Volatility Creates Risk. Poor Data Compounds It.

First published on: 30/6/2026

Author: Alex Cerbone, VP Solutions Engineering at Zema Global 

In volatile markets, the quality of a decision depends on the quality of the data behind it. As commodity market teams work with tighter timelines and faster-moving signals, clean, timely, trusted data becomes essential to managing risk. 

This blog explores:

• Why volatility is only part of the risk companies are managing.

• How poor-quality data can make difficult market conditions harder to navigate.

• Why delays, inconsistencies, and manual workarounds matter more when timing is critical.

• What AI and advanced analytics need before they can deliver reliable insight.

• How companies can reduce uncertainty by building stronger data foundations. 

Recent disruptions across global supply chains have introduced a new level of uncertainty into already complex commodity markets. The Global Supply Chain Pressure Index eased slightly in May 2026 but remained close to its highest reading since July 2022, reflecting continued strain across transportation, capacity, and cost conditions. At the same time, U.S. consumer sentiment has been sitting near record-low levels, underscoring how quickly market pressure can translate into broader economic uncertainty.

Price swings are sharper, timelines are less predictable, and signals are moving faster across regions and asset classes than many companies are equipped to process.

But volatility itself isn’t the real story. Volatility is constant. It’s expected. It’s part of how these markets function.

What’s changed is something else entirely: the tolerance for error has collapsed.

Decisions that once had a buffer — time to validate, time to adjust, time to recover — now happen in compressed windows where the cost of being wrong shows up immediately. In that environment, the risk that comes with volatility is compounded when the data behind those decisions isn’t clean, timely, or trusted. It’s the interaction between external volatility and uncertainty in your own data that creates real exposure.

While everyone is operating under the same market conditions, not everyone is carrying the same level of uncertainty. And increasingly, that’s the differentiator — not the volatility itself, but how much additional risk is introduced by the data you rely on to navigate it.

The margin for error is closing

For years, many companies operated with an implicit safety net. Data could arrive late and still be useful. Minor inconsistencies could be worked through downstream. Teams could align after the fact and still land on the right outcome.

That safety net is disappearing.

Today, signals move across markets faster than manual processes can keep up. Dependencies between commodities, geographies, and asset classes are tighter, and delays don’t just slow decisions — they actively distort them. What might have once been a small discrepancy now has the potential to materially impact outcomes.

We are seeing this especially clearly in energy markets. As of June 22, 2026, the 90-day trailing average of the CBOE Crude Oil Volatility Index was 73.86, more than double its 2025 average. That kind of sustained volatility compresses decision windows and raises the cost of acting on incomplete or inconsistent information.

In practical terms, this means that “good enough” data and workflows are no longer sufficient. When the margin for error closes, the quality of your inputs becomes inseparable from the quality of your decisions.

You can’t out-analyze broken inputs

In uncertain markets, the instinct is often to respond with more analysis — more models, more scenarios, more layers of interpretation. But analysis doesn’t solve foundational issues.

If the underlying data is incomplete, delayed, or inconsistent, then the output is simply a more sophisticated version of the same problem. The illusion of precision increases, but the reliability does not.

This matters even more as companies begin applying AI to commercial and operational decision-making. McKinsey recently found that 65% to 85% of organizations expect to adopt generative AI or agentic AI in pricing over the next one to three years, up from just 10% to 30% today. But the same research found that some of the most significant barriers to adoption are data quality and availability, integration complexity, and security and compliance concerns.

That is the point. AI does not eliminate the need for trusted data. It raises the standard for it.

In conversations with customers, this is one of the most consistent themes we hear. The bottleneck isn’t the ability to analyze — it’s the effort required to trust data enough to begin. Teams are spending more time reconciling and validating than they are actually making decisions. And in an environment where timing matters, the delay is not neutral — it’s a disadvantage.

Control what you can control

No organization can control external volatility. Markets will move, disruptions will happen, and uncertainty will persist.

What organizations can control is the integrity of the data driving their decisions.

That requires a shift in focus, from reacting to external uncertainty to removing internal uncertainty. It means addressing the friction points that are often accepted as just part of the process: manual interventions that introduce lag, fragmented systems that produce conflicting views, and validation steps that happen too late to be useful.

Because when the margin for error narrows, “mostly correct” is no longer enough. Confidence in your data is a prerequisite for sound action.

Data as operational infrastructure

This is where we’re seeing a meaningful shift in how companies approach data. It’s no longer treated as a downstream input or a reporting layer. Increasingly, it is being recognized as core operational infrastructure — something that needs to be engineered with the same rigor as the systems it supports.

At its foundation, that means ensuring data is:

• Clean — standardized, validated, and consistent.

• Timely — available when decisions need to be made.

• Trusted — transparent, traceable, and defensible.

In conversations with customers, the question is no longer simply if data is available; it’s whether it can be relied on quickly, in moments where decisions carry real consequences.

A higher standard for decision–making

Markets will continue to evolve. Volatility will ebb and flow. External conditions will remain outside of any one organization’s control.

But the standard for decision-making has already changed.

The companies that will navigate this environment most effectively are not the ones reacting the fastest, doing the most analysis, or consuming the most data. They are the ones that have built a foundation where data is accessible and dependable under pressure.

Sources: University of Michigan Consumer Sentiment Index via FRED; Global Supply Chain Pressure Index via the Federal Reserve Bank of New York and Trading Economics; Cboe Crude Oil Volatility Index; McKinsey & Company, “B2B pricing: Navigating the next phase of the AI revolution.”