Gold — weekly cycle & the long-term cycle
Chart & cycle-analysis proxy: SPDR Gold Shares (GLD) · $382.27 per ETF share · as of 2026-10-06
GLD is an ETF proxy. Price is USD per share, not spot metal per ounce. Counts are automated swing-based estimates; yearly structure is not confirmation of an eight-year cycle.
Gold’s major top may be in.
Our base case is that gold has entered the declining leg of the cycle anchored at the 2022 low. In this view, weekly rallies are countertrend: they should peak earlier, fade faster and increasingly leave left-translated cycles behind.
The evidence
- GLD weekly structure: declining; left translation so far.
- 2 of the last 3 completed weekly swing cycles were left-translated.
- GLD monthly structure: declining.
- Gold futures are 24.9% below their highest high since the 2022 anchor.
The countercase
The gold futures weekly cycle is even-translated, while GLD is left-translated. The ETF and futures counts differ; that prevents an unqualified top call. The 2022 low is the working eight-year anchor, not a calendar guarantee.
What changes our view
A sustained recovery above the post-2022 peak would invalidate the current top candidate. Before that, repeated right-translated weekly cycles and higher lows would weaken the bearish case.
2022-11-03Post-anchor high
$5586.20 · 2026-01-29Below that high
24.9%
Anchor and peak use gold futures (GC=F), USD per troy ounce; market date 2026-10-06. GLD charts below use USD per ETF share.
See the recent rally lengths
A left-translated cycle spends fewer bars advancing than declining. Open cycles are provisional.
| Weekly low → low | Rally | Decline | Translation |
|---|---|---|---|
| 2026-03-23 → 2026-06-08 | 3w | 8w | Left |
| 2026-06-08 → 2026-07-13 | 1w | 4w | Left |
| 2026-07-13 → 2026-07-27 | 1w | 1w | Even |
How this view updates
The bearish working view requires a drawdown of at least 15% from the post-2022 futures high, declining weekly GLD and monthly futures structure, and left translation in at least two of the last three completed GLD weekly swing cycles. These are editorial rules, not a backtested prediction model.
These are consecutive confirmed swing cycles, not analyst-selected 18–30-week intermediate cycles. Short swings can be nested inside a larger weekly cycle. A phase call is an interpretation, not a probability. Read the framework.
The structure underneath the view
Gold · cycle chart
Gold — frustrating chop, but the bigger picture is still coiling for a breakout into the 8-year cycle top. Watching support for a dip-and-reclaim as the actionable tell.
Engine read — auto-derived from the cycle data
Gold is in the declining phase of the weekly swing cycle — 10 weeks into a cycle that has historically run a median of about 5 weeks. The last ICL printed 2026-07-27, and the cycle is left-translated so far. On the separate yearly-bar view, the automated read is advancing, 11 years in, right-translated. This yearly-bar count is separate from the eight-year thesis above.
Automatically derived from confirmed swing lows. Published analysis, where available, is dated separately.
Yearly swing history
| Low → Low | Len | Transl. |
|---|---|---|
| 2005-01 → 2015-01 | 10y0m | Right |
| 2015-01 → now | open | Right |