WHEN NOISE ISN'T RISK

Media coverage of commodities markets and price moves has always, by and large, accurately reflected the strategic stories that matter to the largest audiences and connect to broader trends and themes.  They are a great measure of what audiences and markets are talking about and are also a good indication of how a commodity will behave. However, purchasers that rely too heavily on the number of headlines and articles (noise) to identify ongoing price risks (and even, rely on algorithms that, in effect, weigh rather than read coverage) without interrogating market dynamics and what the media are actually saying are increasingly missing a trick:  Lead attracted 14 times more coverage than palladium in 2025. Palladium was three times more volatile. Across 33 industrial commodities, the pattern holds: what markets talk about and what markets move are different things.   In commodity markets, attention and risk are often assumed to travel together.  So much so that corporate buyers and speculators use attention levels as a major buy or sell signal.  After all, when a commodity dominates headlines, it reflects that something important is happening — that prices are moving, supply is disrupted, or buyers should be paying attention.

But what if attention and risk follow different logics?  Markets are famously irrational after all.

An analysis of 33 industrial commodities — spanning energy, metals, and plastics — finds that the correlation between volume of  trade press coverage and realised price volatility in 2025 was just 0.33. Statistically weak. Practically, almost no relationship at all.

This isn't a criticism of commodity journalism. Coverage, by and large, accurately  reflects how a commodity price will behave and what's important to discuss: policy shifts, supply chain risks, strategic trends.  It is also biased towards the commodities that interest and  impact the largest audience - a small movement in steel will generate far more interest and headlines than say a major movement in an obscure but essential plastic. Volatility reflects something different: how much prices actually moved. Both matter. But they're not the same thing — and conflating them can lead procurement teams astray.

HOW WE MEASURED IT

Coverage was measured using Diffbot's Knowledge Graph, aggregating de-duplicated articles from specialist commodity publications including Metal.com, Reuters, Argus Media, and other industry sources — the publications procurement professionals actually read.

Volatility was calculated as annualised standard deviation of log returns from exchange settlement prices across calendar year 2025. Dataset: 33 commodities (8 energy, 14 metals, 11 plastics/petrochemicals).

Where coverage concentrates

Trade press coverage naturally concentrates where the stories are. In 2025, that meant metals.

Where trade press coverage comes from — and what it covers

Metals attracted 72% of all coverage across our dataset — over 26,000 articles. This reflects genuine editorial logic: metals sit at the intersection of trade policy, the energy transition, infrastructure spending, and China's economic trajectory. They're important.

Energy attracted 27% of coverage. Plastics and petrochemicals, despite representing a third of the commodities in our dataset, attracted just 1.4%.

None of this is surprising. Metals have better narratives. But here's the catch: narrative importance and price volatility are different things.

The coverage-volatility gap

Coverage per commodity vs average volatility by sector

Metals: Average 1,904 articles per commodity. Average volatility: 24%.

Energy: Average 1,226 articles per commodity. Average volatility: 40%.

Plastics: Average 46 articles per commodity. Average volatility: 13%.

Energy commodities were significantly more volatile than metals, yet attracted less coverage per commodity. This makes sense: a single natural gas price spike is less narratively rich than the ongoing EV transition story, even if the gas price creates more immediate P&L impact.

The question for procurement teams: which matters more for your hedging decisions?

The divergence in detail

The five most covered commodities in 2025 were Power (5,481 articles), Lead (4,992), Steel (4,066), Copper (3,008), and Iron Ore (2,611). Average volatility across these five: 27%.

The five most volatile were Natural Gas (71%), Power (70%), Palladium (46%), Platinum (41%), and Silver (38%). Average coverage: 2,041 articles.

Only four commodities appear in both top-ten lists: Natural Gas, Power, Silver, and WTI.

Coverage vs volatility across 33 industrial commodities. Correlation: 0.33.

The narrative commodities

Some commodities attract coverage because they're embedded in important stories, not because their prices are moving dramatically:

Steel connects to infrastructure bills, trade policy, and construction activity. It featured in 4,066 articles while delivering 14% volatility — well below the median.

Iron Ore is shorthand for Chinese property markets and industrial demand. 2,611 articles, 14% volatility.

Lead appears constantly in EV battery discussions. 4,992 articles, 15% volatility.

These are legitimately important commodities. But their coverage reflects strategic significance, not price instability.

Commodities with high coverage relative to their volatility

The quiet movers

Conversely, some commodities moved substantially while generating relatively little discussion:

RBOB Gasoline: 28 articles, 30% volatility. The blendstock that determines pump prices across North America — technically important but narratively unglamorous.

Rhodium: 54 articles, 28% volatility. Critical for catalytic converters, but lacking a compelling broader story.

Carbon Credits: 40 articles, 25% volatility. Increasingly material to industrial costs, but still finding its narrative footing.

Palladium: 352 articles, 46% volatility. The third most volatile commodity in our dataset, with half the coverage of comparably-sized metals markets.

The quiet movers: high volatility relative to coverage

A tale of two metals

The Lead-Palladium comparison crystallises the disconnect. Same asset class, same exchanges, similar industrial applications. Very different profiles:

Lead: 4,992 articles. 15% volatility.

Palladium: 352 articles. 46% volatility.

Lead's coverage reflects its role in energy storage narratives. Palladium's volatility reflects concentrated supply (Russia and South Africa), limited inventory buffers, and active substitution dynamics with platinum.

Both perspectives are valid. Lead is strategically important. Palladium is price-volatile. But a procurement team using coverage volume as a risk proxy would systematically over-attend to one and under-attend to the other.

Lead vs Palladium: 14x more coverage, 3x less volatility

Two different questions

Volume of coverage and volatility answer different questions:

Coverage answers: "What's strategically important? What connects to broader trends? What should I understand about the market?"

Volatility answers: "What's moving? Where is my P&L exposure? What requires active hedging attention?"

Both questions matter. But they're not the same question, and they don't have the same answer.

Trade publications serve the first question well. They help readers understand why commodities matter, how markets connect to policy and technology trends, the drivers of commodity price movements and where strategic risks lie. This is valuable - especially if the audience looks in detail at the tonality and content (see below.)

But for the second question — where is price risk right now? — coverage volume is a weak signal. The commodities generating the most discussion are not systematically the commodities generating the most volatility.

Read, don’t weigh the coverage

It should perhaps not be necessary to state this but the statistics show that readers who pay attention to content and tonality of coverage — not just volume of coverage  — would still receive accurate signals.

To test this, we analysed every headline in our dataset. Using GPT-4, we scored each headline on a 0-10 scale for how strongly it signalled upcoming price volatility. Headlines predicting surges scored the same as those predicting crashes — both indicate movement.

HEADLINE SCORING METHODOLOGY

Each headline was scored 0-10 by GPT-4 for volatility indication:

• 0-3: Stability signals ("prices steady", "market unchanged")

• 4-6: Moderate movement signals ("prices fluctuate", "some uncertainty")

• 7-10: High volatility signals ("prices surge", "market disruption", "supply crisis")

We computed each commodity's average volatility score and percentage of "high-volatility" headlines (score ≥7).

Media as an accurate lead indicator

Statistically speaking: commodity journalists do accurately distinguish between volatile and stable markets in their coverage.

For commodities with substantial coverage (200+ articles), headline tone correlates meaningfully with actual volatility (ρ = 0.46). The stable commodities — Lead, Steel, Iron Ore — have the lowest average headline volatility scores. The more volatile commodities — WTI, Palladium, Gas Oil — have higher scores.

In other words: when journalists have enough material to establish a pattern, that pattern reflects reality.

Left: Headline tone vs actual volatility. Right: High-volatility commodities receive more urgent headlines.

Stable commodities, calm headlines

The commodities that delivered low volatility in 2025 received appropriately measured coverage:

Iron Ore: The lowest headline volatility score in our metals dataset (3.4/10), matching its 14% actual volatility. Despite thousands of articles about China demand and property markets, the tone stayed measured — correctly reflecting prices that moved less than most assume.

Steel: Headline score of 4.0/10, actual volatility 14%. Coverage discussed policy developments and demand shifts, but rarely signalled crisis — because there wasn't one.

Lead: Headline score of 4.2/10, actual volatility 15%. The EV battery narrative generated volume, but journalists correctly avoided crisis framing.

Volatile commodities, urgent headlines

Conversely, commodities that moved sharply received more volatility-signalling coverage:

WTI: Headline score of 4.9/10, with 21% of headlines scoring 7 or above. Actual volatility: 32%. The coverage correctly flagged instability.

Palladium: Headline score of 4.7/10, with 26% high-volatility headlines — nearly double Lead's rate. Actual volatility: 46%. The signal was there.

Rhodium: The highest headline volatility score in our dataset (5.4/10), with 32% high-volatility headlines. Actual volatility: 28%. Journalists covering this thin market correctly emphasised its instability.

Lead vs Palladium: the signal was there

Return to the Lead-Palladium comparison. The coverage gap was stark — 4,992 articles vs 352. But within that coverage, did journalists differentiate?

Yes.

Palladium's 352 articles were more volatility-signalling than Lead's 4,992

Palladium's average headline volatility score was 12% higher than Lead's (4.7 vs 4.2). Its share of high-volatility headlines was 86% higher (26% vs 14%). A reader paying close attention to Palladium coverage would have encountered noticeably more urgent language than a reader of Lead coverage.

The journalists correctly identified which metal was unstable. Their readers just had to find the coverage first.

Attention allocation, not analytical failure

This analysis is important to understand the coverage-volatility disconnect. The issue isn't that commodity journalists misjudge volatility — when they cover a market, their tone and language generally reflects its behaviour. The issue is that some volatile markets receive far less coverage than stable ones.

A reader who followed Palladium coverage closely would have received accurate volatility signals. But most readers don't follow Palladium closely — because there are only 352 articles to follow, compared to 4,992 for Lead.

The trade press is analytically sound. The attention it allocates across commodities reflects factors beyond price volatility: narrative richness, policy relevance, trading volume, reader interest. These are legitimate editorial considerations. But they mean that coverage volume alone is not a reliable proxy for price risk.

Implications for procurement

Read the commodity press for strategic context and as a lead indicator. It's good at explaining why commodities matter and how markets are evolving.  Its also sound when it comes to identifying strategic risk and potential market movements.

Don't use coverage volume as a volatility signal. The correlation is too weak to be useful for hedging prioritisation.  Especially if you need to purchase or trade niche commodities that inherently matter to a smaller audience.

Build independent price monitoring. Historical settlement prices are widely available. Computing realised volatility is straightforward. For active risk management, there's no substitute for looking at the numbers directly.

Pay attention to the quiet movers. RBOB, rhodium, carbon credits, palladium — these delivered 25-46% volatility with relatively modest coverage. If you're exposed, make sure you're monitoring them directly.

The bottom line

Volume of Media coverage reflects narrative importance, the commodities that matter to the most people get the most coverage. Price volatility reflects market structure. In 2025, these pointed in different directions more often than not.

Lead generated 4,992 articles and 15% volatility. Palladium generated 352 articles and 46% volatility. Both facts are true. Both are useful. But they answer different questions.

For procurement teams managing real commodity exposure, the lesson is simple: noise and risk are not the same thing. Know which one you're measuring.