GLD/SLV · > +7% above its 65-day MA · forward 21 sessions · recency window from 2021-01-01 · half-life 9mo · de-clustered ≥42d
Now: +8.6% vs the 65-day MA · 95th pctile (2021+) · 95th all-time
Close 7.10 · MA 6.53 · IN BREACH now · all-time range -33.0% … +37.3%
Forward returns by regime
Median forward return (% positive) at 2 / 5 / 8 / 11 / 15 / 21 sessions, conditioned on each market regime — and the recency-weighted read (the headline). Watch how the shape shifts as the window tightens toward now.
Full historyPost-CovidAI eraRallyRecency-wtdbaseline
Horizon
Full historyn=27
Post-Covidn=8
AI eran=4
Rallyn=4
Recency-wtdn=8
2 sessions
+0.3% 59%
+1.5% 88%
+2.5% 100%
+2.5% 100%
+1.7% 99%
5 sessions
+0.8% 67%
+1.2% 88%
+2.5% 100%
+2.5% 100%
+1.8% 99%
8 sessions
+0.9% 63%
+2.1% 88%
+2.1% 100%
+2.1% 100%
+2.0% 99%
11 sessions
+0.9% 63%
+1.3% 88%
+2.2% 100%
+2.2% 100%
+3.2% 99%
15 sessions
-0.1% 44%
+0.7% 75%
+1.4% 100%
+1.4% 100%
+2.3% 98%
21 sessions
+0.0% 52%
-0.4% 50%
+3.1% 75%
+3.1% 75%
+5.0% 63%
Regime delta: the recency-weighted 21-day median is +5.0 pts vs the full-history read — the recent regime is markedly stronger here.
Headline is recency-weighted (faster 9-month half-life) over 8 post-2021 events. Regime windows: AI era from Apr 1 2023 · Rally from the Oct 30 2023 bottom. Tight regimes carry few events (see the n per column) — read them with the recency-weighted line, which keeps sample by fading rather than cutting.
Forward 21d · median
+5.0%
mean +1.7%
% positive
63%
8 events
vs baseline
+4.5 pts
base +0.5% / 54% pos
Range
-4% … +5%
25th -3.0% · 75th +2.2%
Path · max-gain
+4.6%
median MFE in window
Path · max-draw
-2.1%
median MAE in window
Every event · outcome over time
Each dot is one de-clustered event at its date; height is the forward 21-session return. Gold-ringed = post-2021 — if there are few gold rings, the signal is a pre-Covid-regime artifact.
By decade: 2000s 32010s 142020s 10
The event log
Event date
Stretch
Fwd 21d
Max gain
Max draw
2007-08-16
+8.1%
+1.7%
+3.0%
+0.0%
2008-08-11
+8.0%
+24.3%
+24.3%
-2.1%
2008-10-09
+22.3%
-3.0%
+12.9%
-5.2%
2010-02-04
+9.2%
-6.3%
+1.6%
-6.3%
2010-06-04
+8.1%
-4.2%
+0.0%
-6.3%
2011-05-05
+8.1%
-1.0%
+4.0%
-7.4%
2011-07-11
+8.2%
+6.9%
+6.9%
-8.9%
2011-09-22
+12.1%
+8.1%
+12.9%
+0.0%
2011-11-21
+7.5%
+3.4%
+4.2%
-2.5%
2012-05-15
+7.4%
+1.7%
+2.8%
-1.1%
2013-04-15
+8.5%
+2.7%
+3.8%
-1.5%
2013-06-20
+7.9%
+0.0%
+1.7%
-3.6%
2014-09-25
+7.4%
+2.8%
+3.9%
-0.7%
2014-11-28
+7.1%
-2.2%
+0.8%
-4.7%
2016-02-26
+7.4%
-2.6%
+0.0%
-5.5%
2017-05-03
+7.5%
-3.1%
+0.2%
-3.6%
2018-09-06
+7.0%
-3.0%
+0.1%
-3.7%
2020-02-28
+8.4%
+21.6%
+31.4%
+0.0%
2020-04-29
+9.5%
-13.7%
+2.8%
-13.7%
2021-08-10
+7.2%
+0.8%
+4.4%
-0.4%
2022-05-09
+8.5%
-1.6%
+3.1%
-2.4%
2022-07-11
+8.2%
-3.8%
+2.8%
-4.6%
2023-02-24
+8.4%
-2.8%
+4.7%
-2.8%
2024-08-01
+6.9%
+1.2%
+4.5%
-1.8%
2025-04-03
+8.7%
+5.0%
+6.9%
+0.0%
2026-03-17
+7.1%
-3.6%
+4.8%
-4.3%
2026-06-23
+8.4%
+5.4%
+7.2%
+0.0%
Read it honestly. The recency-weighted read leans on 8 post-2021 events (half-life 9mo); the full-history column carries 27. The regime delta shows how far recent behaviour has drifted from the long base rate. De-clustered by a 42-session gap; overlapping windows remain. A context gauge, not a mechanical signal.
Computed live from adjusted-close history (yfinance). Forward return = close-to-close over 21 sessions from each event. MFE/MAE = max favorable / adverse excursion within the window. Educational research, not investment advice.