LearnBits

Silver hit a record high in January, then lost nearly half its value by the time the US-Iran war broke out in February. It's now trading around $58-59 an ounce, clawing back small gains after a brutal stretch that saw silver futures suffer their worst single-day drop since the 1980s. That kind of round trip — record high to near-collapse to tentative recovery, all inside six months — is the real story behind any "best commodity for 2026" conversation right now, not the headline price itself.
What's driving the swing is almost entirely the Federal Reserve. New Fed Chair Kevin Warsh has taken a notably hawkish line, reaffirming the central bank's inflation fight even as markets had hoped he might ease up under political pressure. May's PCE inflation came in at 4.1% year-over-year, and traders are now pricing meaningful odds of a rate hike as early as September. That's a hostile environment for any non-yielding asset — gold included, which has slipped roughly 7.5% year-to-date and briefly dipped below $4,000 an ounce before recovering to around $4,060.
Gold and silver aren't moving in lockstep right now, and that gap matters. The gold/silver ratio sits near 69, with silver actually outperforming gold on a percentage basis the past two sessions — a sign of the ratio compressing back from extremes, even within an otherwise rough year for both metals. Macquarie's analysts frame it simply: both metals are now "macro driven" again, reacting to Fed signals and dollar strength rather than the safe-haven panic that fueled January's records. They expect gold to average around $4,641 for 2026 overall despite the recent slide, before drifting lower into 2027.
The detail easiest to miss in all the daily price noise: central banks haven't stopped buying. Nearly 90% of central bank respondents in the World Gold Council's latest survey still expect global reserves to grow over the next year, and bar-and-coin demand hit a near-record 474 tonnes in the first quarter — buying driven by long-horizon institutions, not short-term traders chasing momentum.
That's really the split worth understanding heading into the back half of 2026. Silver is the high-beta trade, swinging hard in both directions on Fed expectations and thin market depth. Gold is behaving more like the slower, structurally-supported asset it's historically been, even through a volatile year. Anyone treating "precious metals" as one trade right now is missing how differently these two are actually behaving under the same macro pressure.

Air raid sirens went off in Bahrain twice within hours on Sunday morning. Between 2 and 3 a.m. local time, Iran's Revolutionary Guard launched ballistic missiles and drones at Ali Al Salem Air Base in Kuwait and the US Fifth Fleet headquarters at Port Salman, claiming to have destroyed eight pieces of American infrastructure. The IRGC called it retaliation for a fresh US strike on five Iranian coastal positions hours earlier — the same cycle that's now repeated itself at least five separate times since March.
What makes this pattern worth noticing isn't the violence itself; it's the targeting logic underneath it. Ali Al Salem has now absorbed five IRGC strikes since early March — more than any other American facility in the Gulf. Saudi Arabia, despite hosting Prince Sultan Air Base, hasn't been touched once since a single strike there on March 27. That's not an accident. Saudi Arabia lacks a formal Status of Forces Agreement with Washington the way Kuwait and Bahrain do, which gives Tehran a legal and diplomatic argument for treating those two bases as fair game while leaving the Saudis alone — a calculation that lets Iran keep hitting US assets without dragging the Gulf's most powerful state directly into the fight.
The ceasefire framework holding all of this together is barely six weeks old. Trump and Iran's government signed a memorandum of understanding on June 17, giving both sides sixty days to negotiate an end to the war that began in late February. Sunday's exchange is the clearest sign yet that the document is straining under its own weight: Trump posted that Tehran had violated the agreement and warned of consequences "if that happens," while Iran's foreign ministry fired back that the US strikes themselves broke the same memorandum it's now accusing Iran of violating.
Both governments under direct attack are furious, but for different reasons than Washington might expect. Bahrain called the strikes a violation of its sovereignty and pushed for an urgent UN Security Council session. Kuwait used almost identical language, describing the attacks as a "flagrant violation" of its own territory — language aimed as much at Tehran as it is a quiet signal of frustration at being caught in the middle of someone else's war. Qatar, notably, has stayed out of the firing line entirely, having maintained a $6 billion credit line with Iran and an open diplomatic channel that's kept its bases untouched through four months of regional escalation.
That's the actual mechanics of this conflict: it isn't random. Every strike, every excluded target, every diplomatic statement is calibrated, which is exactly why a ceasefire that keeps almost breaking down hasn't fully collapsed yet — both sides are still leaving each other room to step back, even while trading missiles.
#IRGCSaysItStruckKuwaitAndBahrain

A single AI data center, the kind hyperscalers are now building by the dozen, can chew through copper at a rate that would have sounded absurd five years ago. The going estimate runs anywhere from 30 to 150 tonnes per megawatt depending on whether grid infrastructure is counted, and at gigawatt scale that adds up to roughly 50,000 metric tons per facility. With around 15 gigawatts of new AI capacity going up annually, that's 750,000 tonnes of fresh demand a year from data centers alone — before a single transmission line upgrade or substation gets counted.
That timing couldn't be worse for a market already running short. Disruptions at Indonesia's Grasberg mine and the Kamoa-Kakula complex in the Democratic Republic of Congo have knocked a meaningful chunk of global refined output offline at precisely the moment demand is accelerating. Ore grades at legacy mines have fallen roughly 40% since 1991, and building a new mine from discovery to first production routinely takes 15 to 29 years once permitting is factored in. The International Copper Study Group has revised its 2026 forecast from an expected surplus to an outright 150,000-ton deficit — a swing large enough to reshape how every major bank is pricing the metal.
That repricing is already visible. J.P. Morgan projects copper averaging around $12,075 a tonne for 2026, with UBS reaching further out to $15,000 by early 2027. Not everyone agrees on the timeline — Goldman Sachs still expects a modest surplus this year before structural deficits take hold in 2027, and Red Cloud actually trimmed its near-term forecast on softer US demand even while raising its 2028 and 2030 numbers. The disagreement isn't really about direction. It's about how soon the math catches up with itself.
What makes this cycle different from past copper rallies is that AI isn't competing with electrification for the metal — it's stacking directly on top of it. EVs, grid modernization, and renewable buildouts were already going to strain supply before a single AI campus broke ground. Robert Friedland, the mining veteran behind the Kamoa-Kakula discovery, put it about as bluntly as anyone in the industry has: the world is heading toward a shortage measured in the billions of tonnes over the coming decades, not the thousands. Whether 2026 ends up the year prices break out or merely the year everyone agreed they eventually would, the supply side has nowhere near enough lead time to catch up on its own schedule.

Three and a half million people, give or take, fell for a coin that never actually existed on any blockchain. That's the part of the OneCoin story that still doesn't fully land even years later: no ledger, no mining, no real tokens — just a Bulgarian-born scheme dressed up in cryptocurrency language and sold through stadium conferences and a global multi-level-marketing network. Founders Ruja Ignatova and Karl Sebastian Greenwood defrauded as many as 3.4 million investors worldwide, and total losses are estimated anywhere from $4 billion to $15 billion depending on the count.
Now there's a deadline, and it's almost gone. The Department of Justice's remission program, open since mid-April, closes June 30. Anyone who purchased OneCoin between 2014 and 2019 and suffered a net financial loss can petition for a share of the recovered funds — but the pool available is roughly $40 million, a fraction of what was actually stolen. Filing is free, doesn't require a lawyer, and runs only through the DOJ's official remission portal, managed by Kroll Settlement Administration. The FBI has been blunt about one thing: anyone charging a fee to "help" file a claim is almost certainly running a second scam on top of the first.
The math here is uncomfortable but worth sitting with. Even if every eligible victim files in time, $40 million split across a fraud this size means most people will recover only cents on every dollar lost. That's not a flaw in the program — it's just what's left after years of asset forfeiture against people who spent much of the money before anyone caught up to them.
Greenwood is already serving 20 years and forfeited $300 million of his own. Ignatova, the "Cryptoqueen" who marketed OneCoin as Bitcoin's eventual successor, vanished in 2017 and remains on the FBI's Ten Most Wanted list, with a $5 million reward still sitting on her capture. She built a currency that was never real. The deadline closing this week is about the only thing in this story that is.
#FBIUrgesOneCoinVictimsToSeekDOJCompensation

It took one drone hitting one cargo ship to send oil traders scrambling again. On June 25, the Singapore-flagged Ever Lovely was struck southeast of Oman's coast, forcing the International Maritime Organization to pause its plan to evacuate roughly 11,000 sailors stranded around the Strait of Hormuz. By Friday, CENTCOM had launched retaliatory strikes on Iranian missile and drone infrastructure along the coast. Trump called it "a foolish violation" of the ceasefire on Truth Social — and then, when reporters asked if Iran would face consequences, gave reporters nothing more than: "You'll find out."
This is the fourth time in roughly three months that the same pattern has repeated: a maritime incident near Hormuz, a spike in oil prices, a US strike, and then — somehow — the broader truce holds anyway. Brent crude actually settled down nearly 4.3% on Friday at $71.99 a barrel even as the attack happened, because traders are increasingly pricing this as noise around an agreement rather than its collapse. That's a real shift from March, when the same kind of incident sent Brent above $100.
The reason markets are calmer now traces back to June 17, when Trump and Iranian president Masoud Pezeshkian signed the Islamabad Memorandum — a formal framework meant to end the war within 60 days, built on a prior agreement to reopen Hormuz and begin negotiating Iran's nuclear program. CENTCOM lifted its naval blockade the next day. That document is doing real work even as both sides keep testing its edges.
What hasn't changed is the underlying leverage Iran holds, and it's worth saying plainly: control of the strait isn't really up for negotiation in practice, whatever the paperwork says. Former US energy advisor Amos Hochstein put it bluntly on CNBC weeks ago — regional leaders already assume Tehran controls Hormuz for the foreseeable future regardless of what any deal states. Roughly a quarter of the world's seaborne oil trade and a fifth of its LNG still depends on a waterway that one drone can shut down for days.
So far, neither side has let an incident become the moment everything falls apart. Whether that holds through the rest of the 60-day window — with nuclear talks, frozen asset disputes, and the unresolved Lebanon front all still on the table — is the actual story driving every oil trader's next move, not the headline of a single strike.#USIranCeasefireBreaksDown

The world's largest crypto exchange just got locked out of its biggest regulated market, and the reason traces back to a single rejected application nobody saw fail until it was too late to fix.
Starting July 1, Binance can no longer onboard new clients or run regulated services in France, Italy, Poland, and Spain, after failing to secure a license under the EU's Markets in Crypto-Assets regulation.
The exchange had pinned its entire EU strategy on Greece, applying through the Hellenic Capital Market Commission back in January for a license that would have let it "passport" across all 27 member states from a single approval. That strategy collapsed on June 24, when Binance withdrew the Greek application rather than wait for what multiple reports describe as an imminent rejection.
The reasoning behind that rejection reportedly had less to do with paperwork and more to do with history — lingering concerns over the exchange's anti-money laundering record, and whether founder Changpeng Zhao could clear MiCA's "fit and proper" bar for company leadership.
The numbers around this deadline tell their own story. Of more than 3,000 crypto firms operating in Europe, only around 210 secured MiCA authorization in time — a conversion rate low enough that even an exchange of Binance's scale couldn't force its way through. Rivals including Coinbase, Kraken, and OKX made the cut. Binance, despite reportedly spending 18 months working with regulators, did not.
What's actually happening to customers is narrower than the dramatic headlines suggest. This is a suspension and orderly wind-down, not an exit — withdrawals stay active, funds remain accessible, and Binance insists it isn't telling anyone to pull their assets by the deadline. New orders, deposits, sign-ups, and yield products like staking and launchpool are what's actually freezing. Binance now says it's pursuing licensing through France instead, though that path carries its own complications: French regulators have an active investigation into the company, and any approval realistically lands well after the July 1 cutoff, leaving a gap with no clear end date.
The deeper signal here isn't really about one exchange. MiCA was built to end the old patchwork era, where crypto firms shopped around fragmented national rules for the friendliest regulator. Binance just found out that strategy has limits — and that being the biggest player in the room doesn't guarantee a seat once the rules stop bending.

Two months ago, Switzerland's financial watchdog said giving banks unrestricted access to Anthropic's Mythos model would be dangerous enough to count as a systemic threat to the entire financial system. This week, that same regulator put roughly 100 policy and technology specialists in a room for a hackathon — to build AI tools for supervising crypto markets. The contradiction isn't really a contradiction. It's the whole strategy.
FINMA president Marlene Amstad laid out the logic plainly: hackers are already using AI to find vulnerabilities faster than banks can patch them, so financial watchdogs no longer have the option of staying analog. FINMA helped establish a forum within IOSCO specifically to push AI adoption across supervisory authorities overseeing roughly 95% of global financial markets (CoinGecko) — a coordination effort with a scope few regulatory initiatives can match.
What's actually being built is more specific than "AI for compliance." FINMA is running a generative tool that reads through pre-inspection documents and flags anomalies worth a closer look, paired with a second AI layer whose only job is catching when the first one hallucinates before anything reaches a human supervisor.
On the crypto side specifically, the agency has built a live dashboard that merges quarterly holdings data with daily price feeds, hunting for concentration risk — too much exposure resting on one institution, or one blockchain carrying too much systemic weight. That's a meaningfully different posture than simply reading reports after the fact; it's monitoring built to catch a problem while it's still forming.
The tension Amstad isn't dodging is the same one that triggered the Mythos warning in the first place: restricting access to frontier AI protects against one kind of risk while potentially starving regulators of the exact tools they need to fight another. Her position now is that Switzerland needs access to the most capable models precisely because the threats they're defending against have already arrived. Crypto markets never close, never respect borders, and increasingly get attacked by tools as sophisticated as the ones meant to police them — which is precisely why the regulator chasing the danger decided building its own AI was less risky than not having one at all.
#FINMAAcceleratesAIForCryptoOversight

Micron walked into its June 24 earnings report needing not just a beat, but a beat-and-raise big enough to justify a stock that had already climbed roughly 700% over the past year. It delivered exactly that.
Revenue hit $41.46 billion against estimates of $35.84 billion. EPS came in at $25.11 versus the $20.20 Wall Street expected. Shares jumped 15% in extended trading, pushing the company's market cap past $1 trillion for the first time.
The number that mattered most wasn't even the quarter just closed — it was the guidance. Micron projected next-quarter revenue around $50 billion, more than $6 billion above what analysts had modeled. That gap explains why a company growing revenue 268% year-over-year still felt pressure heading into the print. Broadcom and ASML both delivered solid beats earlier this year, only to see their stocks sink because the guidance attached wasn't aggressive enough. The bar for "good enough" had quietly moved, and Micron knew it.
Behind the headline figures sits a structural story that's been building for over a year. Micron's high-bandwidth memory supply for 2026 is entirely sold out, and CEO Sanjay Mehrotra said the company can fulfill only 50% to two-thirds of customer demand for key products. To lock that demand in, Micron has signed 17 Strategic Customer Agreements worth roughly $100 billion in guaranteed future revenue.
Management went further still, saying market tightness looks "locked in to persist beyond calendar 2027" — a claim that, if it holds, extends the pricing power story well past this single quarter.
None of that erases the stock's volatility. Shares have already endured two separate 20% pullbacks since March even while marching higher overall, and the rally has pushed valuation into territory where any hint of softening demand could trigger fast profit-taking. For now, the memory shortage driving Micron's pricing power shows no sign of easing — and that dynamic, more than any single headline number, is what's actually setting the next move.

OpenAI didn't pick a random partner to challenge Nvidia. It picked Broadcom — and itself. The custom chip, called Jalapeño, was unveiled this week and pitched as delivering better performance per watt than current state-of-the-art hardware, a direct jab at the company OpenAI still buys more chips from than almost anyone else. Broadcom shares ticked up. Nvidia dipped modestly. The market's initial read: manageable, but not nothing.
Context matters here, because Nvidia has already had a genuinely rough June. Early in the month, the Philadelphia Semiconductor Index fell 10% in a single session, with Nvidia alone shedding over $300 billion in market value as a hot jobs report revived fears of Fed rate hikes. The stock briefly dipped below a $5 trillion valuation before clawing back ground. None of that was about chip demand specifically — it was macro fear hitting the most crowded trade on the board.
What makes the OpenAI-Broadcom news different is that it's not macro noise; it's structural. Every major AI lab building its own silicon — and OpenAI is far from the first to try — chips away, literally, at the assumption that Nvidia owns the entire stack indefinitely. The counterweight is that Nvidia isn't standing still either. Jensen Huang's recent push into PC chips with the RTX Spark, plus the upcoming Rubin platform rolling out across AWS, Microsoft, and Google Cloud later this year, shows a company defending share on every front rather than coasting on its data center lead.
For now, Wall Street's read is more nuanced than panic. Hyperscaler capex commitments north of $750 billion for 2026 suggest the overall AI buildout is still intact, custom chips or not — OpenAI remains one of Nvidia's largest buyers even as it diversifies. The real test arrives at the next earnings call, where forward guidance, not the headline numbers, will likely decide whether this becomes a real dent in the growth story or just one more competitive skirmish in a market that was never going to stay a one-company show.

Render is sitting 88% below its all-time high. Stacks is down 96%. Neither network has actually slowed down — fees keep flowing, tokens keep burning, transactions keep clearing — which makes the price charts look almost dishonest next to what's happening on-chain.
That gap between price and fundamentals is the real setup here, more than any single coin pick. A token at $60 can be undervalued; one at a fraction of a cent can be wildly overpriced. The question worth asking isn't "is it cheap," but whether the business behind the token has actually broken, or whether the market just stopped paying attention. Hyperliquid's HYPE token has pulled back about 19% from its mid-June high near $76, yet its on-chain perpetuals exchange recently got name-checked by the head of the New York Stock Exchange's parent company as something punching wildly above its twelve-person team. Injective runs a community buyback that shrinks supply as usage climbs. Render's burn-and-mint model does the same thing mechanically — scarcity tied to demand, not hype.
Timing matters too. Bitcoin dominance is still parked near 58%, and the Altcoin Season Index hasn't broken out of "Bitcoin Season" territory in any convincing way. The old pattern hasn't changed much: Bitcoin moves first, large-cap alts follow, and the smaller infrastructure plays bring up the rear — usually after most people have already stopped checking the charts.
That's exactly why the unglamorous news tends to matter more right now than the loud headlines. NEAR's resharding upgrade this month and Solana clearing over a trillion dollars in quarterly on-chain activity aren't the kind of stories that trend on social media, but they're the kind that quietly reprice a token once attention returns.
None of this is a guarantee. Corrections of fifty to seventy percent have buried plenty of projects that never recovered, and "discounted" doesn't automatically mean "underpriced." But for tokens where revenue and usage haven't followed the chart downward, the math tends to look very different in quiet months than it does once the headlines turn loud again.

More than fifty percent of all circulating Bitcoin is currently sitting on unrealized losses, a level of pain not seen since the depths of the 2022 bear market. The Fear and Greed Index has been pinned in the low-to-mid 20s for days, and spot Bitcoin ETFs have bled around $8 billion over the past month, with BlackRock's IBIT leading the redemptions. By almost every visible metric, this looks like capitulation.
Except one thing doesn't fit the story: corporate treasuries are buying. Firms like Strategy and Strive have collectively scooped up more than 1,279 BTC in recent weeks, even as ETF holders run for the exits. That split matters more than it might first appear. ETF outflows reflect retail and short-term institutional money reacting to a hawkish Fed and a tech-stock selloff that's dragged crypto down with it. Direct corporate accumulation reflects a different kind of investor — one bypassing fund structures entirely to hold Bitcoin on the balance sheet, sidestepping management fees in the process.
Bitcoin itself is trading in the $59,000–63,000 range, comfortably below its 200-day moving average and roughly half of October's all-time high near $126,000. Technically, almost every signal — RSI, MACD, moving averages — points bearish. But sentiment extremes and price action don't always move in lockstep, and that's the tension defining this exact moment. History shows two ways an extreme fear stretch like this resolves: either the macro backdrop eases and price finds a floor, or the bleeding continues until there's simply nobody left to sell.
What's different this cycle is the audience watching. A market once driven almost entirely by retail sentiment now has corporate balance sheets and ETF mechanics pulling in opposite directions — and that tug-of-war, more than any single chart pattern, is what's actually setting the next move.