This article is an asset map: panorama and core concepts only. For a deep dive into the futures mechanics, see Chapter 3 · Futures.
"Is the commodity bull here?" — the 2020-2022 energy crisis and the 2024 record highs in cocoa/copper pulled commodities back into public view. But commodity moves are not random noise: they are the stack of three rhythms — decadal "supercycles", 3-5 year "macro cycles", and multi-month "inventory cycles".
This article covers how commodity indices are built (CRB / S&P GSCI / Nanhua), the lead-lag relationship between commodity prices and inflation, the anatomy of two supercycles, the macro linkages of commodities with the dollar/PMI/geopolitics, and the right way — and the common mistakes — for ordinary investors to participate.
⚠️ Risk Warning
This article is for learning and research only and does not constitute investment advice. The index compositions, weights, levels, and cycle dating mentioned here are generic teaching-basis descriptions — always defer to each index provider's latest methodology and each exchange's latest data. Commodity futures carry built-in leverage and roll costs, commodity ETFs suffer tracking decay, and historical cycle patterns guarantee nothing about the future. Assess your risk tolerance before participating.
① The Three Major Commodity Indices: CRB / S&P GSCI / Nanhua
A "commodity index" is the thermometer for the overall commodity market, but different indices are built differently, and each thermometer reads a different "feels-like temperature":
| Index | Publisher | Constituents and weighting | Traits and limits |
|---|---|---|---|
| CRB Index (now TR/CC CRB) | Thomson Reuters/CoreCommodity (lineage back to 1936) | About 19 commodities, arithmetic average, roughly equal weights (oil heavyweight, the rest fairly even) | The longest history, most quoted by media; the equal-weight design amplifies small commodities (cocoa, coffee, etc.), detached from real commodity trade volumes |
| S&P GSCI | S&P Dow Jones | 24 commodities, weighted by global production (bigger output, bigger weight) | Energy weight is huge (historically 50%-70%, per latest methodology) — the index is essentially steered by oil; closer to the "real global supply-demand structure of commodities" |
| Nanhua Commodity Index | Nanhua Futures (domestic) | Main domestic futures market constituents, weighted by liquidity/trading volume (per latest methodology) | Tracks the overall performance of China-priced commodities (ferrous, chemicals, agricultural products) — the "domestic thermometer" for onshore commodity longs and shorts |
- One-line memory hook: CRB is "egalitarian", GSCI is "energy-ist", Nanhua is "China-ist" — use GSCI (or CRB) for the global commodity big picture, Nanhua for domestic ferrous/chemicals.
- The two sources of index returns: price moves + roll gains/losses (the actual return of a futures-based index does not track spot price changes; see ⑥⑦).
Main constituents at a glance (per latest methodology)
| Index | Typical constituents | Weighting traits |
|---|---|---|
| CRB (TR/CC) | About 19 commodities: crude oil, natural gas, corn, soybeans, wheat, cotton, sugar, cocoa, coffee, copper, aluminum, zinc, nickel, etc. | Equal-weight design: agricultural/soft commodities carry relatively more weight |
| S&P GSCI | 24 commodities: energy (crude/refined products/gas), industrial metals (copper/aluminum/nickel/zinc), precious metals (gold/silver), agriculture, livestock | Production-weighted: energy utterly dominant |
| Nanhua Commodity | Ferrous (rebar/iron ore/coke/coking coal), chemicals (PTA/methanol/rubber), non-ferrous (copper/aluminum/zinc), oils and oilseeds, energy-chemicals and other domestic majors | Volume/liquidity weighted: ferrous weights stand out |
- A lesson in picking indices: "is copper at a new high" can be read from GSCI or CRB, but "did oil drag commodities down" only GSCI can show — confirm what's inside the index before reading its moves.
② Commodity Indices and Inflation: A Leading Indicator
The transmission logic
text
Commodity prices rise (upstream raw materials)
↓ (weeks to months)
PPI rises (ex-factory prices of industrial goods)
↓ (months to quarters)
CPI rises (consumer prices) → central bank hikes rates → pressures all assets
| Link | Content |
|---|---|
| Why commodities lead | Commodity prices are set instantly by spot supply and demand, with no "expectations pricing" step — supply shocks (war, production cuts, weather) show up in prices immediately, while CPI waits for costs to pass through layer by layer |
| Typical lags | Crude → transport/logistics prices in weeks; metals → industrial goods in quarters; grain → food prices in about 1-3 months (per latest historical data) |
| Policy implication | Much of the "inflation" central banks watch is first decided in commodity prices — the commodity index is the forward sentry of rate policy (see 08 - Macro Economy and Markets) |
- Practical reminder: a commodity index rising for several months is a high-probability preview of "inflation climbing"; the reverse holds too. This leads the after-the-fact CPI prints by 1-2 quarters.
Two "false correlations" to sort out
| Phenomenon | The truth |
|---|---|
| "Oil is up but CPI isn't" | The lag hasn't arrived, or demand is too weak for downstream to pass it on — watch whether PPI and core CPI follow |
| "Commodities are down but prices still rising" | Services inflation (wages, rents) is decoupled from commodities — commodities are only part of CPI, and the share varies by country |
- Conclusion: the commodity index is a sufficient leading indicator of inflation but not a sufficient condition — read the "commodity index" alongside "PPI month-on-month" and "core CPI month-on-month"; three data points cross-confirming beats any single one.
③ Commodity Supercycles
A "supercycle" is a commodity bull market lasting a decade or more, driven by structural demand — fundamentally different from the ordinary 3-5 year commodity cycle.
Two widely discussed supercycles
| Cycle | Period | Driver | Outcome |
|---|---|---|---|
| China-demand supercycle | About 2000-2014 | Industrialization + urbanization after China joined the WTO: infrastructure, property, and heavy industry exploded demand for copper, iron ore, crude oil, coal | Price centers of copper, iron ore, and crude oil shifted up systematically; iron ore rose by multiples (per latest historical data); ended after 2014 as China's demand growth downshifted |
| Pandemic supply-shock cycle | About 2020-2022 | Supply contraction (lockdowns, logistics disruptions, the Russia-Ukraine war) stacked on unprecedented monetary and fiscal stimulus of demand | Copper, energy, and agriculture hit historic/multi-year highs (2022 crude, 2024 cocoa/copper, per latest data); a shorter cycle — more "supply shock + liquidity" than structural demand |
How to recognize a supercycle
| Trait | Description |
|---|---|
| Structural demand expansion | Not cyclical fluctuation but a "step-up in demand" from population, urbanization, energy transition, industrial upgrading — e.g., the energy transition's demand for copper/lithium/aluminum |
| Persistent supply bottlenecks | Long-underinvestment in mine capex, 5-10 year lead times for new mines — demand running ahead of supply is the hardest evidence of a supercycle |
| Measured in decades | A true supercycle far outlasts the ordinary inventory cycle (1-3 years) |
| Self-reinforcing narrative | "Resource-country" policies and (geopolitical) weaponization of resources amplify the cycle's slope |
💡 A supply-demand gap is the hardest evidence of a supercycle
Long-underinvestment in mine capex and 5-10 year lead times for new mines — demand running ahead of supply is the hardest evidence of a supercycle. To judge whether the current energy-transition supercycle is real, don't watch how hot the narrative is; watch whether demand growth persists and whether supply can catch up.
- The new narrative being discussed today: an energy-transition supercycle (copper, lithium, nickel as beneficiaries) + geopolitical realignment (onshoring critical-mineral supply chains) — whether it qualifies as a supercycle depends on demand growth persisting and supply catching up in time (per latest data and professional-institution research).
- Beware: the word "supercycle" itself is often used by sell-siders to sell — by the time the narrative is everywhere, the move is usually more than half done.
⚠️ By the time the supercycle narrative appears, the move is half over
The word "supercycle" itself is often used by sell-siders to sell — by the time the narrative is everywhere, the move is usually more than half done. Whether the currently discussed energy-transition supercycle holds depends on demand growth persisting and supply catching up in time. Don't get brainwashed by sell-side narrative into going heavy.
④ Commodities and the Macro
| Macro variable | Relation to commodities | Mechanism and notes |
|---|---|---|
| US dollar index | Broadly negative correlation (not absolute, per latest data) | Commodities are mostly dollar-priced: dollar up → commodities cost more in other currencies → demand suppressed; a strong dollar also usually comes with tightening expectations → pressures commodities |
| Global manufacturing PMI | Positively correlated with industrial metals | Copper/aluminum/zinc are the "coincident thermometer" of manufacturing: PMI expansion → demand up → metals stronger ("Dr. Copper" in 06 - Commodity Panorama) |
| Geopolitics | Pulsed shocks for energy/grain | War, sanctions, export bans cut supply directly: the Russia-Ukraine war on oil/gas and wheat, the Red Sea crisis on shipping, Middle East tensions on oil — all recent cases (per latest data); the signature is "fast up, slow back" |
| Real interest rates | Mostly negatively correlated with gold | Real rates up → higher opportunity cost of holding gold → gold pressured (see 09 - Precious Metals and Energy Spot) |
- A simplified mnemonic for the linkages: strong dollar + contracting PMI = broad commodity pressure; weak dollar + expanding PMI = broad commodity strength; geopolitical events = structural pulse moves (energy/grain most sensitive).
Why the dollar and commodities are negatively correlated long term
- Pricing effect: commodities are dollar-priced; dollar depreciation → the same dollar price is cheaper for non-US buyers → demand rises;
- Rates/liquidity channel: a weakening dollar often accompanies easing expectations → global liquidity expands → speculative demand for commodities rises;
- Financial asset character: the dollar is the vehicle of "risk-free assets"; when haven money flows out of dollar assets it often flows into real assets (gold the classic).
- Two exceptions to note: early in a crisis "the dollar and commodities rise together" (dollar haven bid + commodity supply shock, as in early Russia-Ukraine); when inflation runs out of control (both the dollar and commodities can strengthen on their own logic) — the negative correlation is "broadly true", not "always true".
Sorting commodities into macro "conditions groups"
| Group | Constituents | Main drivers |
|---|---|---|
| Growth group | Copper, aluminum, zinc, iron ore | Global/China manufacturing PMI, infrastructure and property |
| Inflation group | Crude oil, natural gas, agricultural products | Supply shocks + monetary easing |
| Haven group | Gold | Real rates + geopolitical risk + central-bank buying |
| Policy group | Ferrous (domestic) | Domestic growth-stabilization policy, production cuts, crude-steel output caps |
- Use: under different macro regimes, which group to buy becomes obvious — expansion: growth group; runaway inflation: inflation group; recession panic: haven group (directional hints only, not investment advice).
⑤ The Role of Commodities in Asset Allocation
| Role | Logic | Caveats |
|---|---|---|
| Inflation hedge | Commodity prices are themselves part of inflation; holding commodities directly hedges "prices rising" risk | Works only in rising-inflation phases; in low-inflation/deflation phases commodities are a drag |
| Low correlation with stocks/bonds | Commodities are set by supply-demand, stocks by earnings and rates; historical correlation is low (per latest data) | Correlation rises in crises (liquidity shocks where stocks and commodities fall together exist) |
| Gold's special status | Gold is the "misfit among commodities": both monetary character (haven, the other side of real rates) and institutional central-bank demand | Gold's allocation logic is entirely different from copper/oil — don't read it as an "industrial commodity" |
| Portfolio effect | A moderate allocation (e.g., 5%-10%, varies by person) can smooth portfolio volatility and improve risk-adjusted returns | Allocate via indices/ETFs, not high-leverage futures |
A simplified allocation sketch (for learning, not advice)
| Portfolio type | Commodity position | Composition of the commodity sleeve |
|---|---|---|
| Conservative | Under 5% | Mostly gold (haven + low volatility) |
| Balanced | 5%-10% | Gold + a commodity index fund (diversified) |
| Aggressive | Above 10% (with explicit risk capacity) | May add energy/industrial-metals themes |
- Two principles: ① participate via indices/ETFs, never use high-leverage futures for allocation; ② book the commodity "allocation position" and "trading position" separately — the allocation position may sit untouched for years; the trading position must run a strict stop-loss (figures are illustrative; per personal risk tolerance).
⑥ How Ordinary Investors Participate in Commodities
| Route | What it is | Threshold and traits | Suited for |
|---|---|---|---|
| Commodity futures | Trade futures contracts directly | High threshold (margin system + leverage), requires professional skill, roll and delivery issues | Professional traders |
| Commodity ETFs / QDII funds | On-exchange or cross-border funds tracking commodity indices (gold ETFs, oil funds, commodity index funds, per the latest product list) | Low threshold, no screen-watching; mind tracking decay (roll costs) and premium risk | Most ordinary investors |
| Commodity-themed stocks | Resource stocks (miners, oil companies, agrochemicals) | Low threshold; but share price ≠ commodity price (company operations and equity beta stack on top) | Those who also want the equity upside |
| Physical spot | Physical gold, etc. | Bars/coins carry premiums and storage costs | Long-term preservation needs |
- An oft-ignored difference: futures/ETF returns ≈ spot price change + roll P&L (contango/backwardation). Buy a "rolling futures-based ETF" in a contango environment, and even with the spot price flat the NAV can keep sliding (compare the VXX decay logic in 13 - Volatility and VIX).
⑦ Common Mistakes in "Commodity Investing"
Mistake 1: Holding commodities like stocks
- Stocks grow on earnings and can be "bought, held, and left to grow"; commodities have no earnings — holding them produces no intrinsic value — commodity profits come from the "spread", so you must answer "when does it rise, by how much, and when do I leave".
- The result: buying commodities with a stockholder's mindset and holding dead at the top for three years is the most common loss script in commodity investing.
Mistake 2: Ignoring roll costs
- Futures-based products roll monthly: in contango the roll bleeds continuously; only in backwardation does the roll earn.
- Assuming the NAV will track "how much spot rose" is the biggest trap in commodity ETF investing — check the product's tracking method and the term structure first, then judge the cost.
Mistake 3: Watching price, not inventories
| Price-only | Price + inventories |
|---|---|
| "Copper at a new high — chase!" | New high + low inventories + spot premium = genuine tightness, the trend has support |
| "Oil fell — buy the dip!" | Oil down + inventory builds + backwardation flipping to contango = ample supply; the dip is dangerous |
- Inventories (LME/SHFE/port stocks, the EIA weekly crude report) and the term structure (contango/backwardation) are earlier and more honest data than price — price is the conclusion; inventories and the term structure are the evidence (see the "commodity watchlist" at the end of 06 - Commodity Panorama).
Mistake 4: Trading sentiment instead of data, buying the "news top"
| News scene | Reality |
|---|---|
| "Oil is surging — get on board now" | Oil prices news in minutes; by the time retail sees the news the move is usually more than half done |
| "Metal X is short, going to price Y" | The "shortage narrative" may already be fully priced or overdrawn — check whether inventories and the term structure confirm |
| "Country X bans exports" | The first pricing of a supply shock is the fiercest; after that comes the back-and-forth of "will it materialize, can it be circumvented" |
- Commodity markets are highly transparent and extremely fast: news is "olds"; inventories and contango/backwardation are the "scene". Accounts that chase news into commodities have historically mostly underperformed do-nothing accounts (per latest market and statistical data).
💀 News is olds; inventories and term structure are the scene
News is "olds"; inventories and contango/backwardation are the "scene". Commodity markets price news within minutes; by the time retail sees the news the move is usually more than half done. Accounts that chase news into commodities have historically mostly underperformed do-nothing accounts.
Risk Warning
⚠️ Risk Warning
- Commodity futures carry built-in leverage and violent swings: 2020's negative WTI crude and the 2022 LME nickel squeeze are both textbook risk events (per latest historical data).
- Futures-based commodity ETFs have roll decay and tracking error: in contango, long-term holding steadily erodes NAV, with premium-collapse risk on top — read the fund prospectus before buying.
- The "supercycle" narrative guarantees no returns: history doesn't simply repeat; when the sell-side's "supercycle thesis" appears, the move is usually half over.
- Commodity prices are frequently shocked by geopolitics, weather, and policy; historical correlations (with equities/the dollar) can break down completely in extreme conditions.
- All index compositions, weights, ranges, and cycle dating in this article are teaching-basis descriptions — defer to index providers' latest methodologies and exchange data; this article does not constitute investment advice.