12 Past Yield Jumps: Stocks Were Higher A Year Later In 11
Today’s big story: Could Anthropic Stock Fall 25%?
The 10-year Treasury yield is the rate the U.S. government pays to borrow for 10 years. It closed at 5.07% on Nov. 18, 2026. A year earlier, it was 4.12%. That is a rise of 0.95 percentage points in 12 months. On Oct. 5 the yield closed at 5.31%, its highest close since 2002. Since 1990, the 10-year has risen 0.9 percentage points or more in a year 13 times, counting episodes at least a year apart. That count includes today’s episode, so 12 past cases have a full year of results.
A rise this size resets what long money costs. It lowers what future profits are worth today, and it pushes up mortgage rates. So the question matters to anyone who owns stocks, long bond funds or homebuilders.
What happened after past jumps? The S&P 500 was higher a year later in 11 of 12 past cases. Long Treasury funds were higher a year later in only 6 of 9.
Even those good odds came with wide swings along the way. The Trefis Income Solution pays a dividend yield about 3x the S&P 500’s, and it lost less than the S&P 500 in all 9 big selloffs since 2018. See the metrics.
What Stocks and Bonds Did After Every Past Jump
The table shows each episode since 1990. SPY tracks the S&P 500. TLT holds Treasury bonds that mature in 20 years or more. Returns are total returns, with dividends included. Bond fund returns are not available for the three 1990s episodes.
| Episode start | 10-year yield | SPY 6 months | SPY 12 months | TLT 6 months | TLT 12 months |
|---|---|---|---|---|---|
| Apr. 1994 | 7.16% | +4.7% | +18.2% | n/a | n/a |
| Jul. 1996 | 7.05% | +15.5% | +43.5% | n/a | n/a |
| Aug. 1999 | 5.98% | +4.6% | +16.5% | n/a | n/a |
| May 2004 | 4.79% | +7.5% | +8.6% | +10.6% | +17.3% |
| May 2006 | 5.14% | +5.9% | +16.1% | +10.7% | +11.4% |
| Dec. 2009 | 3.48% | -2.9% | +12.9% | +7.0% | +5.9% |
| Jun. 2013 | 2.60% | +16.6% | +25.8% | -3.2% | +8.0% |
| Jul. 2017 | 2.33% | +13.8% | +14.6% | +2.3% | +0.8% |
| Nov. 2018 | 3.22% | +7.5% | +13.8% | +11.8% | +25.9% |
| Mar. 2021 | 1.55% | +16.7% | +11.9% | +8.0% | -0.8% |
| Apr. 2022 | 2.61% | -15.7% | -6.8% | -19.0% | -12.6% |
| May 2023 | 3.72% | +9.4% | +28.3% | -8.2% | -5.5% |
| Sep. 2026 | 4.95% | too recent | too recent | too recent | too recent |
A year after the trigger, the S&P 500 fund’s median gain was 15.4%. Its only loss was 6.8%, in the year from April 2022. The long bond fund’s median was a smaller 5.9%. Its worst was a 12.6% loss in that same year, and its best was a 25.9% gain from November 2018.
Homebuilders were higher a year later in every episode since 2009, though the sample is smaller. The ITB homebuilder fund rose in all 7 episodes from Dec. 2009 on, with a median gain of 14.4%. That held even though higher yields push up mortgage rates. The data does not show why, and 7 cases is a small sample. The one earlier episode ITB covers, May 2006, came at the start of the housing bust, and homebuilders fell. Today’s episode, which began on Sep. 10, 2026, is too recent to count as a success or a failure.
The Closest Match Is 2006
May 2006 is the nearest episode to today’s yield of 5.07%. The 10-year stood at 5.14% when that episode began. Measured at the trigger, today’s 4.95% sits between May 2004 at 4.79% and May 2006 at 5.14%. In 2006, stocks wobbled first. SPY was down 2.0% after three months. It then finished the year up 16.1%. Long bonds did well from the start. TLT gained 3.9% in three months and 11.4% over the year.
Since its Oct. 5 peak, the 10-year has eased about a quarter of a percentage point. Over the same stretch, long Treasuries (TLT) returned 4.5%, with dividends, through Nov. 18. That is measured from the peak, not from the Sep. 10 trigger, so it is not yet a test of the pattern.
The counter-case is 2022. That rise began in April 2022 at 2.61%, and it did not stop. The next trigger came in May 2023, at 3.72%. That is another 1.11 percentage points on top of the first rise. Long bonds lost money in the year after both the 2022 and 2023 triggers. Those are two of the three cases where TLT fell over 12 months.
What Is Different This Time
First, the Fed’s policy rate ended this rise where it started. The effective federal funds rate was 3.88% on Nov. 17, 2025, and 3.88% on Nov. 17, 2026. So the jump in long yields did not come from a higher Fed rate, and a Fed cut may not reverse it. That cuts against a quick bond rally.
Second, bond supply is heavy. The federal deficit totaled $2.11 trillion in the 12 months to October. The U.S. is also projected to hit its $41.1 trillion debt limit in 2027, and a fight in Congress over raising it could unsettle bonds. The 10-year term premium was 0.99 percentage points as of Nov. 13. That is the extra yield investors demand to hold a long bond instead of rolling short ones. All three cut against bonds.
Third, the starting yield is high. Even after expected inflation, the 10-year pays 2.89% (as of Nov. 17). A bond bought near 5% pays far more income than one bought at 1.55% in 2021 or 2.33% in 2017. That income cushions price losses. This cuts in favor of bonds.
The Odds, Plainly
In 11 of 12 cases, the S&P 500 was higher a year after the 10-year rose 0.9 percentage points or more in a year. The median gain was 15.4%, and the worst was a 6.8% loss. For long Treasuries, 6 of 9 cases were higher, the median was 5.9%, and the worst was a 12.6% loss. These are small samples. This is a base rate, not a forecast.
History becomes a poor guide if yields keep climbing for another year, as they did after April 2022. Heavy deficits and a debt-ceiling fight are the clearest ways that could happen this time. Watch the term premium, the debt-ceiling talks and whether the 10-year retakes its Oct. 5 high of 5.31%.
No one could tell at the start which past rises would fade and which would keep going. A portfolio does not have to make that call if it is built to hold up either way.
Stop Guessing. Start Positioning.
No one can predict surprises like this one. The Trefis High Quality (HQ) Portfolio doesn’t try: it holds 30 quality stocks in equal weight, so no single surprise decides your result. Come see how it works.
Spot an error or want more like this? Tell us in 10 seconds.