From price-yield mechanics to fiscal dominance. Master the framework used by Gundlach, Druckenmiller, and Howard Marks — real yields, credit cycles, repo plumbing, and the global central bank signal system.
Bonds are the market’s first language. Every serious macro investor reads bond markets before anything else — yields, spreads, and the shape of the curve contain more forward-looking information than any other asset class. This course teaches you to read that language fluently: from price-yield mechanics, through the Fed plumbing system, to a professional 8-panel signal dashboard. By the end, you will understand why a 15bp BOJ hike can collapse the Nikkei in a day, why the 2022 inflation shock was visible in TIPS markets a year early, and why Howard Marks checks credit spreads before anything else every morning.
“The bond market is the most important market in the world. I look at it first every single morning.”
— Stanley DruckenmillerNo jargon. No formula without intuition first. Before anything else, you need to feel why bonds matter — not just know it as a fact. By the end of this module, you’ll have a simple two-gauge framework that will serve you for the rest of this course and the rest of your investing life.
The single most important idea in this entire course is this: bonds react to stress before stocks do.
Why? The bond market is dominated by institutional players — banks, insurance companies, pension funds. These players are extremely sensitive to funding costs and risk. When something starts going wrong, they adjust their bond portfolios first, often days or weeks before equity investors notice. Stocks are where retail investors watch the news and panic. Bonds are where professionals quietly move billions at the first whiff of trouble.
Learning to read bonds is like learning to read the professionals’ body language.
Bond prices and yields move in opposite directions. You’ve probably heard this. Let’s make sure you feel it.
Imagine you buy a bond today that pays 5% per year. You’re happy. Then tomorrow, new bonds come out paying 6%. Suddenly your 5% bond looks less attractive. If you tried to sell it, you’d have to offer a discount. Your bond’s price drops.
That’s it. When yields rise (new bonds pay more), existing bond prices fall. When yields fall (new bonds pay less), existing bond prices rise. They’re on a seesaw.
Why this matters: When investors get scared, they rush into Treasuries — the safest bonds on earth. That massive wave of buying pushes Treasury prices up and yields down. So when you see the 10-year Treasury yield dropping fast on a random Tuesday morning, money is probably fleeing risky assets and hiding in safety. That falling yield is your first signal.
You’ll hear this word constantly in macro. The simplest way to think about it:
Liquidity = how easily can money flow through the system?
When liquidity is abundant, banks lend freely, companies borrow cheaply, and investors take risk. When it dries up, borrowing gets expensive, banks pull back, and prices fall — sometimes violently.
Think of it like water pressure in a building. When pressure is good, every tap works fine. When it drops, the top floors lose water first (risky assets like stocks and high-yield bonds), while the ground floor still has some flow (government bonds). Bond markets are the pressure gauge.
Not all bonds are equal. Treasuries are issued by the U.S. government — essentially risk-free. Corporate bonds are issued by companies that can go bankrupt. So when you lend to a company, you demand extra yield on top of the Treasury rate. That extra yield is called the credit spread.
Example: 10-year Treasury yields 4.0%. Ford’s 10-year bond yields 5.5%. The credit spread is 1.5% (150 basis points). That 150bp is the market’s price for Ford’s risk.
The critical insight: credit spreads move based on fear, not just the company’s actual finances. When the system gets nervous about anything, investors demand wider spreads on all corporate bonds. It’s a collective risk appetite signal.
You now have two gauges:
| Spreads tight | Spreads widening | |
|---|---|---|
| Yields falling | Calm — rate cuts expected | ⚠ Red alert — liquidity crunch |
| Yields rising | Growth optimism | Worst case — stress + inflation |
Yields falling + spreads widening simultaneously = investors rushing to safety AND refusing to lend to corporations. That’s the genuine alarm. Think of it as: everyone is selling their houses (stocks fall), some people move into a safe building downtown (Treasuries), and banks are simultaneously refusing to approve mortgages (spreads widen). When all three happen together, the system is seizing up.
The Federal Reserve sets one key rate — the federal funds rate. Every interest rate in the economy is influenced by it. But bond markets don’t wait for the Fed to act. They anticipate.
If traders believe the Fed will cut rates next year — because inflation is cooling or growth is slowing — they want to lock in today’s higher yield before it disappears. So they buy 10-year Treasuries now. That wave of buying pushes yields down before the Fed has done anything.
Real example: In late 2023, the Fed hadn’t cut rates at all. But inflation was cooling and markets priced in 4–6 cuts for 2024. The 10-year yield dropped from ~5% in October to ~3.9% by December — purely on the expectation of future cuts. The bond market overshot. The Fed only delivered 3 cuts. Yields had to adjust back up through 2024.
The lesson: when you see yields moving, always ask what expectation is being priced in — and whether it could be wrong. If the market is pricing an aggressive outcome that may not play out, you’re spotting a potential reversal before it happens.
Short-term bonds (2-year Treasuries) are closely tied to Fed policy expectations. They basically ask: where does the market think the Fed funds rate will be over the next two years?
Long-term bonds (30-year Treasuries) respond to big-picture fear. When a genuine crisis hits, institutions don’t want to park cash for 2 years — they want the longest, safest asset they can find. And because of duration, a 30-year bond is roughly 10x more price-sensitive than a 2-year bond. So a big rally in long bonds requires enormous buying pressure — which only happens when major institutions are making a decisive run to safety.
March 2020: the 30-year Treasury yield dropped from 2.0% to 1.0% in two weeks. That kind of move screams "institutions are terrified" — and it happened while stocks were still in freefall.
Duration measures how sensitive a bond’s price is to yield changes. A bond with duration of 8 means: if yields move 1%, the bond’s price moves roughly 8% in the opposite direction.
Higher duration = more price sensitivity = bigger swings when yields move.
During a liquidity panic, long-duration bonds move dramatically. A 2-year Treasury barely flinches. A 30-year Treasury might jump 15–20% in a crisis. When you see headlines like “long bonds are rallying hard” — that’s duration amplifying the flight-to-safety signal. The bigger the move, the louder the alarm.
Most beginners watch the level of a signal. The professionals also watch the speed of change.
A credit spread widening 50 basis points over 3 months is a gradual deterioration — you have time to think. The same 50 basis points in one week is an alarm requiring immediate attention.
Level answers: are we in trouble?
Speed answers: how much time do I have?
Later in this course we’ll formalise this into a velocity score you can run every week. But the intuition starts here: watch not just where spreads are, but how fast they’re moving.
In mid-2007, U.S. stock markets were near all-time highs. The S&P 500 peaked in October 2007. But Moody’s Baa credit spread had already started climbing in mid-2007 — moving from ~1.2% to ~1.7% before most equity investors noticed anything wrong.
By November 2007, the Baa spread had climbed to ~2.0%. The stock market still hadn’t broken down. By March 2008 (Bear Stearns collapse), the spread was near 3%. Stocks were down ~15%.
The entire time, the credit spread was screaming while equities were still rationalising. The 10-year Treasury yield was also falling — the two-gauge red-alert pattern was visible for months before the equity collapse. Level told you something was wrong. Velocity told you it was accelerating.
In 2022, both Treasury yields and credit spreads rose at the same time — the bottom-right box of the matrix. The 10-year yield climbed from ~1.5% to ~4.2%. HY credit spreads widened from ~300bp to ~600bp.
The 60/40 portfolio lost ~16% — its worst year since 1937. Both stocks and bonds fell simultaneously. The traditional playbook failed because this was the one scenario where bonds offer no shelter: yields rising + spreads widening = inflation shock + stress. Nothing hides you.
The early warning sign was there in January 2022 — inflation expectations (5-year breakevens) were already above 2.5% and climbing. The regime had shifted. Those watching it saw it coming.
Go to FRED and pull up two series: GS10 (10-year Treasury yield) and BAA10YM (Moody’s Baa credit spread). Don’t analyse yet. Just scroll through the last 20 years.
Find 2008 and 2020. Look at what happened to the Baa spread — those spikes. Then look at how the 10-year yield moved at the same time. You’re looking for the red-alert pattern: yields falling + spreads spiking. Once you can spot it visually, you’ve internalised the most important pattern in this course.
Find one week in the last 5 years where Treasury yields fell. Then check: did credit spreads widen at the same time, or stay flat? Classify the move using the 2x2 matrix (calm repositioning vs red alert vs growth optimism vs worst case). Write 2–3 sentences on what you conclude. This is the core analytical habit you’ll repeat every week for the rest of this course.
Q1. You see the 10-year Treasury yield drop 25 basis points in one week, while credit spreads barely move. What is the most likely explanation — and is this a red alert?
Q2. Treasury yields are falling AND credit spreads are widening at the same time. What is happening and what should you do?
Q3. A 30-year Treasury bond rallies hard (price up, yield down sharply) while a 2-year Treasury barely moves. What does this tell you?
Q4. Credit spreads are at 550 basis points. Last week they were at 520 bps. Two weeks ago they were at 490 bps. Does level or velocity concern you more here — and why?
“The yield curve has predicted every US recession since 1970 — six to twenty-four months in advance.”
— Estrella & Mishkin, Federal Reserve Bank of New YorkThe yield curve is a distributed forecast of future rates, growth, inflation, and risk appetite. Learning to decompose it into expectations and term premium, to read the BOJ/JGB global interconnections, and to interpret inversion vs. steepening changes how you see every bond market move.
When term premium rises (Oct 2023: TP10 hit +0.45%, highest since 2014), this means investors demand more compensation for duration risk — not that they expect more rate hikes. Different cause, different implication.
| Curve Signal | Typical Lead to Recession |
|---|---|
| Inversion begins | 12–18 months |
| Maximum inversion | 6–12 months |
| Re-steepening begins | Recession arriving NOW |
The recession arrives after the curve re-steepens — as the Fed cuts (short end falls) or long-end rallies on recession fears.
BOJ held 10Y JGB at 0% (later 1%) from 2016–2024, purchasing ~50% of the entire JGB market. Japan became the largest foreign holder of US Treasuries (~$1.1T). When BOJ abandoned YCC, JGB yields could rise freely — making JGBs competitive with US Treasuries for the first time in a decade. This is the most underappreciated structural headwind for US Treasury demand.
| MOVE Level | Signal |
|---|---|
| Below 100 | Complacent / normal |
| 100–130 | Elevated — watch |
| 130–150 | Stress — reduce risk |
| Above 150 | Panic — defensive |
On July 31, 2024, the BOJ raised rates by 15 basis points. The estimated yen carry trade was $4 trillion: borrow JPY at near-zero, invest in USD assets at 4–5%.
The cascade: USD/JPY: 160 → 142 in 3 weeks. Nikkei: −12% in one day (Aug 5 — largest since 1987). VIX: spiked to 65 intraday. US tech sold off as carry participants liquidated to cover yen losses.
Not a recession. Not a credit event. A 15bp BOJ hike triggered a pure carry unwind. The MOVE index signalled first — rising for two weeks before the equity move.
Download the NY Fed ACM term premium data (newyorkfed.org). Import into Excel. Chart the 10Y term premium alongside the 10Y yield from 2010–present. Focus on October 2023 (10Y at 5%). What fraction was expectations vs. term premium? What was TP10 that month? Write a 300-word analysis.
Q1. The 2s10s inverts in Q3 2024. A colleague says “recession is imminent — sell equities now.” What is wrong with this and what signal should you wait for?
Q2. In Oct 2023, 10Y hit 5.0% and the ACM term premium was +0.45%. What does this tell you about the market’s view on future Fed hikes?
Q3. Why is a 15bp BOJ hike potentially more disruptive than a 25bp Fed hike?
“Credit spreads are the most honest assessment of the economy’s health. They cannot be manipulated by narrative or spin.”
— Howard Marks, Oaktree CapitalHoward Marks checks credit spreads before anything else every morning. Spreads aggregate the collective intelligence of thousands of professional investors making real-money decisions about default risk. When spreads widen, someone sophisticated has decided the risk of lending has increased — and they are usually right.
OAS (Option-Adjusted Spread) strips out embedded option value for cleaner comparison. CDS leads cash bond spreads by 1–5 days because CDS is dominated by active institutional hedgers, not passive buy-and-hold holders.
Where EDC = expected default component and EBP = excess bond premium (residual credit supply conditions). When EBP rises, lenders are tightening beyond what default risk justifies. EBP leads recessions by 3–6 months.
(1) Lenders relax → spreads tighten → leverage builds → asset prices rise. (2) Single shock → doubt → lenders pull back. (3) Lenders panic → credit freezes → forced selling → spreads blow out. (4) Reset. The cycle ends when lender behaviour changes — not when borrowers default.
| Regime | HY OAS (bps) | Baa Spread (bps) |
|---|---|---|
| Normal | 300–400 | 100–150 |
| Elevated | 450–550 | 175–225 |
| Stress | 600–800 | 250–350 |
| Crisis | Above 800 | Above 350 |
FRED: BAMLH0A0HYM2 (HY OAS), BAA10YM (Baa), BAMLC0A4CBBB (BBB)
~$4 trillion of BBB-rated bonds outstanding. A mass downgrade to junk would force insurance companies, pension funds, and regulated holders to sell — creating an estimated $1–2T of forced liquidation. This is why BBB spreads deserve as much attention as HY spreads.
Oil collapsed from $100 to $26. Energy HY (~15% of the HY index) imploded. HY OAS: 350 → 840 bps.
But: IG spreads barely moved. Financial sector CDS was calm. S&P only −15%.
The lesson: always decompose the widening. HY OAS at 840bp looks like a 2008-level crisis — but when IG is contained and bank CDS is calm, it is idiosyncratic, not systemic. Correct response: avoid energy credit, do NOT make a global risk-off call.
Download BAMLH0A0HYM2 (HY OAS) and BAMLC0A4CBBB (BBB spread) from FRED. Find every period 2010–present where BBB spread widened >30bps while HY OAS moved <50bps. What does BBB-only widening tell you about the stress? Write a 300-word analysis.
Q1. HY OAS: 680bps. Baa: 140bps. Financial CDS: calm. What does this tell you?
Q2. Why does CDS lead cash bond spreads structurally?
Q3. At what point does the Howard Marks credit cycle end?
“The repo market is the heart of the financial system. When it stops beating, everything stops.”
— Zoltan PozsarThe repurchase agreement market processes over $4 trillion per day in the US alone. It is the plumbing through which banks fund their inventories and money market funds deploy overnight cash. Most investors have never looked at it. That is precisely why understanding it is an edge.
| FRA-OIS Level | Regime |
|---|---|
| Below 10 bps | Normal |
| 10–25 bps | Elevated — watch |
| 25–50 bps | Stress |
| Above 50 bps | Crisis (March 2020: 80bps) |
When implied dollar rate exceeds SOFR, dollars are scarce globally. March 2020: EUR/USD basis hit −120bps — extreme dollar hoarding.
WALCL = Fed balance sheet total. WTREGEN = Treasury General Account. RRPONTSYD = Reverse Repo usage. RRP draining → easing. TGA refilling → tightening — both mechanical, without any Fed rate action.
FRA-OIS spiked to 80bps. EUR/USD basis hit −120bps. Treasury bonds sold off as institutions liquidated to raise dollar cash.
The Fed activated swap lines with 14 central banks. FRA-OIS collapsed from 80bps to 15bps within two weeks of swap line activation — before equity markets bottomed (March 23). The practitioner monitoring FRA-OIS identified the stress resolution as the leading indicator, not the equity low.
Download WALCL, RRPONTSYD, WTREGEN from FRED (weekly). Calculate NL = WALCL − WTREGEN − RRPONTSYD for each week from Jan 2022–present. Plot alongside S&P 500. Identify TGA refill periods in 2023 — did SOFR spike during those windows?
Q1. The TGA rises $400B over 6 weeks (Treasury issuing debt). No QT, no rate changes. What happens to bank reserves?
Q2. FRA-OIS spikes from 12bps to 55bps in 3 days. What does this signal and what Fed tool resolves it fastest?
Q3. QT is running at $60B/month and a $600B TGA refill occurs over 3 months. What is the true liquidity drain?
“When short-term rates hit zero and fiscal stimulus isn’t enough, the central bank must print money and buy assets. That’s Stage 5.”
— Ray Dalio, Principles for Navigating Big Debt CrisesThe Fed balance sheet is the most powerful macro lever in the post-2008 world. Understanding the precise transmission mechanism — from bond purchase to reserve creation to dealer inventory to spread tightening — lets you front-run the effects of policy changes that most participants respond to reactively.
Fed buys bonds → reserves created → dealer funding cost falls → dealers expand inventory → market-making improves → spreads tighten → risk premium falls → equities and credit rally. QE works by reducing the cost of risk-taking for dealers.
Bonds roll off → reserves drain → dealer funding costs rise → dealers reduce inventory → bid-ask spreads widen → market depth falls → volatility rises → risk premium expands → spreads widen. QT is asymmetric: tightening moves happen faster than easing moves.
| Reserve Level | Regime | Risk |
|---|---|---|
| Above $3.0T | Ample | Low |
| $2.0–$3.0T | Adequate | Moderate — watch repo |
| Below $2.0T | Scarce | High — repo stress probable |
Sept 2019 repo seizure occurred at ~$1.4T. $2T is the established lower bound.
UK government announced an unfunded £45B tax package. 30Y gilt yield: +130bp in 3 trading days. LDI pension funds received margin calls → sold gilts → yields rose more → more margin calls → doom loop.
The BOE had been running QT, which reduced the system’s shock-absorption capacity. Without adequate reserves, dealers could not warehouse the forced selling. BOE was forced into emergency QE (£65B) to break the loop.
Lesson: QT removes the system’s buffer against exactly this type of event. The threshold between “adequate” and “scarce” reserves determines whether the system absorbs a shock or collapses into a doom loop.
Calculate monthly NL (WALCL − WTREGEN − RRPONTSYD) from Jan 2020–present. Chart alongside S&P 500 monthly close. Calculate rolling 3-month lead correlation. Isolate 2022 QT period: at what NL level did the equity bear market bottom?
Q1. Fed announces QE $80B/month. Trace the full chain to HY credit spreads. Where is the longest lag?
Q2. Reserves at $1.7T. QT at $60B/month, TGA refill of $200B expected in 6 weeks. What are you watching for?
Q3. Why did BOE’s QT make the 2022 gilt crisis worse?
“Japan is the laboratory for everything that is coming to the rest of the developed world.”
— Lacy Hunt, Hoisington Investment ManagementThe global bond market is not determined solely by the Fed. Japan’s $1.1T in US Treasuries makes the BOJ a global bond market actor. ECB/Fed divergence creates EUR/USD flows. And fiscal dominance — where debt levels are so high that central banks hesitate to raise rates — is already reality in Japan and increasingly relevant in the US.
BOJ held 10Y JGB at 0% (later 1%) 2016–2024, buying ~50% of the JGB market. Japan’s investors — forced out of zero-yield JGBs — recycled capital into US Treasuries (~$1.1T). When BOJ abandoned YCC, JGB yields could rise freely, making JGBs competitive domestically. The incentive to hold US Treasuries declined structurally — the most underappreciated long-term headwind for US Treasury demand.
When Fed hikes faster than ECB, capital flows to dollar assets → USD strengthens → tightens global financial conditions for EM borrowers with $13T in USD-denominated debt.
The reserve currency issuer (US) must run persistent current account deficits to supply global dollars. But permanent deficits undermine long-run dollar confidence. The US is trapped: provide global liquidity (deficits, weak confidence) or protect dollar value (surpluses, global dollar shortage). This tension is structural and unresolved.
When debt/GDP is so high that the central bank is politically and economically unable to raise rates sufficiently:
If r > g (real rate > real growth), debt/GDP grows indefinitely
Japan: debt/GDP ~260%, rates near zero for 30 years. US trajectory: ~125% debt/GDP, $2T+ annual deficits. Lacy Hunt argues the US is on the same path.
May 22, 2013: Bernanke mentioned the possibility of tapering bond purchases — once, in congressional testimony. 10Y yield: 1.63% → 3.0% in four months (+137bps).
EM contagion: Indian rupee: −18% | Indonesian rupiah: −17% | Brazilian real: −15%
These EM currencies collapsed not because their fundamentals changed — but because rising US yields made dollar assets more attractive, triggering capital outflows. A US bond market move is a global macro event.
On TradingView, open FRED:DEXJPUS (USD/JPY). Overlay the US 10Y − Japan 10Y yield differential. Set range 2015–present. When the differential widens, does USD/JPY rise? When it narrows (2022 YCC tweak, 2024 BOJ hike), what happens? Write a 400-word memo: what would cause the yen carry to unwind completely, and what would be the sequence of asset moves?
Q1. Japan’s debt/GDP is 260%. BOJ holds ~50% of JGBs. Why is the BOJ effectively unable to raise rates significantly?
Q2. BOJ abandons YCC and JGB yields rise to 1.5%. Why is this a headwind for US Treasuries specifically?
Q3. What is the Triffin Dilemma and is it resolved?
“The credit markets are a more reliable indicator of stress than equity markets. They price default risk with no narrative overlay.”
— Howard MarksHaving built the foundation in Module 3, we now apply professional-grade analytical tools: spread velocity to distinguish panic from trend, EBP to measure credit supply conditions, and SLOOS as a 3–6 month leading indicator of spread widening.
The rate of spread widening is as important as the level. A slow grind wider suggests a trend. An acceleration suggests a panic or forced selling event:
| Velocity (bps/week) | Classification |
|---|---|
| Below 20 | Low — trend widening |
| 20–50 | Moderate — elevated concern |
| 50–100 | High — stress event |
| Above 100 | Extreme — crisis/panic |
When EBP rises with spreads, lenders are pulling back beyond what default risk justifies — a pure credit supply shock. When EDC rises but EBP is flat, it is a default risk repricing. SLOOS (Senior Loan Officer Opinion Survey, quarterly) precedes EBP moves by 3–6 months: tightening lending standards → EBP rises → spreads widen.
Professional spread assessment integrates three dimensions simultaneously:
| Dimension | Data | Signal |
|---|---|---|
| Level | HY OAS vs. benchmarks | Normal / Elevated / Stress / Crisis |
| Velocity | Week-on-week change | Trend vs. panic |
| Direction | EBP + SLOOS | Supply or demand shock |
From October to December 2018, HY OAS widened from 320 to 530bps. Velocity analysis:
October: ~21bps/week (Moderate). November: ~35bps/week (High). December: 40–60bps/week (Extreme — accelerating).
Fed Chairman Powell signalled a pause in January 2019. Within days of the pivot: spreads reversed sharply. Equities followed weeks later.
The practitioner who tracked velocity saw: (1) the acceleration in December signalling panic, and (2) the immediate reversal on the pivot signal, days before equities confirmed. Spreads led equities by weeks.
Download daily BAMLH0A0HYM2 from FRED. In Excel, calculate the 5-day rolling change (velocity) for each business week from 2015–present. Identify every period where velocity exceeded 50bps/week (stress threshold). For each event: document the cause, duration, and subsequent mean reversion. How many were genuine crises vs. technical/seasonal widenings?
Q1. HY OAS widens from 380 to 430bps over one week. Classify the velocity and interpret.
Q2. SLOOS shows 40% net tightening for C&I loans. HY OAS is at 380bps (normal). What is the forward implication and lead time?
Q3. EBP is rising fast but EDC is flat. What does this tell you about the nature of spread widening?
“The bond market is smarter than the equity market. Watch MOVE before you watch VIX.”
— Practitioner maximThe MOVE index measures implied volatility in US Treasury options — the bond market’s own fear gauge. But the deeper analytical layer is convexity hedging: the $9T US mortgage-backed securities market has negative convexity, creating feedback loops that amplify yield moves by 20–30bps beyond what fundamentals warrant.
| MOVE Level | Signal | Action |
|---|---|---|
| Below 100 | Complacent | Normal positioning |
| 100–130 | Elevated | Reduce duration risk |
| 130–150 | Stress | Defensive positioning |
| Above 150 | Panic | Maximum caution; watch for reversal |
Normally MOVE and VIX are correlated. When MOVE rises sharply while VIX stays low, it signals stress that is isolated to bond markets — not yet priced into equities. This is an early warning of a coming equity repricing. The SVB crisis (March 2023) showed MOVE >170 while VIX only reached 26 initially. The MOVE/VIX divergence was the early warning.
The US MBS market is ~$9 trillion. MBS have negative convexity: when rates rise, prepayments slow and the MBS duration extends — forcing holders to sell Treasuries to rebalance duration. This creates a feedback loop:
Rates rise → MBS duration extends → MBS holders sell Treasuries → rates rise further → more duration extension → more Treasury selling
This mechanical loop can amplify rate moves by 20–30bps beyond fundamentals in rising rate environments.
SRVIX (swaption volatility index) measures implied vol on long-dated interest rate swaps. When SRVIX > MOVE, stress is concentrated in long-dated rate expectations — typically signalling fiscal concerns, term premium spikes, or long-end supply anxiety rather than short-term funding stress.
In 1994, the Fed hiked 250bps in 12 months. The 10Y yield moved from 5.7% to 8.0% — a 230bp rise, the most violent bond move since World War II.
The mechanism: as rates rose, MBS convexity hedgers sold Treasuries to rebalance duration. Each round of hedging pushed rates higher, triggering more hedging — a convexity feedback loop that amplified the move well beyond what the rate hikes alone would have caused.
Orange County, California went bankrupt with a $1.7 billion loss — the result of leveraged bets on short-term rates staying low. The fund’s duration had been dramatically underestimated. The massacre was not just about the Fed hikes — it was about the convexity amplification that nobody had modelled.
Overlay MOVE and VIX on TradingView from 2018–present. Find every period where MOVE exceeded 130 and VIX was below 20. Document: (1) cause, (2) how long the divergence lasted, (3) what happened to VIX subsequently. Was MOVE always the early warning? Write a one-page analysis of the SVB crisis (March 2023) using this lens.
Q1. MOVE is at 155, VIX is at 16. What does this divergence signal?
Q2. Explain why the MBS market amplifies Treasury yield moves in a rising rate environment.
Q3. SRVIX is at 130 and MOVE is at 95. What type of stress does this indicate?
“The edge isn’t in having the data — everyone has FRED. It’s in having a framework to interpret it consistently and the discipline to act when signals converge.”
— @ADZOEight panels, one daily routine. The dashboard integrates everything from Phases I–III into a single morning workflow that takes 10 minutes. After 30 days it becomes instinct. After 90 days you will identify regime shifts before consensus names them.
| Panel | FRED Code | What It Tells You |
|---|---|---|
| 1. 10Y Treasury | GS10 | Direction of the risk-free rate |
| 2. Baa Spread + 50dma | BAA10YM | IG credit stress + trend |
| 3. HY OAS | BAMLH0A0HYM2 | Risk appetite in credit markets |
| 4. MOVE Index | TVC:MOVE | Bond market fear gauge |
| Panel | Source | What It Tells You |
|---|---|---|
| 5. SOFR vs. FF Rate | FRED: SOFR | Overnight funding conditions |
| 6. RRP Balance | FRED: RRPONTSYD | Net liquidity direction |
| Panel | FRED Code | What It Tells You |
|---|---|---|
| 7. Real Yield (10Y TIPS) | DFII10 | The master cross-asset variable |
| 8. 2s10s Curve | T10Y2Y | Recession lead + term premium |
| Panel | Concern Threshold | Alarm Threshold |
|---|---|---|
| HY OAS | Above 450bps | Above 500bps |
| Baa Spread | Above 175bps | Above 200bps |
| MOVE | Above 120 | Above 130 |
| 2s10s | Crosses zero (inverts) | Re-steepens after inversion |
| Real Yield (DFII10) | Crosses +1.5% | Crosses +2.0% (headwind); −0.5% (tailwind) |
| SOFR vs. FF | SOFR > 10bps above FF | SOFR > 25bps above FF |
Convergence rule: when 4 or more panels simultaneously signal in the same direction, conviction is high enough to act. Fewer than 4 converging = monitor, do not size aggressively.
Open TradingView and FRED side by side. Check overnight: did any panel move more than 5bps (yields/spreads) or 5 points (MOVE/VIX)? Note the direction.
Did any panel cross a concern or alarm threshold? Has anything that was in “Elevated” moved to “Stress”? Is the 2s10s re-steepening after a long inversion?
Where do today’s readings put you in the 4-scenario matrix (Module 10)? Has the scenario changed from yesterday? From last week?
Example: “HY OAS widened 12bps today to 478bps (Elevated). MOVE at 118 (approaching concern threshold). Real yield unchanged at +1.4%. No threshold breach — monitoring. Scenario remains Scenario 2 (yields rising, spreads contained).”
Q1. You open the dashboard: HY OAS = 510bps, MOVE = 138, real yield = +1.8%, 2s10s = −0.4% (inverted). How many alarm thresholds are breached and what is the convergence call?
Q2. You have done the dashboard routine for 45 days. Today, DFII10 crosses −0.5% from above. What is the signal and what does it suggest for gold and EM equities?
Q3. What is the convergence rule and why is it important for position sizing?
“In 2022, the textbook broke. Bonds and equities fell together. The investors who knew why — and had prepared — were not surprised.”
— @ADZOThe four-scenario matrix is the operational framework that converts dashboard readings into positioning calls. But a critical complication: the bond-equity correlation that made 60/40 portfolios work for 30 years can flip in inflationary regimes — as 2022 brutally demonstrated.
| Scenario | Yields | Spreads | Regime | Positioning |
|---|---|---|---|---|
| 1 | Falling | Tightening | Liquidity flood | Long equities + credit (TLT, HYG, SPY) |
| 2 | Rising | Tightening | Growth boom | Cyclicals, short duration (XLF, XLE, BIL) |
| 3 | Falling | Widening | Flight-to-safety | TLT, BIL, short credit |
| 4 | Rising | Widening | Stagflation | GLD, DBC, TIP, BIL — AVOID bonds AND equities |
Scenario 4 is the rarest and most dangerous — traditional 60/40 portfolios lose on both legs simultaneously.
Disinflationary regime (1990s–2021): Bond-equity correlation = negative. Bonds up = stocks down. 60/40 works because bonds hedge equity drawdowns.
Inflationary regime (1970s, 2022): Bond-equity correlation = positive. Both fall simultaneously. 60/40 fails on both legs. The portfolio has no hedge.
Detection signal: 5Y5Y forward breakeven (FRED: T5YIFR) above 2.5% and rising = inflation regime risk. When this signal appeared in late 2021, the 2022 regime flip was foreseeable.
The classic 60/40 portfolio (60% SPY, 40% TLT) lost approximately 16% in 2022 — its worst year since 1937.
SPY: −18% | TLT: −31%
Both legs fell simultaneously. Bond-equity correlation flipped from approximately −0.4 (2010–2021 average) to +0.5. Bonds provided no hedge — they amplified the drawdown.
The early warning: FRED T5YIFR (5Y5Y forward breakeven) crossed above 2.5% in late 2021 and was rising. This was the signal that an inflationary regime — where the correlation would flip — was underway. Investors watching this signal had 3–6 months to reposition before both legs declined.
Score each panel from −2 (strongly bearish for risk) to +2 (strongly bullish for risk). Sum for a composite signal.
| Panel | −2 | −1 | 0 | +1 | +2 |
|---|---|---|---|---|---|
| HY OAS | >600bps | 450–600 | 350–450 | 300–350 | <300bps |
| Baa Spread | >250bps | 175–250 | 130–175 | 100–130 | <100bps |
| MOVE | >150 | 130–150 | 100–130 | 80–100 | <80 |
| Real Yield | >+2% | +1–2% | −0.5–0% | <−0.5% |
Score of +6 to +8: maximum risk-on. Score of −6 to −8: maximum defensive. ±2: neutral/monitor.
Q1. Yields are rising and spreads are tightening simultaneously. Which scenario is this and what is the right positioning?
Q2. FRED T5YIFR (5Y5Y forward breakeven) is at 2.8% and rising. What regime risk does this signal for a 60/40 portfolio?
Q3. On the signal scoring table, you get: HY OAS = −1, Baa = −1, MOVE = −2, Real Yield = −1. What is the composite score and what does it imply?
“The market has always taught me everything I know. The key is to study the cases closely enough that you don’t have to learn the same lesson twice.”
— Howard MarksFive case studies. For each: the dashboard signal timeline, the early warning that was available, and the exercise of writing a “What would I have done?” memo using only the data that was available at the time — no hindsight.
HY OAS timeline: Feb 2007 (280bps) → Jul 2007 (400bps) → Nov 2007 (500bps) → Mar 2008 (700bps) → Sep 2008 (2,000bps). MOVE exceeded 260.
Early warning available: HY OAS broke above 400bps in July 2007 — 14 months before Lehman. BBB spreads widened. SLOOS began tightening in Q1 2007. The credit warning was clear 6–12 months before the equity market peaked in October 2007.
Rule: When HY OAS breaks 400bps on a sustained basis with rising velocity, begin defensive positioning regardless of equity market levels.
Speed: HY OAS moved from 360bps to 1,100bps in 23 trading days — the fastest widening in history. MOVE exceeded 160.
Key insight: The Fed announced unlimited QE on March 23, 2020. Spreads tightened before a single bond was purchased — the mere announcement broke the doom loop. This demonstrates that credible CB commitment, not actual purchases, is often the mechanism.
Rule: In a liquidity crisis (velocity >100bps/week), watch for Fed intervention announcements as the primary reversal trigger — act before confirmation of purchases.
Scenario 4 all year. 10Y yield: 1.5% → 4.2%. HY OAS: 300 → 600bps. Bond-equity correlation flipped from −0.4 to +0.5.
Early warning: T5YIFR (5Y5Y breakeven) crossed 2.5% in late 2021. DFII10 (real yield) turning from −1% toward zero in Q4 2021. Both signals were publicly available on FRED months before the equity drawdown began.
Rule: When T5YIFR crosses 2.5% rising AND real yields are turning from deeply negative, exit long duration and begin building inflation hedges.
Speed and divergence: 2Y Treasury yield plunged 100bps in 3 days (fastest move since 1987). MOVE exceeded 170. VIX only reached 26 — a textbook MOVE/VIX divergence.
Early warning: MOVE had been elevated (>130) for weeks before the SVB failure. The VIX/MOVE divergence was visible. The Federal Reserve’s BTFP (Bank Term Funding Program) announced March 12 arrested the cascade within 48 hours.
Rule: MOVE/VIX divergence (>130 MOVE, <20 VIX) is the early warning that bond-market-specific stress has not yet been priced into equities. Reduce equity duration exposure when this signal appears.
1994: Fed hiked 250bps in 12 months. 10Y: 5.7% → 8.0% (+230bps). MBS convexity hedging amplified the move significantly beyond the rate hikes alone. Orange County bankrupt ($1.7B leveraged rate loss).
2013: Bernanke mentioned tapering once. 10Y: 1.63% → 3.0% (+137bps in 4 months). EM contagion: Indian rupee −18%, Indonesian rupiah −17%.
Combined rule: When the Fed pivots from easing to tightening rhetoric AND MBS duration is extended (rising rate environment), expect convexity-amplified moves. The fundamental move + convexity overshoot = total move is 30–50% larger than the rate change alone implies.
For each of the five case studies, write a 300-word memo using only the dashboard data available at the early warning point (not at the crisis peak). For 2008: use July 2007 data. For 2020: use February 28, 2020 data. For 2022: use November 2021 data. For 2023 SVB: use February 28, 2023 data. For 1994/2013: use February 1994 and May 22, 2013 data respectively.
Each memo should state: (1) current dashboard readings, (2) which scenario you are in, (3) what you would do and why, (4) what would change your view.
Q1. In the 2008 crisis, HY OAS was at 400bps in July 2007. The S&P 500 was still rising. Would you have acted? What is the argument for and against?
Q2. In 2020 COVID, why did spreads tighten before the Fed actually purchased bonds?
Q3. What is the single most distinguishing feature between the 2023 SVB crisis and the 2008 GFC in terms of spread dynamics?
“Carry is the most seductive strategy in finance. And the most dangerous when it unwinds.”
— Practitioners’ maximModule 12 integrates everything into a portfolio framework: how to express macro bond views in actual positions, how carry trades build and unwind, and the honest backtesting reality check — including look-ahead bias, regime dependence, and why percentile ranks beat fixed thresholds.
For the yen carry: borrow JPY at ~0%, convert to USD, invest in US Treasuries at 4–5%. Carry = 4–5%. Profit until: (1) JPY appreciates (FX loss on funding leg), (2) US yields fall, or (3) risk event triggers forced unwind. Estimated yen carry size in 2023–2024: ~$4 trillion notional.
Risk parity allocates capital to equalise risk contribution (volatility) across asset classes — not capital. In a disinflationary regime, this works: bonds are low-vol and negatively correlated with equities. In an inflationary regime (2022), bond vol rises AND correlation with equities goes positive — risk parity rebalancing sells bonds (as vol rises) into a falling bond market, amplifying losses on both legs.
SLOOS tightening (net % tightening >30%) + Conference Board LEI falling: High-conviction defensive signal. This combination preceded every US recession since 1990 with a 3–9 month lead time. Action: maximum defensive positioning, move to Scenario 3 or Scenario 4 playbook.
Look-ahead bias: Using today’s data to make yesterday’s decision. Always construct signals using only lagged data.
Regime dependence: Signals calibrated in a disinflationary regime fail in an inflationary one. Use percentile rank of spread level within current trailing 2-year window — not fixed levels.
Overfitting: Weekly signals are cleaner than daily. Max 2% portfolio per liquidity-regime trade to limit damage from false positives.
| Scenario | Regime | Long | Short / Avoid |
|---|---|---|---|
| 1: Yields ↓, Spreads ↓ | Liquidity Flood | TLT, HYG, SPY, EEM | BIL, SHV |
| 2: Yields ↑, Spreads ↓ | Growth Boom | XLF, XLE, XLI, BIL | TLT, TIP |
| 3: Yields ↓, Spreads ↑ | Flight-to-Safety | TLT, BIL, GLD | HYG, JNK, EEM |
| 4: Yields ↑, Spreads ↑ | Stagflation | GLD, DBC, TIP, BIL | TLT, SPY, HYG |
ETF proxies: TLT=long Treasuries, HYG=HY credit, BIL=short-term T-bills, GLD=gold, DBC=commodities, TIP=TIPS, EEM=EM equities, XLF/XLE/XLI=US sector ETFs.
Using monthly FRED data from 2010–2024, classify each month into one of the 4 scenarios (yield direction + spread direction, using month-over-month changes). For each scenario classification, calculate the subsequent 1-month and 3-month return of: TLT, HYG, SPY, GLD. Do the returns match the scenario predictions? Identify the worst-performing months where the framework failed — what was the cause?
Q1. You borrow JPY at 0.1% and invest in a US Treasury at 4.6%. The yen appreciates 8% over 6 months. What is your net P&L on the carry trade?
Q2. SLOOS shows 45% net tightening AND Conference Board LEI has fallen for 6 consecutive months. What is the positioning implication?
Q3. Why does risk parity fail specifically in inflationary regimes?