. "Gambling is about hoping, not about owning." A get-in/get-out mentality is a bet on two moments in time, not a strategy.
  3. Check the odds structure: over any reasonably long horizon the odds favor the investor; in a casino they never do. If your holding period is short enough that the odds flip against you, you are gambling.
  4. Use the same test as a market-froth gauge, not just a personal one. Count the vehicles that only make sense to a spectator: single-stock ETFs, leveraged and inverse products, heavy retail options activity, sports betting and prediction-market crossover among younger investors.
  5. Compare against the last episode — "shades of 2021," the meme-stock and SPAC crazes — to size how far the froth has run.
Here: "pockets of sentiment froth" rather than a systemic bubble — and she is explicit that this blurring, which she and Kevin Gordon wrote up as Gamblers Blues (April, still on the Schwab site), worries her more "than some '08 crisis ahead of us and a deep long-lasting bear market."
Watch for

Methods distilled from the public YouTube video (transcript in transcript.txt) for personal study. Not investment advice. © The Master Investor Podcast / Master Investor Limited for source material.