Research · Credit & Liquidity
Credit Before Price
The short version: stock prices are what most people watch, but the bond market is where lenders decide whether companies can keep borrowing cheaply. When lenders get nervous, they demand more yield over Treasuries. That extra yield, the credit spread, has usually widened before stocks broke. That is why the 65-signal framework anchors its macro phase on one number: the investment-grade (IG) credit spread.
What the homepage is telling you right now
When this was written (October 19, 2026), the 65 Signals macro dashboard read Phase 1: Expansion. The engine's IG spread reading was 84 basis points, comfortably under the 150 bps line that defines Phase 1. The dashboard updates every day, so check it for the current reading; this post explains what the number means, not where it will go.
The framework's rule (the "Master Rule") is deliberately simple:
- Below 150 bps: Phase 1, expansion. Credit is cheap and plentiful.
- 150 to 300 bps: Phase 2, cracks. Lenders are charging more for risk.
- 300 bps and above: Phase 3, forced liquidation. The engine only leaves Phase 3 once the spread falls back below 250 bps.
The Master Rule
84 bpsOct 19, 2026 reading (check the dashboard for today)
The series behind it is the ICE BofA US Corporate Index option-adjusted spread (FRED: BAMLC0A0CM). You can read about it on the IG Credit Spreads (OAS) and IG Credit Spread Cycle signal pages.
Why credit tends to move first
A stock is a claim on whatever is left after a company pays its lenders. A bond is a claim that gets paid first. So the people with the most to lose from a slowdown, and the strongest reason to watch for one, are the lenders. They re-price risk continuously, across thousands of bonds, without waiting for quarterly earnings or monthly economic data.
When borrowing gets more expensive, companies feel it before shareholders do: refinancing costs rise, buybacks and expansion plans get trimmed, and weaker borrowers start to struggle. Equity markets often keep climbing on momentum while that happens underneath. The spread is the earlier, quieter signal.
"Usually" matters in that sentence. The record below shows one cycle where credit clearly led and one where it did not.
The record: daily, since December 1996
The framework's credit record is the daily IG spread from December 31, 1996 forward: 7,688 daily observations in the archive the engine reads. Across that record the spread sat below 150 bps on about 65% of days. Expansion is the normal state; the other phases are the exceptions that matter.
The record in three numbers
The phase readings for 2007–2009 and 2020 below come from applying today's rule to that archived daily series. They are a historical reconstruction, not calls the engine made in real time.
Two episodes, one rule
2007–2009
- IG crosses 150 bps: Phase 2
- S&P 500 monthly average peaks at 1,539.66
- 303 bps: first Phase 3 reading
- Back to Phase 2
- Phase 3 again
- Peak: 656 bps
- Below 250 bps: Phase 3 ends
2020
- S&P 500 record close 3,386.15; IG at 102 bps (Phase 1)
- 188 bps: Phase 2
- 303 bps: Phase 3
- Peak: 401 bps; S&P 500 closes at 2,237.40
- Below 250 bps: Phase 3 ends
- Back to Phase 1
2007–2009: credit cracked first
- September 6, 2007: the IG spread crossed 150 bps and the rule moved to Phase 2. It slipped back under the line for a few weeks, then held above it from October 31, 2007.
- October 2007: the S&P 500 composite's monthly average peaked at 1,539.66 (Shiller monthly data). Credit had crossed its line the month before stocks topped.
- March 17, 2008: the spread reached 303 bps, the framework's first Phase 3 (forced liquidation) reading of the cycle. It eased back into Phase 2 on May 16, then returned to Phase 3 on July 31, 2008.
- December 5, 2008: the spread peaked at 656 bps, the widest in the record.
- September 10, 2009: the spread fell back below 250 bps and the rule left Phase 3.
March 2020: a shock, not a slow build
- February 19, 2020: the S&P 500 closed at a record 3,386.15. The IG spread that day was 102 bps, deep in Phase 1. This time credit gave no early warning.
- March 9, 2020: the spread jumped to 188 bps (Phase 2).
- March 18, 2020: 303 bps, a Phase 3 (forced liquidation) reading, seven trading days after Phase 2 began.
- March 23, 2020: the spread peaked at 401 bps. The same day the S&P 500 closed at 2,237.40, its lowest close of that episode.
- April 13, 2020: back below 250 bps and out of Phase 3. The rule returned to Phase 1 on July 15, 2020.
The two episodes teach different lessons. In 2007 the spread widened while stocks were still near their highs; that is the classic "credit before price" pattern. In 2020 the shock hit everything at once, and the spread's value was in describing how severe the stress was and when it was easing, not in calling the top.
How to use this, and how not to
- It is a regime gauge, not a timing tool. The spread tells you which kind of market you are in. It does not tell you what any single asset will do next week.
- Thresholds are bright lines on a noisy series. Readings near 150 or 300 bps can flicker across the line, as in September 2007. Read the trend, not one print.
- Policy can change the path. In 2020 the spread was back under 250 bps three weeks after its peak; after the December 2008 peak it took about nine months.
- Check it next to the other signals. The macro dashboard shows the phase alongside volatility, the yield curve and the rest of the framework.
Sources
- IG spread: ICE BofA US Corporate Index Option-Adjusted Spread (FRED series BAMLC0A0CM), daily, December 31, 1996 to June 10, 2026, from the engine's archive (an archived FRED capture joined to the live FRED window; the overlap matches exactly). Current reading: the 65 Signals engine, October 19, 2026.
- Phase readings: the framework's Master Rule (below 150 / 150–300 / 300 and above, exiting Phase 3 below 250 bps), recomputed over that series.
- S&P 500: daily closes (2020); Shiller monthly S&P composite averages (2007).
This article is for educational and informational purposes only. It is not investment advice, and nothing in it is a recommendation to take any position. Historical patterns describe what happened; past performance doesn't guarantee future results.