Neel_Proshun_DXC

If you’ve been following the "copper supercycle" narrative, you know the vibe has been pretty electric lately. We’ve seen record highs, talk of massive shortages and a "buy everything" mentality driven by the green energy transition.
But Goldman Sachs just tapped the brakes. In their latest update, they’ve shifted their 2026 outlook and it’s a classic case of high prices being their own best cure.
Here’s the breakdown of what’s changing and why the "supply gap" isn’t hitting quite as hard as we expected.
1. The Numbers: From Scarcity to Surplus
Earlier this year, the narrative was all about deficits. Now, Goldman has revised their 2026 global copper surplus forecast to 300,000 tonnes, up significantly from their previous estimate of 160,000 tonnes.
For context, they also bumped their 2025 surplus estimate to a whopping 500,000 tonnes.

What does this mean for the price?
Goldman expects copper to cool off from its recent peaks, targeting around $11,000 per metric tonne by the end of 2026 roughly an 18% drop from the highs we saw earlier this year.
2. Why the Shift? (The "Human" Factors)
Markets aren't just spreadsheets they react to people and policy. Three things are driving this surplus:
The Scrap Response: When copper prices shot past $12,000, people didn't just keep buying they started digging through the "junk drawer." High prices incentivized a massive wave of recycling and scrap recovery, which effectively acted as a "shadow mine" bringing more supply to the market than expected.
The "EV Diet": We’ve heard for years that EVs need tons of copper. That’s still true, but engineers are getting efficient. Newer EV models are being designed with lower "copper intensity" basically finding ways to use less of the red metal to keep costs down.
The Tariff Factor: There’s a lot of talk about a potential 15% U.S. tariff on refined copper. This led to a massive bout of "front-running" or stockpiling in late 2025 and early 2026. Once that stockpiling ends and the policy becomes clear, that artificial demand disappears, leaving the market with plenty of extra metal sitting in warehouses.
3. Is the Bull Run Over?
Not exactly. It’s more of a breather than a collapse.
Goldman is still very bullish on the long-term. They view this 2025–2026 surplus as a "transient condition." While we have enough copper right now because of scrap and efficiency, the long-term structural problems haven't gone away.

Ore grades are still falling (it’s getting harder to find high-quality rock).
New mines take 10+ years to permit and build.
The Grid is the new EV: Power infrastructure and data centers (hello, AI) are projected to drive 60% of copper demand growth through 2030.
Goldman is essentially saying, "The party got a little too loud, too fast." By 2026, the market will be better balanced, prices will likely settle into a more sustainable range ($10,000–$11,000) and the "scarcity premium" will fade at least until the next big demand wave hits.
#Copper #GoldManSachs #RedMetal #EnergyTransition #RenewableEnergyRevolution $SOL $BTC $ETH

One of the less discussed aspects of modern data-driven systems is how they handle uncertainty. Most systems today are designed to process inputs, validate them and produce outputs in a consistent and reliable way. That structure works well in environments where data is clear and decisions can be derived directly from it.

But not all situations fit that model.

In many real-world cases, data exists without fully capturing the context needed to make a strong decision. Information can be accurate but incomplete valid but insufficient. These are the kinds of situations where uncertainty is not a flaw, but a natural part of the environment.

The problem is that most systems are not built to express that.

Instead of signaling uncertainty, they tend to convert whatever data is available into a usable output. Verification ensures that the data is authentic, and once that condition is met, the system proceeds. There is no built-in mechanism to pause and acknowledge that the available information may not be enough to support a meaningful conclusion.

This creates a subtle but important distortion.

From the outside, everything appears certain. Inputs are validated, outputs are generated, and decisions are made. There is no visible indication that the underlying data might be incomplete or that alternative interpretations could exist.

Over time, this can lead to a form of misplaced confidence.

Users begin to rely on the system not just for verification, but for judgment. The presence of an output is interpreted as a sign that the system has enough information to support it, even when that may not be the case.

The issue is not that the system is incorrect.

It is that the system is not designed to express the limits of what it knows.

In traditional decision-making processes, uncertainty often plays a visible role. Experts may disagree, additional information may be requested or decisions may be delayed until more clarity is available. These mechanisms allow uncertainty to be acknowledged and managed.

In contrast, systems that prioritize efficiency and consistency tend to move forward as soon as minimum conditions are met. They reduce friction by avoiding hesitation but in doing so, they also reduce the visibility of uncertainty.

This becomes more significant as systems scale and are applied to more complex scenarios.

The range of situations they encounter expands, including cases where data is ambiguous, conflicting or incomplete. Without a way to represent uncertainty, these systems continue to produce outputs that may appear equally reliable, even when the underlying conditions differ significantly.

That is where the risk lies.

Not in the failure of the system, but in its inability to communicate the limits of its knowledge.

A system that cannot say “I don’t know” may still function correctly at a technical level. But it also creates an environment where uncertainty is hidden rather than addressed and where decisions can carry more confidence than the data actually supports.

In the long run, the challenge is not just improving verification or increasing efficiency.

It is finding ways to make uncertainty visible again.

Because without that, even accurate systems can lead to outcomes that feel certain, while quietly resting on incomplete understanding.
#SignDigitalSovereignInfra $SIGN @SignOfficial

One thing I’ve been thinking about lately is how most verification-driven systems are designed to answer one specific question, is this data valid? If the answer is yes, the system moves forward. If not, it stops. That binary clarity is what makes these systems efficient and scalable.

But there’s another question that doesn’t get the same attention, who is responsible when valid data leads to a bad outcome?

At first, this might not seem like a major concern. If the data is correct and the process is followed, then the system is technically working as intended. Verification ensures that nothing has been altered, that the source is legitimate and that the structure is consistent. From a technical standpoint, that is success.

The issue starts to appear when decisions are built directly on top of that verified data.

In many cases, systems treat verified inputs as reliable enough to act on without further interpretation. That reduces friction and speeds up processes, which is exactly what these systems are meant to do. But it also shifts the burden of judgment away from the system itself.

The system verifies. It does not evaluate.

That distinction becomes important when outcomes are not as expected.

Imagine a situation where all inputs are valid, every check passes and the system executes exactly as designed yet the final result is still flawed. In traditional setups, there is usually a point where responsibility can be traced, whether it is a decision-maker, an institution or a process that failed to apply proper judgment.

In a verification-driven environment, that clarity becomes less obvious.

Each component can point to the fact that it did its job correctly. The issuer created valid data, the system verified it accurately and the application used it as intended. No single part appears to be at fault, yet the outcome still raises questions.

This creates a form of distributed responsibility where accountability becomes difficult to assign.

Over time, this can lead to a subtle but important shift in how systems are trusted. Users may begin to assume that if something is verified, it is also safe to rely on. But verification was never designed to guarantee good outcomes. It only guarantees consistency and authenticity.

That gap can remain hidden as long as outcomes align with expectations.

However, as systems scale and are applied to more complex situations edge cases become more common. Decisions that require context, nuance or judgment may not fit cleanly into a framework that prioritizes structured validation over interpretation.

What makes this challenging is that the system does not fail in a visible way?

There is no error message, no obvious breakdown. Everything functions correctly at a technical level. Yet the results can still feel incomplete or misaligned with real-world expectations.

I don’t think this means verification systems are flawed. They solve important problems and make coordination significantly more efficient. But they also redefine where responsibility sits.

Instead of being embedded within the system, responsibility moves outward to the edges, to issuers, to users and to the contexts in which data is applied.

The system proves that something is valid.

It does not take responsibility for what happens next.

As more decisions rely on these systems, that distinction becomes harder to ignore.
@SignOfficial #SignDigitalSovereignInfra $SIGN