← Analysis page  ·  Michael Lebowitz hub  ·  Research hub

Actionable insights — Bonds To Reverse Soon As Yields Approach "Line In The Sand"?

The repeatable analysis behind the bond call: not what he'd buy, but how he decides — written so the process can be rerun at the next yield scare.
2026-SEP-10 · Thoughtful Money with Adam Taggart · Michael Lebowitz (RIA Advisors) · ▶ Watch · full analysis · transcript
How to read this page: each insight is a method — the check he runs, the steps that turn it into a decision, 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.

10:13 1. The fundamentals-vs-narratives ledger — audit a yield move before you fear it

The repeatable method
  1. Draw two columns. Fundamentals (what sets yields in the long run): core CPI, trimmed-mean PCE, market breakevens, payroll trend after revisions with ADP/JOLTS as confirmation, real wages, trend growth with and without the AI build-out.
  2. Narratives (what moves yields in the short run): each story in the headlines — oil, deficits, AI debt crowding out, chip prices, foreign intervention, Fed silence, downgrades. Tag each true / false / debatable.
  3. Compare each fundamental with its level before the shock that started the move. If they are unchanged while yields are up, the move is narrative-driven: mean reversion is likely but the timing is not.
  4. Remember that narratives run both ways: when the tide turns, stories will push yields lower too.
Fundamentals (say yields too high)Narratives (driving yields up)
Core CPI 2.5% · trimmed-mean PCE 2.3% · breakevens 2.4–2.5%, all pre-Iran levelsOil re-igniting inflation · daily deficit headlines
Payrolls heavily revised; ADP and JOLTS don't confirmHyperscaler AI debt crowding out · memory-chip prices (Apple price rises)
Growth ~1/3 of trend · real wages flat · ex-AI growth flat to negativeBoJ yen intervention · "cocktail-party inflation" · Warsh gives no guidance · downgrades
Here: every fundamental matched its pre-Iran level, so he treats the climb toward 5% as a narrative overshoot: Government bonds Positive, but "bond yields can go higher from here" (16:59).
Watch for

16:25 2. Price the tightening the market has already done

The repeatable method
  1. Before debating a Fed hike, measure how far the rates that actually reach borrowers have moved: auto, card, mortgage and corporate loans key off the 3- to 10-year yields.
  2. Treat that rise as a hike already delivered, with its usual 3–9-month lag into activity.
  3. If the market has already tightened meaningfully, an extra policy hike raises the odds of a policy mistake, and the long end should respond by flattening rather than rising.
Here: the 3–10-year rates are up 50–75bp, so he calls a Warsh hike (especially several) "a big policy mistake," and reads the post-Jackson Hole flattening (short up, long stable) as the market agreeing (35:33).
Watch for

32:22 3. Strip the outliers, weight the basket — don't count price rises

The repeatable method
  1. When someone cites breadth ("54% of goods rose more than 3%"), switch to the trimmed mean (Dallas Fed): drop the biggest gainers and losers and read the middle. Cross-check median PCE and core CPI.
  2. Ask whether the risers are heavily weighted. Count matters less than weight: orange juice up 10% is noise, exactly as the 499th S&P company is next to Nvidia at 8%.
  3. Check the trimmed mean's recent stability (months at the same level) and whether Fed research attributes the impulse to a fading one-off (tariffs).
  4. Separate the price level from the rate of change: eggs from $3 to $6 that then stay at $6 is zero inflation. Discount "cocktail-party inflation."
Here: Warsh's Jackson Hole breadth statistic vs his own preferred measure, the Dallas Fed trimmed mean at 2.28% for four or five months; median PCE 2.7%, core CPI 2.5%. The mismatch is why Lebowitz reads the pivot as talk, not a hiking cycle (33:51).
Watch for

21:25 4. Inverse-CAPE for bonds — bad trailing decades predict good ones

The repeatable method
  1. Plot trailing 10-year annualized bond returns. Mark the troughs (the 1950s, 1861): forward returns have shot up after them.
  2. Set it against equities' CAPE, where high valuations mean low forward returns. Asset-allocation signal: equities rich and bonds after a lost decade.
  3. Adjust for starting yield: a bear market with no coupon cushion (low-for-long) exaggerates trailing losses; today's ~5% coupon means you're paid to wait.
Here: CAPE ~41 implies roughly zero 10-year equity returns while bonds sit at a historic trailing low. "There's opportunities… but for the patient" (22:09).
Watch for

24:33 5. The held-to-maturity free option

The repeatable method
  1. Buy an actual bond (not a fund) in the 5–7-year sector and decide up front that you will hold it to maturity.
  2. Write the worst case: your money back plus the yield every year (5% here). If that outcome meets your needs, the downside is accepted before you buy.
  3. Write the upside branch: if yields collapse (say to 2.5%) while stocks fall 30%, sell the bond at a price gain and rotate into equities on sale.
  4. For institutions or larger sums, ladder in (some now, more higher up) rather than waiting for a round-number yield.
Here: "essentially a free option" on a 5-year at 5% (24:33); for retirees, "buy a five or seven-year bond… tell yourself you're holding it to maturity" (45:32); insurers and pensions already laddering in below 5%; Corporate bonds above 5% at a spread.
Watch for

28:42 6. Wait for confirmation — buy 4.50 heading to 2.50, not a knife at 5

The repeatable method
  1. If you already own the asset, don't root for lower prices to add. Define what confirmation of a reversal looks like before acting.
  2. His checklist: technical signs the trend has turned, the narratives starting to change, the Fed's path over the next few meetings, and a run of benign inflation prints (three months of 0.1–0.2).
  3. Accept giving up the top tick: a worse level with much higher confidence beats the best level with none.
  4. Pick the instrument only when the shape of the move is known: a front-end rally with a pinned long end argues for 3- and 5-year paper, corporates, or options rather than duration.
Here: RIA's 60/40 (about 10% cash, a 20-year bond, shorter bonds, mortgages) holds and waits: "there's none of that right now." He'd rather buy the 10-year at 4.50% heading to 2.50% than catch it at 5% (26:46).
Watch for

47:56 7. Benchmark to the plan's required return, not the S&P

The repeatable method
  1. Build the financial plan first: planned spending (a car every five years, a wedding, travel), inflation, and assets. Let the software output the probability of success at a given return.
  2. Take the return that gives a high probability (e.g. 98% at 5%) as your wealth benchmark.
  3. Look for the safest instrument that locks that return in: a 30-year Treasury, a 5- or 10-year corporate. If it exists, lock it in rather than chase the index.
  4. Stress-test the plan (holdings cut in half, lost income) to find true vulnerabilities, not imagined ones.
Here: "My goal isn't to beat the S&P… my wealth benchmark in this example is 5%"; with 5% on offer in Government bonds and Corporate bonds, adding fixed income "secures that a little bit more" (48:19).
Watch for

Methods distilled from the public YouTube video (transcript in transcript.txt) for personal study. Not investment advice. © Thoughtful Money / Adam Taggart for source material.