Actionable insights — Gundlach Unlocked: Positioning for Higher Rates and Persistent Inflation

4:02 1. The Fed follows the two-year Treasury — watch the gap

1. Plot the 2-year Treasury yield against the fed-funds rate. The 2-year leads; the Fed lags. When the 2y pulls away from funds, the Fed is "behind the curve" and will be dragged to follow.
2. Read the direction of the gap: 2y rising well above funds → hikes are coming; 2y collapsing well below funds (the chart's red shaded area) → cuts are coming, possibly a jumbo catch-up cut.
3. Use it to fade consensus Fed forecasts: the market's rate path is usually wrong about timing because it underweights what the 2-year is already telling you.

• The 2y-minus-fed-funds spread each meeting; a 2y that peaks/troughs ahead of the Fed; jumbo "catch-up" cuts when the gap got too wide.

6:46 2. The prices-paid / employment scatter — are you in a "hike zone"?

1. Plot manufacturing employment (x) against manufacturing prices-paid (y), then color each month by what the Fed did next — red for hikes, blue for cuts (the JPM Asset Management chart).
2. The upper-right quadrant (tight labor + high input prices) is overwhelmingly red — historically the Fed hikes there. Blue dots in that zone are the anomalies worth naming: politically pressured easing (Burns under Nixon, 1970s) or genuine emergencies (Bear Stearns, March 2008).
3. Locate today's reading. If it's deep in the upper-right hike zone, a cut would be a historical anomaly — so position for "higher for longer / possible hike," not for easing.

• ISM prices-paid and employment readings; whether the current dot drifts toward the cut-quadrant or stays in the hike-quadrant; whether any easing would be "emergency" or "political" (the anomaly tells).

14:45 3. DoubleLine's two-input 10-year fair-value model

1. Estimate where the 10-year US Treasury "should" be from just two inputs: the 7-year moving average of US nominal GDP, plus the German 10-year yield. (Counter-intuitive, but it's tracked uncannily since 2021.)
2. Compare the model output to the spot 10-year. When spot is near the model, the yield "makes sense"; gaps flag dislocation.
3. Forecast the next move from the inputs: rising nominal GDP (from inflation) and rising German/developed yields both push the fair value up.

• The 7-year nominal-GDP average and the German 10-year each month; a widening spot-vs-model gap as a dislocation signal.

16:46 4. Standard-deviation bands as a regime-break detector

1. Draw a long-run trend line through a yield series with ±1 and ±2 standard-deviation bands (e.g. the 30-year Treasury since 1990).
2. A clean break outside the band that has contained the series for decades is a regime change, not noise — don't assume mean reversion back into the band.
3. Pair it with the structural cause (here: the 2020 secular bottom in rates ending the 1980–2020 declining-rate era) before concluding the old relationships still hold.

• Series persisting outside long-run bands; extraordinary policy (YCC) as the only plausible path back in; the structural driver that broke the prior relationship.

33:05 5. Use "unfettered" import/export price indices as the honest inflation gauge

1. Prefer inflation series with the fewest adjustments. Import/export price indices are "real prices" — no seasonal adjustment, no hedonics, no quality tweaks — so they're harder to massage than CPI/PCE.
2. Average the export and import year-over-year rates for a cleaner read of underlying inflation.
3. Cross-check against the official series; if the unfettered gauge runs hot while headline looks tame, distrust the headline (and the trimmed-mean tricks that strip the highest inputs more than the lowest).

• The import/export YoY prints; divergence between the unfettered average and headline CPI/PCE; trimmed-mean asymmetry (top-31% vs bottom-24% trims).

31:25 6. Energy & ISM prices-paid lead services inflation by ~8 months

1. Lag the energy CPI component (and the ISM prices-paid index) forward ~8 months and overlay it on CPI services inflation — both lead services by roughly the same window.
2. Read the implied path: a recent energy surge (and ISM prices-paid rising 55→71) projects services inflation toward 5%+ in the coming months.
3. Use it to anticipate the Fed's discomfort before the data confirm it, rather than reacting to the lagging services print.

• Energy CPI and ISM prices-paid turning points; the ~8-month-forward services-inflation implication; whether services actually follows the lead.

35:27 7. The mega-IPO wave as a market-top clock

1. Treat the largest IPOs in history as a top signal — they cluster near market peaks (not to the day, but the vicinity) because euphoria is what lets companies raise the most.
2. Sharpen it with the seller's identity: when mega-cap private companies that could stay private choose to sell now, that's a statement that they don't think the price gets better — "not suggestive these stocks are cheap."
3. Combine with a valuation extreme (Shiller CAPE near all-time highs) to confirm you're late-cycle, then de-risk concentration rather than chase the listings.

• Record-size IPOs; private mega-caps rushing to sell; CAPE at historical extremes; selloffs blamed on "making room" for the new supply.

37:44 8. Concentration analogs — measure how few names own the index

1. Quantify concentration: what share of the index sits in its top handful of names/one sector? Benchmark against the historical bubble peaks — Nifty-50 (40% in 50 stocks → 1974 bear), Japan 1989 (44% of the world index → 35 years to recover), dot-com (41% in tech/telecom).
2. Note the severity multiplier: today's "AI Big 10" is 41% of the S&P in just 10 names — far more concentrated than any prior analog.
3. Look for the rollover already underway (the share ticking down) and confirm with a parabolic component (Philadelphia semiconductor index's biggest month ever, echoing the Feb-2000 NASDAQ top). Conclusion: don't double down on the concentrated portfolio — diversify away from it.

• Top-10 / single-sector index weight vs the bubble analogs; that weight rolling over; a parabolic sector index (semis) as the late-stage tell.

43:52 9. US-vs-RoW relative performance is a dollar trade

1. Anchor the relative-value case in valuation: MSCI price-to-book, US 5.7 vs rest-of-world 2.4 — under half, and there are historical moments when they converged, so a large US underperformance is "not unthinkable."
2. Overlay US-vs-EM relative performance on the trade-weighted dollar — they "zig and zag almost perfectly together." A falling dollar → EM (and RoW) outperform the S&P.
3. Extend the same logic to debt: EM local-currency debt (JPM index) vs US corporates moves inversely with the dollar and has outperformed for ~4 years. Form the dollar view first, then let it drive the equity and debt allocation.

• US-vs-RoW price/book gap; the EM-vs-S&P line tracking the (inverted) trade-weighted dollar; EM local-currency debt vs US corporates.

46:12 10. Watch private-credit / BDC marks for the cracks

1. Track the reported marks on the largest BDCs/private-credit vehicles over time. A vehicle moving from 100 to 77 in five months means deep impairment — every loan down 23 points, or half down 46, or a quarter down 92 — "neither is good."
2. Distrust private-market disclosure: firms underreport concentration (a software-for-healthcare loan booked as "healthcare"), and the biggest private borrowers (OpenAI, Anthropic) won't show their books at all.
3. Use the marks as an early-warning gauge for the over-concentration that the public software/AI selloff will surface in private credit.

• BDC/private-credit mark declines; mislabeled sector exposure; refusals to share financials; the public-software selloff bleeding into private-credit headlines.