Traders said volumes were likely to remain muted due to a U.S. holiday, with attention focused on copper and zinc.

Benchmark copper on the London Metal Exchange touched a record high of $14,533 a metric ton, surpassing its previous all-time peak of $14,527.50 reached in January. It later pared gains slightly to $14,518 by 1515 GMT, up 0.7%.

Also read: Brent crude oil price nears $100 as attacks on tankers threaten to worsen supply shock

Traders and producers have been shipping large volumes of copper to the U.S. since President Donald Trump first mooted import tariffs in February last year. Comex copper stocks currently stand at a record 766,795 short tons, or 695,624 metric tons.

Live Events

“It’s hard to ​know what will happen on tariffs, but the longer ⁠there is ‌uncertainty, the longer prices will remain elevated as material flows ​to the U.S.,” ​Albert Mackenzie, an analyst at Benchmark Mineral Intelligence, told Reuters.
Elevated premiums, or backwardations, for nearby contracts over longer-dated forwards have encouraged some copper to flow back to the LME.However, cancelled warrants, which represent metal earmarked for delivery, stood at 51%, indicating that more than 121,000 tons of copper is likely to leave the LME system over the next few weeks.

The premium for cash copper over the three-month forward contract climbed above $430 a ton in the middle of August, its highest level since 2021. It closed at around $74 on Friday.

Another indication of tightening supplies outside the U.S. is the level of stocks in warehouses monitored by the Shanghai Futures Exchange. Inventories stood at 63,000 tons, down 85% since the middle of March and at their lowest level since January 2024.

Copper prices on the SHFE are also in backwardation, suggesting that the industry in top consumer China is increasingly concerned about supplies.

Elsewhere, zinc gained 1.1% to $3,988 a ton, supported by tight supplies, after touching its highest level since May 2022 at $3,998.

A weaker U.S. currency, which makes dollar-priced metals cheaper for holders of other currencies, provided broader support to base metals.

Aluminium rose 0.7% to $3,315 a ton and tin added 0.4% to $55,100, while nickel slipped 0.7% to $16,725 and lead declined 0.3% to $1,903.

(Disclaimer: This article is based on inputs from agencies. These do not represent the views of The Economic Times)

MIAMI — An Amazon Prime Air cargo jet was still moving at about 129 mph when it ran off a Miami International Airport runway, crossed a perimeter road and struck vehicles, killing five people and injuring five others on a busy Labor Day weekend afternoon.

The Boeing 767-300 freighter, Prime Air Flight 7598, arrived from San Juan, Puerto Rico, and overran Runway 30 around 2 p.m. Sunday. Tracking data compiled by Flightradar24 put the jet at 112 knots — roughly 129 mph — as it left the usable pavement. Miami-Dade Fire Rescue Chief Ray Jadallah said five people died, three were taken to a trauma center in critical condition and two others were hospitalized with less severe injuries. Officials had not said by Sunday night whether the dead were in the cockpit, in cars on the road, or both.

The National Transportation Safety Board sent a go-team led by Chair Jennifer Homendy and scheduled the agency’s first public briefing for Monday. Investigators are expected to start with a simple question captured on video and flight tracks: where the airplane touched down, and why the crew did not abandon the landing if the jet floated past the normal zone.

Aviation safety specialist Steve Arroyo said the typical touchdown area is within the first 3,000 feet of a runway. If the cargo jet was beyond that, he said, the pilots should have gone around. Mary Schiavo, a pilot and former inspector general of the Transportation Department, said online video showed the Amazon plane remaining airborne well down the strip without the usual flare that puts the main gear on the pavement. It was not raining, she said, but dark storm clouds were nearby and strong winds had been reported. “A tail wind would certainly add to your troubles if you’ve already given up a good chunk of the runway by not touching down and landing right away,” Schiavo said.

CNN’s review of air-traffic control audio found the crew did not declare an emergency before landing. Flightradar24 data showed the jet rose above the normal approach path to Runway 30, then dropped back toward the glide path. The runway is more than 9,000 feet long. That length did not keep the airplane on airport property.

The 767 crossed a safety area and a fence, then hit vehicles on a road used by warehouse workers and businesses that ring the field. It came to rest on its belly near two semitrucks and a parking lot that serves an Amazon warehouse. Thick black smoke rose as fire burned. Jadallah said crews trapped in the cockpit had to be reached through a window by ladder and that occupants were cut from cars. Firefighters also freed a person pinned under a vehicle. Miami-Dade Fire Rescue sent more than 60 units and about 200 people. Jadallah said crews arrived within about 30 seconds of the first call and were still dealing with a fuel leak later in the afternoon.

Miami-Dade Mayor Daniella Levine Cava said at least five people had lost their lives and five others were injured, and that rescue crews reached the scene within seconds. Sheriff Rosie Cordero-Stutz put the time near 1:58 p.m. A student pilot, Rohan Ellis, who saw the wreckage, told a reporter the vehicles “were mangled.” Names of the dead had not been released.

The airplane, registered N1997A, is 32 years old. It was built as a passenger 767 around 1994, flew for airlines for more than two decades and was converted to a freighter in 2015, according to Cirium and Flightradar24 records. 21 Air, a Greensboro, North Carolina, cargo carrier that also flies for DHL, operated the trip under contract. Amazon does not hold the air-operator certificate for those flights. 21 Air said it was devastated by the accident involving one of its aircraft. CEO Keith Winters said the company was cooperating with federal investigators. Boeing said it was supporting the inquiry.

Amazon spokesperson Kelly Nantel said the company was “heartbroken to learn that five people lost their lives” and extended sympathies to families and others affected. In another statement she said Amazon was “still gathering details” and would “cooperate fully with any investigation.” Schiavo said investigators will review the jet’s maintenance history, including periods of foreign ownership.

Miami International is among the busiest U.S. airports for freight. It does not have an engineered materials arresting system — the crushable bed the FAA says has been installed at more than 120 airports and credited with helping save at least 497 lives in overruns. Mike O’Donnell, the FAA’s former director of airport safety and accident investigations, said Miami appears to meet federal rules because it has a 1,000-foot safety area beyond the runway. That standard followed the 1999 American Airlines Flight 1420 overrun in Little Rock, Arkansas, which killed nine.

The crash shut the airport’s runways and taxiways on a holiday Sunday. FlightAware logged more than 160 cancellations and nearly 325 delays by late day. Some flights diverted to Orlando, more than 200 miles north. Two of four runways later reopened and flights resumed Sunday evening, but the disabled 767 remained at the northwest end of the field.

The accident follows last year’s UPS crash in Louisville, Kentucky, in which a jet lost an engine on takeoff and killed three people in the cockpit and 12 on the ground. Sunday’s sequence was different: a landing, not a departure; a long runway; no declared emergency on the frequency CNN reviewed; and enough leftover speed to leave the airport and hit traffic.

What investigators will pull first is mechanical and human at once. Flight data and cockpit voice recorders should show airspeed, thrust, spoiler and reverse-thrust use, brake pressure and what the pilots said in the last minutes. Weather records will show whether a tail wind was on Runway 30. Maintenance logs will show the converted 767’s recent work. Videos and ADS-B tracks already sketch a long float and a high-speed exit. They do not explain it.

Until Homendy speaks, the confirmed picture is narrow and grim. A contracted Amazon 767 from San Juan did not stop on Runway 30. It left the pavement at highway-plus speed, hit vehicles on a working road and burned. Five people are dead. Five are in hospitals. The airport is moving airplanes again. The reason the jet used up the runway is the question the NTSB came to Miami to answer.

Simon Crutchley
CEO & Executive Director

Good afternoon since it is almost 12:00 afternoon, not morning. It’s not a very good morning in Johannesburg. I don’t think we like presenting results unless the sun is shining. But welcome to everybody who’s taken the trouble to come and join us at the JSE today. Thank you and also to everybody online.

These are our results for the 30th of June 2026. We’ve got a pretty normal presentation. Key features and results history, I think I’ll cover. Justin will talk to you about some of the more detailed group financial numbers, and then we’ll go into the performance and prospects and then take questions. It’s certainly a tale of 2 halves. The first semester was much stronger. Second was much more challenging. Certainly, from February onwards, steep price increases in energy — and certainly, quarter 4 of the second semester was particularly challenging from a sales point of view where we saw a market slowdown with respect to anxieties with our wholesale and distribution channel.

So revenue only increasing by 1.4%, some leverage, 4.4% at the operating profit level, but strong protection of the margins across the portfolio, and we’ll go into some of that in more detail. Pretty tough semester, particularly the second semester with creamer, but still a very, very strong results, but a lot more competitive than the prior year.

I&J, a good recovery in the fishing

AUSTIN, Texas — Tesla’s official Robotaxi account said riders can run a Cybercab cabin by talking to Grok, amplifying a passenger video that showed climate, video and music changing without a finger on the screen.

“Talk to Grok to control your Cybercab,” the @robotaxi account posted on Sept. 6, quoting Owen Sparks, who had written: “Cybercab has some pretty in-depth Grok integration, so for many actions, you never have to touch the screen. You can just use natural language to control the climate, YouTube, Music, and it will even send a notification to your phone so you can easily edit your destination.”

The post did not list a software version, a city-by-city rollout or a claim that every Austin Cybercab already behaves that way. It did put Tesla’s own stamp on a feature that matters in a car with no steering wheel and no pedals: if no one is driving, the cabin has to work like a lounge, not a cockpit.

Sparks’s clip, about three and a half minutes long, shows the production two-seater moving through Austin traffic while a large landscape display handles the trip. The visualization puts the Cybercab on a digital roadway with other vehicles. On-screen copy tracks an arrival window measured in minutes. A Grok chip sits in the interface. The rider opens YouTube and Apple Music, adjusts climate settings labeled low, medium and high, and later holds a phone showing an “Edit Dropoff” card for the Courtyard by Marriott Austin Downtown Convention Center, with a fare near $26. Buttons visible on the bar include Pull Over and Support.

That is the product Tesla is selling this week, three days after it began offering paid Cybercab rides in limited parts of Austin. The gold two-seater joined driverless Model Ys already in the Robotaxi app. Elon Musk called the vehicle, in a launch video, “the first car that is specifically built for unsupervised full self-driving.” Texas records listed about 45 Cybercabs registered to the fleet around the launch. NHTSA said it was “in contact with Tesla and is evaluating the situation” after the wheel-free design entered passenger service.

Voice control is not a new Tesla idea. It is new as the public pitch for this body style. In Tesla’s 2026 Summer Update notes the company wrote that Grok “can now make phone calls, search and play music, adjust climate, open the glovebox, and answer questions about your Tesla.” Tesla also said the summer release uses “Grok Think Fast 2.0.” In privately owned cars with the right infotainment hardware, testers have described chained commands — climate, wipers, seat heaters — in one spoken sentence. Cybercab applies that stack to a cabin where there is no driver to reach a stalk or a wheel.

The destination piece is the one that crosses from entertainment into the trip itself. Sparks said Grok can push a notification to the rider’s phone so the drop-off can be edited there. The video shows the phone card, not the car rewriting the route from the cabin screen alone. That split is practical. Tesla’s robotaxi service is geofenced. Changing an address mid-ride is a billing and routing event as much as a voice trick. The phone remains the booking tool. Grok becomes the hands-free remote for comfort and media.

Musk has sketched a deeper layer that the Robotaxi post does not claim is live. Asked in June about talking to Grok “like we can with an Uber driver,” he wrote that “this functionality will be there in about 3 months or so.” He said Tesla would eventually have “fully integrated voice control commands that can control the vehicle,” but that it was “just a bit too early for that.” Three months from late June lands near this launch window. The official Cybercab line still describes cabin control — climate, YouTube, music, a phone ping for the pin — not “turn right here” as a spoken override of the driving stack.

That distinction matters to regulators and to riders. A voice that lowers the temperature is an amenity. A voice that changes lanes on command is a driving control. Cybercab’s Rider Guide still centers the in-cabin screen for trip functions and a stop or pull-over control that brings the car to a safe stop and connects the passenger to support. Tesla has also said riders will be able to plug in a PlayStation or Xbox and has teased theater, music, gaming and later Starlink connectivity. A reply under the Robotaxi post asked for TIDAL HiFi and Spotify Lossless because the car has “Starlink onboard + great sound system.” Tesla did not answer in that thread.

The commercial context is still small. Paid Cybercab trips are Austin-only. Tesla’s Model Y robotaxis run in additional Texas and Florida markets. The company has not posted a Cybercab expansion calendar or a consumer MSRP. Musk has previously said the car should cost under $30,000 and that Tesla would sell units to customers, not only operate them. The September event did not confirm either number. Ashok Elluswamy, Tesla’s head of AI software, posted last month that “the streets won’t be the same anymore.” Asked whether Cybercabs would flood Austin, Musk replied, “Yes.” Forty-five registered cars and a voice demo are not a flood.

They are, however, a consistent software strategy. Tesla wants the same assistant in a Model Y glovebox command and in a robotaxi that has no glovebox driver. Grok already lives in Tesla cars as a chatbot with a “Hey Grok” wake word. The summer notes extended it into calls and climate. Cybercab is the first vehicle where talking to that assistant is not a convenience for a person who is also steering. It is the intended way to live in the cabin for the length of a $26 downtown hop.

What Tesla has not published is a reliability rate for those spoken commands in the robotaxi fleet, or whether every public Cybercab in Austin is on the same Grok build Sparks filmed. Community tests of the summer update in regular Teslas found mixed hit rates on long command lists. A robotaxi ride is shorter and the command set Sparks named is smaller: temperature, YouTube, music, a destination edit on the phone. That is an easier demo than “drive like my usual Uber.”

The company that posted “Talk to Grok to control your Cybercab” is the same company asking riders to trust a car with no wheel. The sentence is marketing. It is also an accurate description of the cabin Tesla showed: a screen that can stay dark while a voice changes the climate, a phone that still owns the pin, and a two-seat car that keeps moving through Austin whether anyone touches glass or not.

Harbor Capital is an asset manager focused on curating an intentionally select suite of active ETFs that they believe have the potential to produce compelling, risk-adjusted returns within a portfolio. Note: This account is not managed or monitored by Harbor Capital, and any messages sent via Seeking Alpha will not receive a response. For inquiries or communication, please use Harbor Capital’s official channels.

NEW YORK — Gmail users across the United States reported blank inboxes, stalled logins and messages that would not send on Labor Day, even as Google’s public Workspace dashboard listed no incident.

DownDetector said complaints about Gmail rose above typical levels beginning at 11:01 a.m. Eastern time on Monday, Sept. 7. The tracker posted its standard alert — “User reports indicate problems with Gmail since 11:01 AM EDT. How is it affecting you? #GmailDown” — a few minutes before a similar warning for Google as a whole. Those posts measure a surge in people saying the same product failed. They are not a company admission.

Independent checkers split the difference. Down For Everyone Or Just Me logged several short Gmail events on Monday, including stretches it timed at roughly 32 to 43 minutes, with users tagging send-mail and receive-mail failures into the late afternoon Eastern. UptimeRobot said probes to gmail.com completed without errors. IsDown, which watches Google’s official status page, said Gmail was “working normally” when it last checked. The Google Workspace Status Dashboard displayed “No incidents” in updates through Monday afternoon Coordinated Universal Time, with Gmail marked available.

That mismatch is the story. Gmail is not a single website. It is a web app, iOS and Android apps, IMAP and POP connections, Workspace tenants for schools and companies, and sign-in that also unlocks Calendar, Drive and Meet. A regional DNS problem, an authentication hiccup or a path that fails on one internet provider can look like “Gmail is down” in a kitchen in San Diego while Google’s global monitors still reach the front door.

Comments posted on DownDetector described exactly that kind of patchy failure. A user in San Diego wrote, “No gmail, no maps, no calendar.” Another in the same city, on AT&T Wi-Fi around 2 a.m. Monday, said the “GMail suite not loading, webpage slow.” In California’s Central Valley, a poster said they could not load Gmail or YouTube around 3 a.m. while “everything else is running fine.” In Sugar Land, Texas, a user said Google had been down “all day” and that “only workaround is to turn on VPN.” Others reported the inverse: turning on a VPN made Gmail load. One Windows user said the Google app launcher returned, “Can’t seem to load the app launcher right now.”

Those are first-person reports, not a root-cause finding. Several of them mention AT&T. DownDetector also posted a separate AT&T alert early Monday. A provider-side problem can impersonate a Gmail outage because the mailbox never gets a chance to answer.

The Labor Day wave followed a louder Sunday. The Houston Chronicle reported that DownDetector first showed excessive Google complaints about 2 a.m. Sept. 6 and that by about 9 a.m. the volume was running about 10 times higher than normal. Readers described Gmail that would not open in a browser or an app, search pages that sat blank and YouTube thumbnails that never appeared. Sportskeeda, writing Sunday night, said Google’s Workspace dashboard still showed no confirmed incident while users reported trouble with Search, Gmail and YouTube. The Grants Pass Tribune later summarized the same pattern: hours of intermittent trouble, no matching major incident on Google’s public boards.

Google had not, by late Monday in the United States, published a named Gmail outage explaining the holiday reports. The company has been more specific in other recent disruptions. In June, after Gemini inside Workspace failed, Google wrote that “the issue was triggered by a performance issue in our backend database which impacted the retrieval of Gemini App tools catalog.” That statement belongs to a different product and a different month. It is useful only as a reminder that Google will sometimes describe a backend fault after users have already spent a workday refreshing.

Gmail’s scale makes even a short, uneven stall expensive. Google has said the service has more than 1.8 billion users. A holiday Monday is when travel plans, school schedules, small-business invoices and family threads all sit in the same inbox. If the web client fails but a desktop mail app still downloads messages over IMAP, the store is probably healthy and the front end is not. If every device on one household connection fails and a phone on cellular works, the fault may be the home network. If a VPN fixes it, routing or DNS is a better suspect than a wiped mailbox.

Support desks will keep reciting the same checks because they still isolate the problem. Open Google’s Workspace status page before social media. Try an incognito window. Try basic HTML Gmail. Switch from the app to the mobile site. Test a different network. Confirm the account is not over its storage quota. Look at whether Calendar and Drive fail at the same moment, which often points to sign-in rather than mail delivery.

None of that converts a DownDetector spike into a historic blackout. The tracker itself says it flags an incident when reports jump well above the usual volume for that hour. A holiday can produce more people poking the same app at the same time. It can also hide a real, regional failure that never crosses Google’s internal line for a public incident.

What can be said on the record is limited and enough. On Labor Day 2026, enough people could not load, send or receive Gmail that DownDetector raised an alert at 11:01 a.m. Eastern. Similar complaints had already run through Sunday morning. Google’s Workspace dashboard did not match that noise with a posted outage. Users from California to Texas described blank pages, dead apps and VPNs that suddenly made mail appear. Until Google names a cause — or the charts fall back to baseline and stay there — the accurate description is an unconfirmed, uneven disruption of one of the internet’s default inboxes, reported by the people who could not open it.

This article was written by

Follow

Analyst’s Disclosure: I/we have no stock, option or similar derivative position in any of the companies mentioned, and no plans to initiate any such positions within the next 72 hours. I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it (other than from Seeking Alpha). I have no business relationship with any company whose stock is mentioned in this article.

Bohdan Kucheriavyi is not a financial/investment advisor, broker, or dealer. He’s solely sharing personal experience and opinion; therefore, all strategies, tips, suggestions, and recommendations shared are solely for informational purposes. There are risks associated with investing in securities. Investing in stocks, bonds, options, exchange-traded funds, mutual funds, and money market funds involves the risk of loss. Loss of principal is possible. Some high-risk investments may use leverage, which will accentuate gains & losses. Foreign investing involves special risks, including greater volatility and political, economic, and currency risks and differences in accounting methods. A security’s or a firm’s past investment performance is not a guarantee or predictor of future investment performance.

Seeking Alpha’s Disclosure: Past performance is no guarantee of future results. No recommendation or advice is being given as to whether any investment is suitable for a particular investor. Any views or opinions expressed above may not reflect those of Seeking Alpha as a whole. Seeking Alpha is not a licensed securities dealer, broker or US investment adviser or investment bank. Our analysts are third party authors that include both professional investors and individual investors who may not be licensed or certified by any institute or regulatory body.

The chancellor may have thought about shifting the venue of his first major speech this morning.

It was an absolute and total coincidence that he chose to make it in the Coventry Manufacturing and Technology Centre (MTC), just a few minute’s drive from the city’s totemic HQ of Jaguar Land Rover (JLR).

In the event, no one was pulling punches. The 4,000 office-based job losses at JLR were confirmed as John Healey answered questions a few miles away.

Healey chose to lean into it as an example of the global turbulence against which the UK needs more resilience.

The difference between this chancellor and his predecessor was he also feels that stressing fiscal discipline should instil confidence in consumers, businesses and investors, and not sap it away amid rolling fears of tax hikes.

So the prime minister and his chancellor have been engaged in a conscious attempt to boost the economic vibes – “a new story”, as he called it.

In his speech, and in a BBC interview afterwards, this was clear. There are some positive underlying signs from consumer, business and recruiter confidence measures that indicate this message has been heard, although the hot weather and World Cup have also been factors.

The Downing Street duo are trying to maintain this precious commodity of confidence and the promotion of economic animal spirits that have been suppressed by years of political and economic chaos.

“Borrowing costs are too high,” Healey acknowledged at the start of his speech. The global rise in bond yields is like a boa constrictor squeezing the Budget maths, slowly and relentlessly, occasionally loosening its grip, but never actually slithering away entirely.

It raises some reasonable questions therefore about how to balance reining in borrowing with protecting mildly better confidence?

The speech did not seem to me to prepare the ground – a “pitch roll” – for significant tax rises, as I heard at the equivalent Budget “scene setters” for the past two years.

“Am I right?” I asked the chancellor on this point. A pause as he sought precisely the right word formulation. “I won’t comment. I can’t comment. No chancellor can ahead of a Budget I will take and announce on October 28th”.

OK, what about the advice from one of the PM’s favourite economists Lord O’Neill that the rise in borrowing rates is a golden opportunity to scrap the triple lock?

This lock guarantees the state pension rising each year in line with either inflation, wage increases or 2.5% – whichever is the highest – but some have argued it is unaffordable.

“The prime minister has said, like I have, that we must bring down welfare costs, but we are also responding to the extreme pressure that is there in wider markets,” Healey said.

Neither answer was a denial, but nor were they any type of confirmation. Like a stubborn batsman, the chancellor is going to stay at the crease, giving little away, in his case, about next month’s Budget.