1. Bottom Line & Directional Bias
Call: Bearish palladium (PA=F), invalidation on a settle above pivot P 1181.7 and confirmation of that invalidation above R1 1200.5.
The 2026-10-02 settle of 1172.2 sits at the 3rd percentile of the 20-day 1163–1439 range — the bottom of the channel, not the middle. Three reasons drive the short. First, trend and location agree: 5D -8.13% and 20D -18.59% (settle-based) with the 52-week low at 1161.3 only ~11 points below the settle means the market is testing a level that has already failed once this year, and there is no evidence of absorption. Second, volatility structure is hostile to longs: ATR14 44.8 is 3.82% of price (full daily range), while RV20 is 32.5% annualized — realized movement is elevated and the last completed weekly bar (2026-09-28–10-02) opened at 1275 and closed at 1172.2, a -8.13% w/w range-down week, which is a supply week, not a base. Third, the 20-session seasonal window (median +4.28%, up 9 of 15 years) is a small-sample positive that the market has already had ample opportunity to discount and has not. Invalidation is a settle above 1181.7 (pivot P); a settle above 1200.5 (R1) would break the lower-high structure outright and force a neutral stance. Until then, the bias is to sell strength, not to buy the 52-week low.
2. Price Action & Technical Analysis
The prior session settle was 1172.2 (2026-10-02), -0.65% on the day, -8.13% over five sessions and -18.59% over twenty sessions. The 20-day channel runs 1163–1439, placing the settle at the 3rd percentile — effectively on the floor. The 52-week range is 1161.3–2169.9, so the settle is roughly 11 points above the 52-week low and about 998 points below the 52-week high; the market is not mid-range, it is at the bottom of a year-long distribution.
In early Asian trade on 2026-10-05 (07:00), the last print was 1177.3, +0.44% versus the settle, with an Asian session range of 1170.3–1181. This is a modest bounce, and it is worth being precise: it is an Asia-bar move, not a settled move, and it has already tagged the pivot zone. Pivot P is 1181.7, R1 1200.5, R2 1228.7; S1 1153.5, S2 1134.7. The Asian high of 1181 is essentially at P — the first supply shelf — and the Asian low of 1170.3 is above S1. The arithmetic matters: a settle above 1181.7 is the first sign the floor is holding; a settle below 1153.5 (S1) opens 1134.7 (S2) and then the 52-week low at 1161.3 is already behind price, so the next reference is the psychological 1100 handle.
ATR14 is 44.8, or 3.82% of price as a full daily range — not ±3.82%. RV20 is 32.5% annualized. The last completed weekly bar (2026-09-28 to 2026-10-02) opened at 1275, high 1275, low 1163, closed 1172.2, -8.13% w/w. That is a wide-range down week that closed near its low, and the current week has no settled bar yet, so no weekly conclusion can be drawn from it. The structure is a sequence of lower highs since the 20-day top at 1439, and the 1163 low of the completed week is the line in the sand. A close below 1163 would confirm the 52-week low at 1161.3 is next, and given ATR of 44.8, a single session can cover that distance without strain.
3. Supply-Demand Balance & Fundamental Drivers
Palladium's fundamental problem is structural and well documented: the market has moved from multi-year deficit to surplus as auto-catalyst substitution (platinum and rhodium loading, plus thrifting) has permanently reduced loadings per vehicle, while secondary supply from recycling has grown. That is the backdrop against which a -18.59% 20-day move makes sense — it is not a positioning accident, it is a repricing of a market that no longer needs to ration supply.
The macro transmission channel is rates and the dollar, and both are headwinds. The US 10-year yield (^TNX) is 5.277, up 0.76% on the last reading, and DXY is 101.86, -0.07%. A 5.28% 10-year is a meaningful real-cost headwind for a zero-carry industrial metal with no monetary role; palladium is not gold, and it does not get a safe-haven bid when real yields rise. The dollar's marginal softening is not enough to offset that. Note also that gold implied vol (^GVZ) at 23.23 sits at the 15th percentile of its 1-year range and VIX at 15.31 is also at the 15th percentile — macro vol is cheap, which historically means the market is not pricing tail risk in precious metals broadly. Palladium, however, is trading with RV20 at 32.5%, well above the gold complex's implied vol. That divergence — palladium realized vol running hot while the broader metals complex is calm — is idiosyncratic weakness, not a macro risk-premium story.
On the physical side, the relevant read is that with the 20-day channel floor at 1163 and the 52-week low at 1161.3, the market is testing whether producers and recyclers will withhold supply at these levels. Historically, South African marginal production becomes uneconomic in the 1100–1200 zone, which is a genuine medium-term support argument. But that is a slow-moving supply response, not a reason to be long into a tape making lower highs with ATR at 3.82% of price. The near-term balance is loose, the macro is hostile, and the price is at the floor. That combination favors continuation over reversal.
4. Positioning & Fund Flows
The most important positioning fact is the one embedded in the price: a -18.59% 20-day move with the settle at the 3rd percentile of the 20-day range is the signature of a market where longs have been liquidated and new shorts have been added into weakness. The last completed weekly bar (2026-09-28–10-02) opened at the high (1275) and closed near the low (1172.2), which is a week of persistent selling pressure rather than a capitulation spike — capitulation weeks typically show a long lower wick and a close well off the low, and this one does not.
On the volatility side, the relationship between implied and realized matters for how to express the view. RV20 is 32.5% annualized. The comparable precious-metal implied measures are subdued: ^GVZ at 23.23 (15th percentile) and ^VXSLV at 36.9. Palladium's realized vol is running above gold's implied vol and near silver's implied vol. For a short, that argues for expressing the view in futures or in defined-risk structures rather than paying for puts, since the realized move is already being delivered and the premium may not compensate.
Crowding: the 20-day percentile of 3.3% is a price-location measure, not a positioning percentile, and the two should not be conflated. What can be said is that the trend is one-directional and extended, which raises the risk of sharp short-covering rallies — the Asian session's +0.44% bounce to 1177.3 is a small example. That is a reason to size and stop correctly, not a reason to abandon the short. The burden of proof is on the bulls to reclaim 1181.7 and then 1200.5.
5. Cross-Asset Relative Value
Palladium's relative position within the metals complex is weak. Gold implied vol at the 15th percentile and VIX at the 15th percentile indicate a market that is broadly calm, yet palladium's RV20 of 32.5% is elevated — the metal is an outlier in its own complex, and outliers in a calm macro regime tend to be idiosyncratic, not systemic. That means the weakness is about palladium's own supply-demand and positioning, not about a broad risk-off impulse that might reverse.
The rates and dollar backdrop reinforces this. With ^TNX at 5.277 and DXY at 101.86, the macro is a mild headwind for non-yielding industrial metals. Palladium has no monetary demand to fall back on, unlike gold, so it does not benefit from the same safe-haven flows when real yields are high. Relative to gold, palladium is the higher-beta, more cyclical expression of the precious complex, and in a 5.28% 10-year environment, the cyclical expression underperforms. The relative-value conclusion is straightforward: if an investor wants precious-metals exposure, palladium is the wrong vehicle right now; if an investor wants to express a bearish industrial-metals view, palladium's high realized vol and weak trend make it an efficient short.
6. Historical & Seasonal Patterns
The seasonality block covers the same calendar start, next 20 sessions, last 15 years: mean +0.88%, median +4.28%, up 9 of 15 years; best 2011 +13.84%, worst 2022 -17.78%. This is context only and the sample is small — 15 observations is not a statistical basis for a trade. The honest read is that the seasonal tilt is mildly positive, with a median that is much stronger than the mean, indicating a right-skewed distribution driven by a few strong years. That skew is exactly what makes the seasonal signal unreliable: the median overstates the typical outcome when the distribution is driven by outliers like 2011.
More importantly, seasonality is a weak prior against a strong trend. The market has just completed a -8.13% w/w bar and is sitting at the 3rd percentile of its 20-day range. When price action and seasonality conflict, price action wins in the near term. The seasonal window is a reason to be disciplined about short-covering risk into mid-October, not a reason to be long. The 2022 analogue (-17.78%) is a reminder that this window can be violently negative when the macro regime is hostile — and with a 5.28% 10-year, the macro regime is hostile.
7. Scenario Analysis (Base / Bull / Bear)
Base case — 55%: continuation lower, 1163 floor breaks. Trigger: a settle below S1 1153.5, which would put the 20-day floor at 1163 and the 52-week low at 1161.3 behind price. Target: 1134.7 (S2) initially, with 1100 as the next psychological reference. Action: hold the short, trail stops to the 1181.7 pivot zone, and add on a failed retest of 1163–1170. This is the path most consistent with the -8.13% 5D and -18.59% 20D trend, the close-near-low weekly bar, and the hostile rates backdrop.
Bull case — 25%: floor holds, short-covering rally. Trigger: a settle above pivot P 1181.7, confirmed by a settle above R1 1200.5. Target: 1228.7 (R2), with 1275 (last completed weekly open) as the stretch. Action: cover the short on the 1181.7 settle and stand aside; only consider a long if 1200.5 is reclaimed and held for two sessions. The Asian session's +0.44% bounce to 1177.3 is the early hint of this path, but it has not cleared the pivot, so it is not yet actionable.
Bear case — 20%: acceleration through the 52-week low. Trigger: a settle below 1161.3 (52-week low) on above-average range. Target: 1134.7 (S2) quickly, then 1100. Action: this is the base case with more velocity — stay short, do not add size into a vertical move, and use ATR14 44.8 as the guide for how far a single session can travel. A move of this magnitude would likely coincide with a broader risk-off impulse, which the current VIX at 15.31 (15th percentile) is not pricing.
Probabilities sum to 100%. The base case agrees with the section 1 call: bearish, invalidation at 1181.7.
8. Trading Strategies & Risk Management
Strategy 1 — Short PA=F on rallies into the pivot zone. Entry: 1178–1182 (the Asian high of 1181 and pivot P 1181.7 define the zone). Stop: 1201 (just beyond R1 1200.5, roughly half an ATR above entry — the level, not the noise, defines the stop). Target: 1135 (S2 1134.7). Horizon: 1–5 sessions. Size: half normal, given the 3.82% ATR and the risk of short-covering squeezes from a 3rd-percentile location. Conviction: 7/10.
Strategy 2 — Momentum short on a settle below S1. Entry: on a settle below 1153.5 (S1). Stop: 1182 (back above pivot P). Target: 1100. Horizon: 3–10 sessions. Size: full normal, because a break of the 20-day floor and the 52-week low at 1161.3 confirms the trend. Conviction: 8/10.
Risk management: both strategies are in the direction of the call. Do not add to shorts into a vertical move; do not average up. If the market settles above 1181.7, Strategy 1 is invalidated and Strategy 2 is off the table. The 20-session seasonal window (median +4.28%) is the main risk to the short, so keep position size disciplined and respect the 1200.5 invalidation.
9. This Week's Data Calendar
BJT 10-05 22:00 | ET 10-05 10:00 — ISM Services PMI (SEP), forecast 54, previous 55.4, surprise if outside 54±1.4; affects GC, SI, DXY. BJT 10-08 02:00 | ET 10-07 14:00 — FOMC Meeting Minutes; affects GC, SI, DXY. Both are macro events that transmit to palladium through the dollar and rates channel; the FOMC minutes are the higher-impact print for the complex.
This report is generated automatically from public quantitative and macro data for research and market tracking only. It does not constitute investment advice or a recommendation to trade.