Credit Spreads — The Smoke Detector

What this article does

If you watch markets, you've heard the phrase "credit spreads are widening" thrown around like everyone already knows what it means. Most people don't, and pretending to is one of the easier ways to lose money.

This article fixes that. By the end, you will:

• Understand what a credit spread actually is, in plain terms
• Know why it matters more than the VIX, the S&P 500, or anything on financial Twitter
• Recognise the four regimes credit spreads cycle between — and what each one means for the broader economy
• Be able to look at the live Lighthouse dashboard and read what spreads are saying right now

This is the foundation. Almost every later Lighthouse article will refer back to credit spreads. Get this one right and the rest of the platform makes sense.

The friend-asking-for-a-loan analogy

Imagine two friends ask to borrow £10,000 from you for a year.

Friend A is a salaried employee with steady savings, no debts, and a good track record of paying you back small amounts in the past. You think: "fine, I'd want maybe 4% interest — basically the going rate at my bank."

Friend B is between jobs, has run up credit card debt, and last time they borrowed money they were six months late paying you back. You think: "I'll do it, but I want 9%. The extra 5% is for the risk that I might not see this money again."

That extra 5% you'd charge Friend B compared to Friend A — that's a credit spread.

In financial markets, "Friend A" is the US government. The US government can in theory always pay you back because it can issue dollars. So whatever the US Treasury pays to borrow is the floor — what economists call the "risk-free rate." Every other borrower pays more than the government, and the difference between what they pay and what the government pays is credit spread.

A credit spread is the market's price for taking on credit risk. It's compensation for the possibility that a borrower might fail to pay you back.

When investors are confident about the economy, they don't demand much extra. Spreads compress. When investors get nervous, they demand more. Spreads widen. When investors are confident about credit risk, spreads compress (narrow). When investors get worried about defaults, spreads widen (expand).^[1]

How credit spreads are actually quoted

You'll see credit spreads quoted in two ways: as a percentage (e.g. "1.7%") or in basis points (e.g. "170 basis points"). They mean the same thing — 1 percentage point equals 100 basis points.^[2]

The standard reference point is always a US Treasury of the same maturity. So a 10-year corporate bond's spread is its yield minus the 10-year Treasury yield. Credit spreads are quoted as the difference in yield between a corporate bond and a US Treasury of similar maturity, measured in basis points or percentage points.^[3]

There's also a more technical version called option-adjusted spread (OAS), which strips out the value of any embedded options in the bond. That's the version Bloomberg, Reuters, and serious credit investors actually watch. The numbers we'll talk about — and the dashboard tracks — are mostly OAS readings.

Why credit spreads matter more than equity markets

Here's the central claim of this article, and the reason Lighthouse exists.

Credit spreads tend to widen weeks or months before equity markets crack, before GDP statistics confirm a slowdown, before the Fed responds. Credit spreads tend to widen weeks or months BEFORE problems become visible in equity markets, GDP statistics, or central bank policy responses.^[4]

That's not a vague intuition — it's an empirical pattern visible across decades. Credit investors are typically more conservative, more institutional, and more focused on downside than equity investors. They also have a clearer view of leverage and default risk because they're literally the ones being asked to lend.

When credit investors get nervous, they demand more compensation. Spreads widen. By the time equity markets register the same nervousness — usually because corporate earnings start disappointing or layoffs hit the news — the credit market has already been pricing it for months.

This is why we call credit spreads the smoke detector of the economy. They go off before the fire is visible. They have false alarms — sometimes spreads widen briefly without anything bad following — but ignoring them is dangerous, because when they consistently widen, something is usually starting to burn.

The other commonly-quoted "fear gauge" is the VIX — the index of expected equity volatility. The VIX is genuinely useful, but as an early warning system it's worse than credit spreads, because by the time the VIX spikes meaningfully, credit spreads have usually already moved. The CBOE Volatility Index (VIX) tends to LAG credit spreads as a stress signal — by the time the VIX spikes meaningfully, credit spreads have usually already widened.^[5]

The investment grade / high yield split

Not all corporate borrowers are equal, and the bond market sorts them by credit rating.

Investment grade is the upper tier — bonds rated BBB- or higher (S&P/Fitch). These are large, established companies. Apple, Microsoft, Coca-Cola, most major banks. They borrow at modest spreads above Treasuries — historically around 1-2% in calm times. Investment grade refers to bonds rated BBB- or higher (S&P/Fitch) or Baa3 or higher (Moody's). High yield (or 'junk') refers to bonds rated below this threshold.^[6]

High yield is everything below that — sometimes called "junk bonds" or "speculative grade." Smaller, more leveraged, more cyclical companies. They borrow at higher spreads — historically 3-6% above Treasuries in calm times, much wider in stress. Investment grade refers to bonds rated BBB- or higher (S&P/Fitch) or Baa3 or higher (Moody's). High yield (or 'junk') refers to bonds rated below this threshold.^[6]

The line between investment grade and high yield isn't just a label. It's structurally important: many institutional investors (pension funds, insurance companies) are restricted by mandate to only hold investment grade. When a company gets downgraded across that line — what's called a "fallen angel" — there's often forced selling regardless of fundamentals. The investment grade / high yield boundary is structurally important because many institutional investors (pension funds, insurance companies) are restricted by mandate to investment-grade only. A downgrade across this line can force concentrated selling.^[7]

This matters for our purposes because the high-yield index is more sensitive to credit stress than investment grade. When investors get nervous, high-yield spreads widen first and most. So most of the time, when people talk about "credit spreads widening," they mean high-yield credit spreads.

The tier-by-tier early warning

Here's where it gets interesting. High-yield itself is divided into three rating tiers:

• BB — the highest tier of high-yield. Companies that nearly qualify as investment grade.
• Single-B — the middle tier.
• CCC and below — the lowest tier. Highly leveraged, often distressed.

These tiers move together in calm times. But when credit conditions start deteriorating, they don't move equally. CCC widens first and most, then Single-B, then BB. When credit conditions deteriorate, CCC-rated spreads typically widen first and most, then Single-B, then BB. This tier-by-tier sequencing is one of the earliest warning signals in credit markets.^[8]

This tier-by-tier sequencing is one of the earliest warning signals in finance. By the time the headline high-yield index has noticeably widened, the CCC tier may already be screaming. Watching the CCC-to-BB ratio is what serious credit investors do — it's the canary in the coalmine for the rest of the market. The US high-yield bond market is divided by credit rating into three main tiers: BB (highest quality), Single-B (middle), and CCC and lower (lowest quality).^[9]

The Lighthouse dashboard shows all four high-yield series and their tier divergence. When CCC pulls ahead of BB, that's the smoke detector starting to chirp.

What spreads have done historically

This is the part that makes credit spreads believable as a smoke detector. We have the data. Let's walk through what it shows.

The Moody's Baa-Treasury spread is the longest-running credit spread series we track on Lighthouse — daily data since 1986. That gives us 40 years of history covering every major credit event of the modern era. Here's what each looked like:

1987 — Black Monday

The October 1987 stock market crash was one of the biggest equity drops in history — the Dow fell 22.6% in a single day. You might expect credit spreads to have exploded too. They didn't. Baa-Treasury spreads peaked at just 2.68% in October 1987. During the 1987 Black Monday era (October-December 1987), the Moody's Baa10Y credit spread peaked at 2.68% on 22 October 1987 — three days after the 19 October crash.^[10]

Why? Because 1987 was a market microstructure event, not a credit event. Portfolio insurance forced selling cascaded into a crash, but the underlying companies were fine. Credit spreads barely flinched.

This is a useful early lesson: credit and equity stress can decouple. Credit spreads tell you about credit problems, not stock market drama.

1990-91 — A textbook recession

The Gulf War oil shock and S&L crisis combined into a textbook US recession. Baa-Treasury spreads peaked at 2.45% in December 1990. During the 1990-91 US recession, the Moody's Baa10Y credit spread peaked at 2.45% on 12 December 1990.^[11] Modest by later standards, but it was a real, broad-based recession with rising unemployment and falling profits.

1998 — The LTCM crisis

Long-Term Capital Management, the most famous hedge fund in the world, collapsed in late 1998. The Federal Reserve organised a private-sector bailout to prevent contagion. Baa-Treasury spreads spiked to 2.77% in October 1998. During the 1998 Long-Term Capital Management (LTCM) crisis, the Moody's Baa10Y credit spread peaked at 2.77% on 16 October 1998 — despite no formal US recession at the time.^[12]

But — and this is important — there was no formal US recession at the time. Credit spreads detected the financial stress without an underlying macroeconomic downturn. Sometimes the smoke detector sees smoke that doesn't end up burning the house down. False alarms exist. But the alarm itself was real: the financial system genuinely was unstable.

2001 — The dot-com recession

The bursting of the dot-com bubble combined with the 9/11 attacks pushed spreads to 3.53% by late September 2001. During the 2001 dot-com recession, the Moody's Baa10Y credit spread peaked at 3.53% on 21 September 2001 — ten days after the 9/11 attacks.^[13] NBER dated the recession as running March-November 2001. Credit spreads widened gradually through the year, then jumped after September 11 when markets reopened.

2008 — The Global Financial Crisis

This is the big one. Baa-Treasury spreads reached 6.16% on 4 December 2008 — the all-time peak in our 1986-onwards window. During the 2008 Global Financial Crisis, the Moody's Baa10Y credit spread reached its all-time high in our 1986-onwards window at 6.16% on 4 December 2008.^[14]

Lehman Brothers had collapsed in September 2008. Credit markets seized up. By December, spreads were screaming. The Fed had already cut rates aggressively and was about to start QE. Equity markets, which had peaked in October 2007, were down over 50%.

Credit spreads had been widening since mid-2007 — over a year before the worst of the equity damage. Anyone watching credit knew something was deeply wrong long before the equity market fully cracked.

2011 — The European sovereign debt crisis

Greek default fears, Italian bond stress, US debt-ceiling drama. Spreads peaked at 3.39% in late September 2011. During the 2011 European sovereign debt crisis, the Moody's Baa10Y credit spread peaked at 3.39% on 27 September 2011.^[15] The S&P 500 dropped about 19% during this period — close to a bear market but not quite.

2020 — COVID-19

This one's particularly instructive. Spreads went from below 2% in mid-February 2020 to 4.31% by 23 March 2020 — the fastest widening in our data. During the COVID-19 market shock, the Moody's Baa10Y credit spread peaked at 4.31% on 23 March 2020 — the same day the Federal Reserve announced unlimited quantitative easing and corporate credit facilities.^[16]

The 23 March peak is not a coincidence. That was the exact day the Federal Reserve announced unlimited quantitative easing and emergency corporate credit facilities. Credit spreads peak when central banks intervene decisively. From that day onwards, spreads compressed back rapidly.

This is the textbook pattern: stress builds → spreads widen → central bank intervenes → spreads compress. Knowing this pattern helps you read every future episode.

2022 — The inflation/rate-hike cycle

The Fed raised rates from 0% to 4.5%+ in nine months — the most aggressive tightening cycle in 40 years. You'd expect credit spreads to have exploded. They didn't. Baa-Treasury spreads peaked at just 2.42% in July 2022. During the 2022 inflation/rate-hike cycle, the Moody's Baa10Y credit spread peaked at 2.42% on 1 July 2022 — modest by historical standards despite aggressive Fed tightening.^[17]

This is a crucial subtlety. The 2022 episode was a rate shock, not a credit shock. Companies' borrowing costs went up because all borrowing costs went up, not because the market suddenly thought defaults were coming. Credit spreads — which measure the premium over Treasuries, not the absolute level — barely moved.

This pattern is part of why the much-predicted 2023 recession didn't actually arrive.

The four regimes

Now you've seen the history, here's the framework Lighthouse uses to classify what credit spreads are saying right now.

We use four regimes, calibrated against the high-yield index since 1996:

Compressed (under 4%) — Calm or complacent. Investors aren't worried about defaults. Often coincides with strong economic conditions, low VIX, rising equities. Can also indicate late-cycle complacency before a downturn. Compressed credit spreads (under 4% on the HY OAS) historically indicate a calm or complacent credit environment with low perceived default risk.^[18]

Normal (4-6%) — Typical mid-cycle. The market is pricing some default risk but no acute stress. Most of the time falls here.

Stressed (6-9%) — Significant credit market stress. Often coincides with recession or major market scare. Central banks are usually already responding (rate cuts, jawboning). Stressed credit spreads (HY OAS between 6% and 9%) historically correspond to recessions or significant market scares without full-blown crisis.^[19]

Crisis (above 9%) — Systemic event. Central banks are intervening with emergency facilities, unlimited QE, or large rate cuts. The Global Financial Crisis and COVID-19 are the only crisis-regime episodes in our 1996-onwards window. Crisis credit spreads (HY OAS above 9%) historically correspond to systemic events where central banks are actively responding (rate cuts, emergency facilities, QE).^[20]

These thresholds are calibrated to history. They are not predictions or guarantees. But they're useful because they translate "the spread is at 7%" — which means nothing on its own — into "we're in a stressed regime, here's what that historically looks like."

Where we are right now

As of late April 2026, the Lighthouse dashboard shows:

• The Moody's Baa-Treasury spread sits at 1.7% — historically very low, well below every major stress episode in our window. As of 27 April 2026, the Moody's Baa10Y credit spread sits at 1.70% — historically low, well below every major stress episode in the 1986-onwards window, and consistent with the compressed credit spread regime.^[21]
• The ICE BofA HY OAS sits at 2.82% — also compressed, near the lower end of its three-year range. As of 28 April 2026, the HY OAS sits at 2.85%, well within the compressed regime range and near the lower end of the spread's three-year window.^[22]

Both readings put us solidly in the compressed regime. By every metric, the credit market is calm.

What does this tell us? Two interpretations are both valid:

Reading 1 — The economy is genuinely fine. Default rates are low, corporate balance sheets are healthy, the Fed has navigated to a soft landing. Compressed spreads reflect actual conditions.

Reading 2 — The market may be complacent. Compressed spreads have historically preceded the biggest stress events. Late 2007 had compressed HY spreads. February 2020 had compressed HY spreads. Compression doesn't mean nothing bad will happen — it means the market isn't pricing anything bad yet.

Both can be true simultaneously. The smoke detector isn't going off, but that's also when houses are most likely to burn down — when no one is watching.

The discipline is: when spreads are this compressed, your job isn't to predict when they'll widen. Your job is to be ready when they do.

What to watch from here

Three signals worth tracking in the coming weeks and months:

1. CCC versus BB divergence. If the riskiest tier of high-yield starts widening relative to the safer tier, that's the smoke detector starting to chirp. Lighthouse tracks this directly.

2. Investment grade widening. Investment grade spreads (BAMLC0A0CM on the dashboard) are normally far less volatile than high-yield. If they start widening too, that's a deeper signal.

3. Velocity, not level. Sometimes the speed of widening matters more than the absolute level. A jump from 2.5% to 4.5% in a month is a much louder alarm than a slow drift from 4% to 4.5% over a year.

The dashboard will track all three. Future Lighthouse articles will dig into each one.

What the smoke detector can't do

Credit spreads aren't magic. Important things they don't tell you:

• Timing. Spreads can stay compressed for years before a crisis. They can also widen and stay wide for months without recession arriving. They tell you about regime, not timing.

• Cause. Spreads measure stress. They don't tell you whether the stress is from inflation, geopolitics, banking trouble, or something else. You need other tools for diagnosis.

• Recovery. Spreads peak when central banks intervene, then compress. But the underlying economic damage (unemployment, business closures, asset price drops) often continues for months after spreads have already started healing.

• Bubbles. Compressed spreads happily coexist with massive equity bubbles. The dot-com bubble had compressed credit spreads. The 2021 everything-bubble had compressed credit spreads. Spreads are a credit signal, not a valuation signal.

Used as one tool among several, credit spreads are arguably the most valuable single market indicator that exists. Used alone, they're incomplete.

The Lighthouse takeaway

If you remember nothing else from this article, remember this:

Credit spreads are the bond market's price for taking on the risk of default. They tend to widen weeks or months before equity markets, GDP statistics, or central bank action confirm a problem. Watching them is one of the cheapest, highest-quality early warning signals available to a non-professional investor.

The Lighthouse dashboard tracks five different credit spread series, divided into tiers, with regime classification and historical context. You don't need to be a credit analyst to use it. You need to know:

• The current regime (compressed / normal / stressed / crisis)
• Whether spreads are compressing or widening (direction matters)
• Whether the riskiest tiers are diverging from the safer ones (early warning)

That's it. Three things. Check it once a week. You'll know more about market stress than 95% of people on financial Twitter.

Test your understanding