This week bonds and gold both fell too. What am I supposed to lean on when stocks wobble? Should I just follow the signal to sell long-term Treasuries and gold?
2026-W39
Both hedges fell at the same time this week
Long-term Treasuries TLT -2.38%, gold GLD -1.93%. The two assets that should protect you when stocks wobble both fell in a week when stocks rose. It's no coincidence. Stocks and bonds moving in opposite directions is not an inherent property of the assets but a feature of particular regimes. When inflation was low, economic worries pulled yields down and bonds offset stock declines. In a regime where inflation is the problem, rising yields hit stocks and bonds at the same time. The correlation between U.S. stocks and bonds has stayed positive since 2022, and that year a 60/40 portfolio returned -17.5%, the worst since 1937 (Medium Technicity, 2026-02).
Gold got caught in the same logic. Spot gold gave up its $4,300 support at $4,274.79 an ounce on Sept. 24, 23.5% below its January peak of $5,589.38 (Discovery Alert, 2026-09-24). As the dollar hit its strongest level in two months and the Fed signaled more hikes, non-yielding gold lost out to interest-paying Treasuries. Despite structural buying, with central banks purchasing 863 tonnes in 2025 alone (GoldSilver, 2026-09-14), the pull of rates was stronger this week.
What the signals say
Here is the week for long-term Treasuries (TLT, safe-haven group, stop-loss at -15% from the high).
| Date | Close | 10-day MA | 50-day MA | 150-day MA | From high | Stop-loss cushion | Entry condition |
|---|---|---|---|---|---|---|---|
| 09-21 | 81.80 | 81.29 | 82.23 | 84.07 | -7.56% | 7.44% | Not met |
| 09-22 | 81.75 | 81.25 | 82.20 | 84.03 | -7.61% | 7.39% | Not met |
| 09-23 | 80.46 | 81.12 | 82.14 | 83.99 | -9.07% | 5.93% | Not met |
| 09-24 | 79.42 | 80.98 | 82.06 | 83.94 | -10.25% | 4.75% | Not met |
| 09-25 | 79.32 | 80.83 | 81.97 | 83.88 | -10.36% | 4.64% | Not met |
The close is below the 10-, 50- and 150-day lines, and the entry condition never turned on. The stop-loss cushion fell below 5% on Thursday, leaving just 4.64% to the stop-loss line.
Here is gold (GLD, same group).
| Date | Close | 10-day MA | 50-day MA | 150-day MA | From high | Stop-loss cushion | Entry condition |
|---|---|---|---|---|---|---|---|
| 09-21 | 398.38 | 397.48 | 393.40 | 413.54 | -19.05% | -4.05% | Not met |
| 09-22 | 400.07 | 397.52 | 394.06 | 413.22 | -18.71% | -3.71% | Not met |
| 09-23 | 392.88 | 396.47 | 394.48 | 412.78 | -20.17% | -5.17% | Not met |
| 09-24 | 391.69 | 396.00 | 394.86 | 412.33 | -20.41% | -5.41% | Not met |
| 09-25 | 393.41 | 395.47 | 395.43 | 411.83 | -20.06% | -5.06% | Not met |
Gold is already -20% from its high, past the -15% stop-loss threshold. A negative stop-loss cushion means that under the rules it should already have been exited. The 10-day and 50-day lines converging around $395 suggest a short-term bottom forming, but the 150-day line ($411.83) is still 4.7% above.
Post-mortem on the verdicts
This week's two Exit verdicts are consistent with the rules. The part worth revisiting is the prior week. The W38 report kept holding gold as a hedge even though its stop-loss cushion was negative (-3.49%) from the start of the week, and that decision cost -1.93% this week. A discretionary call that departed from the rules came precisely in a regime where hedges don't work.
The lesson is simple. Hedge assets are no exception to trend rules. The expectation that they will rise when stocks fall doesn't hold in a regime where rates are rising on economic strength.
So what do you lean on?
| Alternative | Risk it protects against | Weakness |
|---|---|---|
| Short-term Treasuries (SHV-type) | Both rising rates and falling stocks. 3-month yield 4.07% | Doesn't rise when stocks plunge. It only prevents losses |
| Healthcare (XLV) | Rate shocks. +1.76% in a week when rate-sensitive sectors fell across the board | As equities, it takes the hit in a broad market plunge. Eli Lilly is a large weighting at about 15% |
| Trend following (managed futures) | Stocks and bonds falling together. SG Trend Index +27.3% in 2022 | Can lag for long stretches in directionless markets. In 2025 there was a stretch with 12-month returns of -18.6% |
(XLV holdings from StockAnalysis, 2022 trend index from Return Stacked and CNBC compiled, 2025 performance from Man Group data)
The third alternative shone when an inflation shock ran in one direction for a long time, as in 2022. That year, managed futures ETFs returned 14-48% while stocks and 60/40 portfolios posted double-digit losses (eToro, 2026-03-16). But in markets where rates go up and down repeatedly, signals flip often and only costs pile up. It favors weeks like this one, when the direction of rates is clear, and suffers when a shutdown or data vacuum blurs the direction.
What investors should do
- TLT: Exit as signaled. Re-enter only when two conditions occur together: the close returns above the 50-day line ($81.97) and the 10-year falls below 5.0%. As long as rates are rising because of economic strength, long-term Treasuries are a source of losses, not a hedge.
- GLD: Don't re-enter until it returns above the 150-day line ($411.83). The long-term rationale of central bank buying is intact, but it didn't prevent this week's decline.
- Reallocate the hedge budget: Split the share held in long-term Treasuries and gold between short-term Treasuries and XLV. XLV newly met its entry condition on Sept. 25 with a wide margin: a stop-loss cushion of 17.03% and -2.97% from its high. Since it's the first day of meeting the condition, enter in tranches.
- In this regime it makes more sense to redefine a hedge not as an asset that rises when stocks fall but as one that loses less when stocks fall. By that standard, the most reliable asset is short-term Treasuries, which pay 4% without principal swings.
You can find detailed charts and signals for other symbols on the /signals dashboard.