A junk bond crash signal is flashing. Its record is weaker than it looks

“Every time this has happened, the market has gone down at least 10%.” That is the claim of a video by The Trading Fraternity that spread on Instagram this week. The signal is the gap between CCC spreads and single-B spreads, the extra yield over Treasuries on the two lowest grades of US junk bonds. The video says that each time the gap passed 7.8 percentage points, in 2000, 2008, 2016 and 2020, stocks fell.

The first half checks out. On 23 October the CCC and lower index paid 13.56 points over Treasuries and the single-B index 3.43. The gap of 10.13 points is the widest since May 2020. It passed 7.8 on 9 September. But a look at every fall of 10% or more in the S&P 500 since 1997 shows a signal that came late three times and stayed silent through nine falls.

S&P 500 falls of 10% or more since 1997, by when the gap between CCC and single-B spreads passed 7.8 points. Source: Yardeni Research and S&P Dow Jones Indices via FRED (the 16 falls); ICE Data Indices via FRED, with archived FRED files for 1996 to 2023 (the spreads). Before: 1999, 2000–02, 2002–03, 2015–16. Late: 1998, 2007–09, 2020. Never: 1997, 2010, 2011, 2015, 2018 (twice), 2022, 2023, 2025. Our classification.

Before or early in the fall	4
Late in the fall, or after it	3
Never reached 7.8	9

Source: Yardeni Research and S&P Dow Jones Indices via FRED (the 16 falls); ICE Data Indices via FRED, with archived FRED files for 1996 to 2023 (the spreads). Before: 1999, 2000–02, 2002–03, 2015–16. Late: 1998, 2007–09, 2020. Never: 1997, 2010, 2011, 2015, 2018 (twice), 2022, 2023, 2025. Our classification.The Macro Notebook

What CCC spreads measure

Rating agencies grade junk bonds BB, B and CCC, from least to most risky. A CCC company is, in the video’s words, “one problem away” from default. When investors grow nervous they sell the weakest bonds first, so CCC spreads widen faster than the rest. The video calls this spread decompression: the bottom of the market pulls away from the grade above it.

FRED now shows only the last three years of these indices. To test the claim we used copies of FRED’s full daily files saved by the Internet Archive, for CCC bonds and single-B bonds, which start in December 1996.

The years in the video are right. The gap passed 7.8 points in four earlier periods: late 1998, and most of the time until May 2003; October 2008; November 2015; and March 2020. Each of them came with a fall of 10% or more in stocks.

A signal that often came late

The timing is the weak part. We matched the gap against the 14 falls of 10% or more from 1997 to 2022 in Yardeni Research’s tables, and the two since then in the S&P 500’s daily closes: 10.3% in 2023 and 18.9% in 2025.

In 1998 the gap first crossed on 30 November, three months after the low of a 19.3% fall. In 2008 it crossed on 16 October, more than a year after the S&P 500 peaked on 9 October 2007. In 2020 it crossed on 24 March, one trading day after the low. A reader who sold on the signal then would have sold close to the bottom.

It did give warning twice. It was above 7.8 for much of 1999, before the 49.1% fall that began in March 2000. And it crossed on 30 November 2015, four weeks into a fall of 13.3% that ran to February 2016.

Nine falls came with no signal at all, among them both falls of 2018, the 25.4% bear market of 2022 and the 18.9% drop of spring 2025. The gap peaked at 7.41 in November 2022.

Why CCC spreads are wide now

This time the gap is different in kind. When it first crossed in each earlier period, single-B bonds were under strain too: their spread was 5.61 points in November 1998, 17.10 in October 2008, 6.39 in November 2015 and 11.54 in March 2020. On 9 September this year it was 2.80. The spread on the whole high-yield index was 3.38 on 23 October, low by its own history, though up from 2.93 on 25 September, when we wrote that it was the widest since April.

Since the end of 2025, CCC spreads have widened by 4.71 points and single-B spreads by 0.42. On 9 October the CCC spread was 4.04 times the single-B spread, the highest ratio in the data since 1996.

A few borrowers drive it. Janus Henderson, a fund manager, wrote on 15 October that around 20% of the CCC index traded at $60 or below in early October, while “the performing side of the index looks quite normal relative to history”. The worst losses were in cable and satellite and in wireline telecoms, led by CSC Holdings, part of the cable company Optimum, and Altice. CCC bonds are also only 8.6% of the high-yield index, down from about 14% in 2016.

That is a fund manager’s view, and it may prove too calm. But it fits the numbers: so far, stress is confined to the weakest borrowers, while the rest of the junk market lends freely. The S&P 500 closed at 7,623.68 on 23 October, 2.5% below its record of 6 October and almost exactly where it stood on 9 September.

What to watch

The single-B spread is the one to watch. If it starts to rise, the strain will have spread from a few cable companies to the middle of the junk market, as it did in 2008 and 2020. Janus Henderson put much of the recent widening down to higher borrowing costs as central banks turned hawkish, so the Federal Reserve’s decision on 28 October, at the end of its two-day meeting, matters here too.

The second half of the video is about next week’s midterm elections and what a clean sweep for one party could mean for stocks. That is a forecast, not a fact, and we leave it aside. We wrote on 23 October about what the 2018 midterms did to stocks.