Jeffrey Gundlach — Gundlach Unlocked: Positioning for Higher Rates and Persistent Inflation

0:06 Why this webcast exists — funds-and-ETFs, macro-first

• Episode 2 of Gundlach Unlocked. Created it because listeners kept asking "which of your funds do you recommend I buy?" — so this show walks a macro view that supports an asset allocation across DoubleLine's mutual funds and ETFs.

0:35 The Bloomberg Agg: 30-year avg yield 4.05%, today 4.78%

• The Agg's 10-, 20-, and 30-year average yields are within a basis point of each other (~4.05%) — lower than most would guess, because ~15 years of zero-rate "financial suppression" (then COVID money printing → inflation) dragged the average down. Today's 4.78% resembles 2006–08, and so do the market-excess conditions.

2:19 Developed-market long rates have surged since 2019–20

• UK 30-year now 5.60% (rose most), US third at 5.01%, Switzerland lowest at 0.65%. US 1%-yield 2020 bonds fell over 52% by 2022 and never recovered — the 30-year is the highest in the exhibit's history.

3:27 Avoid long-term government bonds — even into recession

• His multi-year warning stands: avoid long-dated developed-market government bonds. 30-year rates won't fall significantly even in recession, and "especially if the Fed were to cut" — the market no longer expects cuts, a dramatic shift from January/March.

4:02 The Fed follows the two-year Treasury (2004→today)

• One chart proves it: the 2-year leads, the Fed lags "dragging kicking and screaming." 2004–07 the Fed chased the 2y up; the 2y then peaked and fell first into 2007–08; same in 2015–18 and 2021–23. In 2022 Gundlach told CNBC the Fed should have hiked 200bp (not 25) that first meeting — and it had to do 200bp over the next meetings; then it stayed too high as the 2y collapsed and needed a jumbo cut to catch up.

6:46 The JPM scatter: today's reading is where the Fed hikes

• JPM scatter (manufacturing employment x prices-paid y; red dots = hikes, blue = cuts). Anomalous blue dots (cuts amid high employment + high prices) cluster in the 1970s (Burns, intimidated by Nixon) and March 2008 (Bear Stearns emergency). The orange "you are here" circle sits in that upper-right hike zone — so a hike in 2026–27 is more prudent than a cut.

10:07 No cuts in 2026 — if anything, a hike

• Investors entered the year banking on ≥2 cuts; Gundlach (flagged on Fed-day in January) sees no chance of cuts in 2026 and would bet on a hike — a headline that "went instantly to top go on Bloomberg."

10:48 Volcker — "when men were men"

• The scatter's lower-left (weak employment + low prices-paid) shows Volcker hiking in Jan/Sep '82 and Aug '85 — "I don't care what happens to unemployment… we've got to get rid of this inflation problem." 1979–83 he ignored the 2-year entirely: hiked from 14% to 20% (2y never above 15.5), cut to 11.5% (2y still ~14.5), then back to 20% — radically agile, not curve-following.

13:37 Will Warsh be Volckeresque? Kill the SEP/dot-plot

• New chair Warsh's first press conference is next week. Indications he'll eliminate the SEP (the economic/inflation projections "never been right") and possibly the dot plots entirely — which Gundlach hopes ushers in more agile, Volcker-style behavior rather than the cynical 2-year-following game since 2004.

14:45 DoubleLine's 10-year fair-value model = 4.53%

• His 10–15-year-old model pegs the 10-year Treasury off two inputs — 7-year moving average of US nominal GDP + the German 10-year yield. It currently reads exactly the spot 10-year at 4.53%, has tracked uncannily since 2021, and points higher (rising nominal GDP from inflation + rising German/developed yields).

16:46 30-year broke its 30-year band — entitlement cliff

• The 30-year sits at 20-year highs and broke above its ±2σ band that held from 1990–2023; he doubts it returns absent yield-curve control or other extraordinary measures. A major retirement-entitlement program just said it runs out of money by 2032 — which he reads as more like 2029 — another reason to avoid long Treasuries.

18:37 Consumer sentiment: lowest in 45 years, all income groups

• University of Michigan sentiment is at a 45-year low. His prediction model (personal spending, U3, CPI, S&P) — which tracked sentiment 1980–2020 — now says it "should" be ~95; instead it's ~44. The break dates to 2020's secular bottom in rates: with rates now rising, the old 1980–2020 relationships no longer hold (homes unaffordable, mortgages from 2% to mid-6%).

22:57 Sentiment by party + a collapsing savings rate

• By party: Democrats 32, independents ~40, Republicans ~84 (oddly high). The personal savings rate is among the lowest on record and falling sharply — people draw down savings out of need, not desire; a drop to 2005–06 levels is "what happened" right before a financial crisis.

24:22 Inflation is running hotter than consensus

• Headline PCE rising sharply (oil/energy + broad commodities); headline CPI 3.8% — a three-year high, far from the 2% target. For five years 2% was a ceiling; since 2021 it's barely a floor.

25:40 The trimmed-mean "embarrassment" + the core-CPI gimmickry

• The Dallas Fed trimmed-mean PCE (2.33%) trims the top 31% of inflation inputs but only the bottom 24% — "obviously designed to engineer a lower look." Same lineage as CPI → ex-energy → ex-food-and-energy → "super core" ex-food-shelter-energy, stripping out the things everyone must consume.

27:44 The 1966–82 inflation overlay — "second burst"

• An uncanny overlay of 1966–82 inflation onto 2024–today; the lines are nearly identical, and we just entered what aligns with the early-'80s second inflation burst. Tracked monthly.

28:42 Math: getting back to 2% trend needs ~zero inflation

• To restore a true 25-year 2% trend, you'd need 0% inflation for 5 years (impossible at ~4% now), 1% for 10, 1.3% for 15, 1.5% for 20, 1.6% for 25 — none likely barring a deflationary spiral. He suspects the Fed retires the 2% target (or any target) and acts Volcker-style; 3% inflation devastates purchasing power over 20 years.

31:25 Forward inflation: energy & ISM lead services by ~8 months

• Warsh on CNBC next week. Energy leads CPI services by ~8 months and just surged → services inflation may run to 5%+. ISM prices-paid (55→71, also an ~8-month lead) corroborates. His favorite "real prices" — the unfettered import/export price indices (no seasonal/hedonic adjustments) — read export +8.9%, import +4.2%, averaging ~6.5% right now.

34:15 Commodities: keep a sleeve, buy nearer the 200-dma

• Bloomberg commodity index +100% from early-2020 to Q2-2022, another bull run from mid-2025, now correcting (broke the 50-dma, met the 100-dma). A move to the 200-dma "would be a tremendous buying opportunity" — commodities serve a role in the portfolio even here.

35:05 Equities: Shiller CAPE near all-time highs + a mega-IPO top signal

• The CAPE (Shiller PE, since the 1880s) is basically the highest ever, matching the dot-com peak — amid a massive AI/SpaceX IPO wave. Largest-IPOs-in-history land near market tops; mega-cap privates (SpaceX ~$1.8T, ~4× oversubscribed; plus OpenAI, Anthropic) choosing to sell now signals a hype cycle on steroids like 2000, not cheap stocks.

36:36 US vs rest-of-world: bet on RoW outperforming

• US outperformance after the GFC reached ~3–4 standard deviations above the long-run average; the rest of the world has recently turned to outperform. "I would bet very heavily that continues in the years ahead" — another leg down in US-vs-RoW relative performance.

37:44 Concentration: AI Big-10 = 41% of the S&P (the bubble analog)

• BofA's bubble-peak chart: Nifty-50 (40% of S&P in 50 stocks, → '74 bear market), Japan '89 (44% of the world index, 35 years to recover), dot-com (41% in tech/telecom). Today's "AI Big 10" is 41% of the S&P in just 10 stocks — "already reversing." A very bad time to double down on a concentrated portfolio. The Philadelphia semiconductor index just had its biggest month ever (April), echoing the Feb-2000 NASDAQ top.

40:15 Data-center overbuild + Bitcoin/software as speculation gauges

• Office construction stalled post-COVID while data-center construction surged after ChatGPT (2022) and crossed over — a coming surplus/overhang that busts the hype cycle driving the market. Bitcoin and the software-sector ETF tracked tick-for-tick (same speculative impulse) and rolled over together into spring 2026.

42:24 Private-credit opacity — software exposure mislabeled

• The software weakness will resurface as private-credit problems with over-concentration. Private-credit firms underreport software exposure (a software-for-healthcare loan gets booked as "healthcare"). OpenAI and Anthropic won't show their books — one refused a lender's request to see numbers, hinting the numbers are being "dressed up" to fuel a higher IPO price.

43:52 US price/book 5.7 vs RoW 2.4 → position for EM & a falling dollar

• MSCI price-to-book: US 5.7 vs rest-of-world 2.388 — under half; the US could substantially underperform non-US stocks, and "I will be recommending that investors position for such an outcome." EM has outperformed the S&P for 1.5 years, tightly correlated to the trade-weighted dollar (the inverse); with the dollar falling and expected to keep falling, EM should keep beating the S&P. EM debt (JPM local-currency) has beaten US corporates for ~4 years — his strongest recommendation a year ago, still in place.

46:12 BDC marks cracking — 100 → 77

• BDC assets have ballooned from nothing six years ago. One of the largest BDCs was marked at 100 on Dec 31, 2025 and at 77 by May 2026 — meaning every loan dropped 23 points, or half dropped 46, or a quarter dropped 92. "Neither of those three things is good." Thanks listeners for their confidence in DoubleLine.