Actualités financières en direct

A 15-minute slide wiped out leveraged bullish bets and extended a choppy stretch for the largest crypto asset Bitcoin slipped below $82,000, and leveraged bulls paid for it almost immediately. Roughly $120 million in …

A 15-minute slide wiped out leveraged bullish bets and extended a choppy stretch for the largest crypto asset Bitcoin slipped below $82,000, and leveraged bulls paid for it almost immediately. Roughly $120 million in long positions across crypto were liquidated in the span of 15 minutes. What happened in those 15 minutes The drop pushed Bitcoin under the $82,000 mark during intraday trading on October 12. As the price fell, exchanges began forcibly closing leveraged long positions, which are bets that the price would keep rising. That forced selling is called a liquidation. It also tends to feed on itself, because every automatic sell order adds more downward pressure on the price. Data from platforms including CoinGlass and Hyperliquid pointed to heavy long exposure among traders heading into the move. CoinGlass reported that long positions made up the bulk of the liquidations in recent events. Shorts, the traders betting on declines, had a considerably better afternoon. Advertisement By the latest readings, Bitcoin was changing hands in the low $82,000 range. Traders were watching a support zone between $81,300 and $82,000, a band where buyers have previously stepped in. Not the first wobble this month On October 8, Bitcoin dropped below $81,000, touching intraday lows of approximately $80,400 to $80,500. That earlier slide triggered over $1 billion in liquidations within 24 hours, mostly on the long side. A further $696 million in related liquidations came alongside it. The news moving money, markets, and the world—before your day starts. Daily. Free. Join 34,000+ readers across crypto, finance, and policy. Email address Subscribe free We respect your privacy. Unsubscribe anytime. Compared with that, Sunday’s $120 million flush looks almost restrained. The speed, though, is the notable part: the October 8 figure accumulated over a full day, while the latest one landed in a quarter of an hour. Both declines followed a run that took Bitcoin to highs in the $85,000 to $87,000 range earlier in October. Profit-taking near the $87,000 peak helped kick off the retreat. US spot Bitcoin ETFs reportedly saw outflows of $487 million in a single day, which removes a meaningful source of steady buying. No significant regulatory actions or protocol changes were reported around this stretch of price action. The swings appear tied more to traders unwinding leverage than to any single fundamental headline. The bounce pattern After earlier sharp drops, Bitcoin has rebounded fast, at one point climbing back above $83,000. What this means for traders and investors The $81,300 to $82,000 support band is the near-term level to watch. A clean break below it could trigger another round of forced selling if traders remain heavily leveraged. A single $487 million outflow day doesn’t define a trend, but continued redemptions would mean less institutional demand absorbing the selling when leveraged traders get flushed. The October 8 low near $80,400 also serves as a reference point. Revisiting that level would put the market back where the $1 billion-plus liquidation day bottomed out.

Leaked tiers point to a single plan ranging from around $8 to $200 a month, though X has yet to confirm pricing or a launch date X wants to be the only subscription line …

Leaked tiers point to a single plan ranging from around $8 to $200 a month, though X has yet to confirm pricing or a launch date X wants to be the only subscription line on your credit card statement. A post on the platform describes a worldwide rollout of one plan that merges Grok, Grok Bot, Cursor and X Premium into a single bill. As of October 11-12, 2026, neither X nor its affiliates had announced a launch date or confirmed pricing for the bundle. What the bundle reportedly includes The product is being tested under the name “X Subscription.” Some users have unofficially nicknamed it “XPass.” Perks from X Premium would sit alongside a full menu of Grok AI features: chat, voice, Imagine, Build and Grok Bot. Cursor, the AI coding environment, would also be included. Advertisement Leaked screens from early October show an “X Subscription” upgrade option tying these services together. The same leaks outline four potential tiers: Premium: around $8 per month around $8 per month Plus: around $30 per month around $30 per month Super: around $100 per month around $100 per month Ultra: around $200 per month The structure is expected to replace three existing standalone products: X Premium, SuperGrok and Cursor’s own subscription. Instead of juggling separate plans, users would draw from a shared pool of usage credits across all the services. AI, tech, and the markets they move—in one daily briefing. Daily. Free. Join 34,000+ readers across crypto, finance, and policy. Email address Subscribe free We respect your privacy. Unsubscribe anytime. How X got here Grok Bot entered beta on August 11, 2026, and access was initially limited to select subscription tiers. On September 12, 2026, product lead Maxime Prades said the team was working on unifying subscription plans and indicated a rollout was expected soon, though no firm date was attached. The push follows the integration of xAI with Cursor that has been ongoing since mid-2026. What this means for users, X and its rivals Someone already paying separately for X Premium, SuperGrok and Cursor could collapse three bills into one, with credits that flow to whichever tool they use most that month. A jump from around $8 to around $30 is meaningful for casual users, and the around $100 and around $200 tiers are clearly aimed at heavy AI and coding users. Exactly which features land in which tier remains unconfirmed. Cursor’s inclusion stands out in particular. Developers are one of the most valuable audiences in AI, and folding a dedicated coding environment into a mass-market social subscription is an unusual move. Existing subscribers on SuperGrok or a standalone Cursor plan will be looking for answers on how their current pricing and usage limits translate. The things to watch are an official announcement from X, confirmed pricing for each tier, and a clear breakdown of what every level actually includes.

The exchange says TIA sends and receives are lagging, while buying, selling and fiat transactions remain unaffected Coinbase says users are facing delays when moving Celestia’s TIA token on and off the exchange. Buying, …

The exchange says TIA sends and receives are lagging, while buying, selling and fiat transactions remain unaffected Coinbase says users are facing delays when moving Celestia’s TIA token on and off the exchange. Buying, selling and fiat transactions are still working as usual. What Coinbase is reporting The issue is limited to sends and receives on the Celestia network. Those are the on-chain transfers that move tokens between a Coinbase account and an outside address. Everything that happens inside Coinbase’s own books keeps running. That covers order matching, purchases with dollars, and sales back into fiat. Coinbase has posted this kind of notice before. On November 26, 2025, the exchange flagged delayed Celestia sends and receives around 13:22 PST. The issue was resolved by November 27, 2025. During that episode, Coinbase confirmed that trading activity, including buying, selling and fiat transactions, was not affected. It also assured customers that their funds were secure. Advertisement Similar disruptions affecting Celestia transfers for Coinbase users were also noted in January 2026 and April 2026. Coverage of the November 2025 incident came primarily from Coinbase’s own status updates and Reuters. Neither pointed to broader market implications. Why Celestia transfers work differently Celestia launched in October 2023 as a modular data availability network, with TIA as its native token for transactions. The news moving money, markets, and the world—before your day starts. Daily. Free. Join 34,000+ readers across crypto, finance, and policy. Email address Subscribe free We respect your privacy. Unsubscribe anytime. The modular design splits up jobs that older chains handle all at once. Celestia focuses on consensus and data availability, meaning it orders transactions and guarantees the underlying data is published and retrievable. Execution, the part where transactions actually run, gets handed off to other layers. Coinbase offered staking and custody support on day one of the network’s mainnet launch, and TIA remains actively traded on the platform. The upgrade backdrop The delays arrive while Celestia is pushing ambitious technical changes. Its v10 “Fibre” upgrade is focused on raising throughput, with testnet benchmarks reaching approximately 3.07 terabits per second. As of October 2026, the network had an active development roadmap. Celestia’s circulating supply sits at approximately 920 million TIA. What this means for TIA holders and Coinbase users The practical impact depends on how you use TIA. Traders who keep their tokens on Coinbase can keep buying and selling as normal, according to the exchange. The people who feel it most are those moving tokens in or out. That includes users who want to stake through a third-party wallet, shift TIA to another exchange, or deposit tokens they want to sell on Coinbase. After the November 2025 incident, there was no significant market reaction or expert analysis tying it to changes in trading behavior for related assets. That incident also wrapped up within roughly a day. The thing to watch now is Coinbase’s status page. A quick resolution would match the pattern from November 2025.

As the workforces of the U.S., U.K., and similar wealthy Western nations age, other countries with younger populations grow at a clip. Nowhere is this more true than India, where the median age is …

As the workforces of the U.S., U.K., and similar wealthy Western nations age, other countries with younger populations grow at a clip. Nowhere is this more true than India, where the median age is 29 (a decade younger than the U.S.) and GDP is growing at near… The median age in India is 29 years old 10 years younger than in the United States. But despite its youth and high growth, the countrys economy faces hurdles.

Traders loaded up on downside protection after a sharp price drop, even as longer-term open interest still leans toward calls Bitcoin traders spent the week shopping for insurance. Bearish options accounted for 53.5% of …

Traders loaded up on downside protection after a sharp price drop, even as longer-term open interest still leans toward calls Bitcoin traders spent the week shopping for insurance. Bearish options accounted for 53.5% of Bitcoin’s options volume this week, according to data shared on X. The post framed the shift as a sign of institutional demand for downside protection. A selloff, then a scramble for hedges Bitcoin fell roughly 7.3% from a peak of $86,728 to a low of $80,394 on October 8. As of October 12, Bitcoin had stabilized in the $83,500 to $83,750 range. That bounce did not calm the options market. For the October 16 expiry, the put-to-call ratio hit 1.43, which means puts outnumbered calls by 43%. A put option gives the holder the right to sell at a set price, so it pays off when prices fall. A call does the opposite and profits when prices rise. Advertisement Around $451 million in notional value for those near-term puts sits at strike prices between $76,000 and $80,000. The bigger picture still leans bullish Total Bitcoin options open interest across platforms sits somewhere between $37 billion and $37.45 billion. Calls make up about 60% of that total. Open interest put-to-call ratios have hovered between 0.66 and 0.71, which is firmly call-leaning. The news moving money, markets, and the world—before your day starts. Daily. Free. Join 34,000+ readers across crypto, finance, and policy. Email address Subscribe free We respect your privacy. Unsubscribe anytime. Much of that call positioning targets longer-dated strikes, such as $95,000. Volatility and the macro backdrop The 30-day at-the-money implied volatility ranged from 34% to 35.7% during the period. Realized volatility, which measures how much prices actually moved, came in at about 37% to 38%. When realized volatility runs above implied, options are pricing in calmer conditions than traders have recently lived through, which can make protection look relatively cheap compared with recent price action. Overall sentiment in the options market shifted from a calls-richer stance to a puts-bid posture. Traders have also been positioning ahead of releases like the FOMC minutes, which can move risk assets as investors parse the Federal Reserve’s thinking on rates. What this means for traders and investors The research points to institutional participants protecting against bearish moves rather than abandoning their longer-term view on Bitcoin. The concentration of puts at $76,000 to $80,000 deserves attention. Large clusters of open contracts at specific strikes can become reference points that traders watch closely as expiry approaches. The October 16 expiry is the near-term date to monitor, with a put-to-call ratio of 1.43 tied to that date. With open interest in the tens of billions, derivatives positioning is no longer a side show to spot trading. If realized swings stay elevated and implied volatility catches up, protection gets more expensive, and late hedgers pay the premium.

Google is currently leading in the competition to develop the best AI model for October 2026, as reported by Seeking Alpha. The company’s flagship AI model, Google Gemini 4 Argon, has been highlighted as …

Google is currently leading in the competition to develop the best AI model for October 2026, as reported by Seeking Alpha. The company’s flagship AI model, Google Gemini 4 Argon, has been highlighted as a frontrunner among its competitors. According to prediction markets, Google has a 70% likelihood of having the top AI model by the end of October. This figure significantly surpasses the probabilities for Anthropic and OpenAI, which are estimated at 23-26% and 1%, respectively. While Argon performs well on some benchmarks, other evaluations place it behind Anthropic’s Claude Opus 5.5, suggesting variability in its competitive ranking depending on the criteria used. Key Takeaways Market pricing suggests Google’s Gemini 4 Argon currently leads in the race for the best AI model in October. Anthropic’s probability for achieving the top AI model status by November end has decreased, reflecting recent developments and competitive standings. OpenAI maintains a lower probability in this race, consistent with current market pricing and benchmark evaluations. What to Watch Markets will be closely watching the performance of AI models on independent benchmarks, which could shift perceptions and rankings. Notable is whether Google can maintain its lead through November, as further releases or updates from Anthropic or Meta might alter standings. The ongoing development and public availability of these models will be crucial indicators of market dynamics and potential shifts in perceived leadership.

This article examines how recent events may relate to prediction market pricing. It reflects interpretive analysis of publicly available information and is provided for informational purposes only. Observers should monitor announcements from the Saudi …

This article examines how recent events may relate to prediction market pricing. It reflects interpretive analysis of publicly available information and is provided for informational purposes only. Observers should monitor announcements from the Saudi General Authority of Civil Aviation and Saudi Air Navigation Services for any escalation in security measures or further suspensions. The actions of the Houthi movement and any additional missile strikes on Saudi infrastructure could influence market perceptions of risk. Developments in diplomatic relations between Saudi Arabia and Iran may also impact the likelihood of airspace closures. Saudi Arabia has prolonged the suspension of operations at Riyadh’s King Khalid International Airport as authorities conduct thorough safety inspections. This development follows recent attacks by the Iran-backed Houthi movement, as announced by the General Authority of Civil Aviation (GACA). The Houthi attacks have previously resulted in significant casualties and infrastructure damage, prompting the need for extensive safety assessments. Flight operations remain halted, with diversions and cancellations affecting numerous travelers. Disclaimer This article contains analysis of publicly available information and market data and is for informational purposes only. It does not constitute investment advice or a recommendation to buy, sell, or hold any asset or contract. Content may include AI-assisted interpretation and may be incomplete or subject to change. Market conditions may evolve rapidly, and the timing of information may affect how it is interpreted. Market participants may act on similar information at or around the time it becomes available. You are solely responsible for any decisions made based on this content. For additional details, please review our full Disclaimer & Risk Disclosure.

Shares have rebounded more than 50% from June 2026 lows as cloud demand and a $678 billion contract backlog ease AI spending fears Microsoft is back in the $4 trillion club. The company closed …

Shares have rebounded more than 50% from June 2026 lows as cloud demand and a $678 billion contract backlog ease AI spending fears Microsoft is back in the $4 trillion club. The company closed at a market capitalization of $4.02 trillion on October 12, 2026, its first sustained stay above that line since October 29, 2025. The rebound matters because it answers a question that has hung over the entire tech sector. Investors had been asking whether all that money poured into artificial intelligence would ever pay off. At least for Microsoft, the market now seems to think the answer is yes. How Microsoft got its groove back Microsoft shares have recovered by more than 50% from their lows in June 2026. The engine behind the rally is Azure, Microsoft’s cloud computing business. Azure revenue grew 43% year over year in the June quarter. Management expects that pace to pick up further. Guidance points to growth of 45% in constant currency, a measure that strips out the effect of exchange rate swings. Advertisement Microsoft’s remaining performance obligations, or RPO, for cloud and AI contracts reached $678 billion, an increase of 84%. RPO is essentially revenue that customers have signed up for but Microsoft has not yet recognized. For investors worried about AI spending, that figure carried a lot of weight. It suggests enterprises are committing real money to AI infrastructure, not just experimenting with pilot projects. AI, tech, and the markets they move—in one daily briefing. Daily. Free. Join 34,000+ readers across crypto, finance, and policy. Email address Subscribe free We respect your privacy. Unsubscribe anytime. A chip of its own Microsoft also used the same day to make a hardware statement. The company unveiled its Azure Maia 300 AI chip at the OCP Global Summit on October 12, 2026. The chip is expected to help Microsoft reduce its dependence on graphics chips from Nvidia and AMD. It is also expected to improve efficiency inside the company’s data centers. The road from $4 trillion and back Microsoft first crossed the $4 trillion threshold in July 2025. At the time, it became only the second company in the world to reach that valuation. The stock slipped after that peak as concerns mounted over the profitability of AI spending and the trajectory of cloud growth. Even after reclaiming the milestone, Microsoft now ranks as the fourth largest US company by market capitalization, behind Nvidia, Apple, and Amazon. What this means for investors and the tech sector Contracted demand, visible in that $678 billion backlog, appears to be validating the company’s heavy spending on cloud and AI infrastructure. There are still risks. Guidance for 45% Azure growth is a forecast, not a guarantee. A backlog also has to be converted into recognized revenue over time. Key items to watch include whether Azure hits its constant currency growth target, how quickly the RPO figure translates into reported sales, and whether the Maia 300 meaningfully improves data center economics.

The wafer-scale AI chipmaker now trades about 21% below its $185 IPO price as lock-up expiry, insider selling and Nvidia competition weigh on shares Cerebras Systems shares sank to a post-IPO low between $146.90 …

The wafer-scale AI chipmaker now trades about 21% below its $185 IPO price as lock-up expiry, insider selling and Nvidia competition weigh on shares Cerebras Systems shares sank to a post-IPO low between $146.90 and $146.98 on October 12, 2026. That leaves the AI chip designer roughly 21% below its $185 IPO price and approximately 62% under its first-day intraday high of $386. From $386 to under $147 The stock (NASDAQ: CBRS) debuted on May 13, 2026, with an IPO price of $185 per share. The company sold 30 million shares in the offering, raising between $5.55 billion and $6.4 billion. Shares opened at $350, touched an intraday high of $386, and closed at $311.07. That close represented a 68% gain over the offering price. At its first-day peak, the market valued Cerebras at around $95 billion. By early October, Cerebras’ market capitalization had fallen to approximately $39 billion. Advertisement Three pressure points Lock-up expiration. IPO lock-ups bar insiders and early backers from selling shares for a set period after a listing. When that window closes, a wave of previously restricted stock can suddenly hit the market. Insider selling. Insider sales of approximately $84 million were planned around October 1, 2026. AI, tech, and the markets they move—in one daily briefing. Daily. Free. Join 34,000+ readers across crypto, finance, and policy. Email address Subscribe free We respect your privacy. Unsubscribe anytime. Nvidia competition. Over the summer, a report from SemiAnalysis indicated that critical workloads were shifting from Cerebras to Nvidia’s technology. How Cerebras got here Cerebras was founded in 2015 and is based in Sunnyvale, California. Its core idea is architecturally bold: instead of slicing a silicon wafer into many small chips, it uses the full wafer as a single AI accelerator. The road to the public market was not smooth. Cerebras had sought to go public in 2024, but that effort hit regulatory delays tied to a significant investment from UAE-based Group 42. When the company finally listed in 2026, its customer roster included OpenAI and Amazon Web Services, with multi-year contracts covering substantial compute capacity. What this means for AI chip investors Cerebras went from a 68% first-day pop to trading below its offering price within five months. Anyone who received IPO allocations at $185 and held is now underwater. Anyone who bought in the opening frenzy near $350 or higher is sitting on far steeper losses. Several things are worth watching from here. First, whether insider selling continues beyond the roughly $84 million planned around the start of October. Second, any updates on the OpenAI and AWS contracts, including signs of expansion or reduction in committed capacity. Third, further independent analysis, like the SemiAnalysis report, on where AI labs are placing their critical workloads. For now, the $146.90 to $146.98 range marks the floor Cerebras has hit since listing.

The post XRP Traders Are Leaning Long by Two to One, and Whales Have Stopped Sending Coins to Exchanges. One Analyst Says $1.38 Gets Tested Before Any Rally. appeared first on 24/7 Wall St.. …

The post XRP Traders Are Leaning Long by Two to One, and Whales Have Stopped Sending Coins to Exchanges. One Analyst Says $1.38 Gets Tested Before Any Rally. appeared first on 24/7 Wall St.. The XRP (CRYPTO: XRP) long-to-short ratio now stands at 2.32, accord… The post XRP Traders Are Leaning Long by Two to One, and Whales Have Stopped Sending Coins to Exchanges. One Analyst Says $1.38 Gets Tested Before Any Rally. appeared first on 24/7 Wall St..The XRP (…