Bonds and gold, which were supposed to cushion stocks when they wobble, fell together this week. So what on earth should I hold just in case?
2026-W40
What broke this week
Textbook diversification rests on the assumption that bonds rise when stocks fall and gold rises when anxiety grows. This Monday put that assumption to a direct test. President Trump rejected Iran's proposal to reopen the strait, raising geopolitical risk, yet gold futures fell -3.54% in a day and long-term yields rose. Both hedges reacted to rates and the dollar, not to fear.
This phenomenon has a name: a regime in which the stock-bond correlation turns positive. When inflation and supply shocks drive markets, rising rates weigh on stocks and bonds together. Oxford Economics expected the term premium to keep rising on supply shocks, geopolitical risk and expansionary fiscal policy, and the correlation to turn positive again in 2026 (Oxford Economics, 2025-10-08).
"traditional approaches to asset allocation that rely on government bonds ... to provide a ballast to portfolios" will face difficulties — Oxford Economics (2025-10-08)
Things are now playing out as that forecast predicted.
Why did gold lose its hedging function?
Gold hit a record $5,602 an ounce on January 29, 2026, then fell about 30% to $3,942 in June (Morung Express, 2026-07). This week it closed at $4,162.30. Goldman Sachs' year-end target of $5,400 was premised on the Fed cutting another 50bp this year (KuCoin News, citing a Goldman report). In reality, the Fed raised rates on 9/16. The consensus has yet to fully reflect that the premise behind the target is already broken.
The state of both hedges through the signals
| Date | TLT close | TLT stop-loss buffer | GLD close | GLD stop-loss buffer |
|---|---|---|---|---|
| 09-28 | 78.62 | 3.85% | 377.91 | -8.21% |
| 09-29 | 78.23 | 3.41% | 382.89 | -7.20% |
| 09-30 | 77.78 | 2.90% | 380.84 | -7.62% |
| 10-01 | 77.71 | 2.82% | 382.76 | -7.23% |
| 10-02 | 77.48 | 3.47% | 380.14 | -7.76% |
Neither met its entry condition at any point this week. As of 10/2, the long bond ETF (TLT) at $77.48 sits below its 10-day MA of 79.00, 50-day MA of 81.15 and 150-day MA of 83.23, which are stacked in bearish order. The gold ETF (GLD) at $380.14 is below its 10-day MA of 388.10, 50-day MA of 396.29 and 150-day MA of 408.59, and at -22.76% from its $492.15 peak it has already breached the stop-loss line (-15%). Under the trend rules, there is no reason to hold either asset.
So what does work?
| Candidate | Regime where it works | This week | Weakness |
|---|---|---|---|
| Short-term Treasuries (SHV, SGOV) | Almost all regimes | 13-week yield of 3.99%, almost no price movement | Doesn't rise when stocks fall. A parking place, not insurance |
| Energy (XLE) | Oil-driven supply shocks | +1.26%, the only positive among the four asset classes | Can plunge around 5% on a single day of ceasefire agreement |
| Trend-following funds | Trends where stocks and bonds fall together | About +9.28% in the first half of this year | Losses during sharp reversals |
| Long bonds (TLT) | Recession, disinflation | -1.92% | Periods of rising term premium |
| Gold (GLD) | Falling real yields, weaker dollar | -3.37% | Fed hiking cycles |
Trend-following strategies are designed to turn downtrends into gains when stocks and bonds fall together. The WisdomTree trend-following fund (WTMF) grew $1 to about $1.07 through July 2026, while a 60% stock / 40% bond portfolio slid to about $0.96 over the same period (WisdomTree, 2026). However, its correlation with equities is 0.55, so it is not a pure opposite-direction asset.
There is one point Korean investors can easily miss. The dollar index rose +0.95% this week, but USD/KRW fell -1.82% as the won strengthened even more. Measured in won, dollar assets were also a source of loss rather than a hedge.
So what should you hold?
- Split your hedge from one into three. Park cash in short-term Treasuries, put 3-5% in energy against supply shocks, and add a small amount of trend-following if you wish. This system's signal strategy also parks funds that don't meet the entry condition in the short-term bond ETF (SHV)
- Buy long bonds and gold after the trend returns. For TLT, the first signals are reclaiming the 10-day MA of $79.00 and the 30-year settling below 5.5%; for GLD, reclaiming the 10-day MA of $388.10 and the stop-loss buffer turning positive. Both assets will likely rebound together on the day rates turn, but there is no need to rack up losses trying to call that day in advance
- The most reliable hedge is your equity weight itself. In a regime where hedging tools are broken, holding fewer stocks and collecting 3.99% on short-term bonds is more efficient than buying expensive insurance
- Won-based investors should look at the exchange rate separately. If you add new dollar assets while the won keeps strengthening, mix in currency-hedged products
You can find detailed charts and signals for other tickers on the /signals dashboard.