Track market-moving stories before they get noisy

A high yield on a restaurant stock can mean a bargain or a trap, and the difference comes down to numbers most investors never check before buying.

This post may contain links from our sponsors and affiliates, and Flywheel Publishing may receive compensation for actions taken through them.

Restaurant chains have spent 2026 finding out that weak traffic eventually hits the dividend. Wendy’s (NASDAQ:WEN | WEN Price Prediction) shares have fallen 65.14% over five years, and its board has cut the quarterly payout twice since early 2025. A high yield only matters if the company can keep paying it. Three well-known restaurant names show how fast that ability can wear down, and how to tell real stress from a stock that has simply fallen out of favor.

How a Falling Stock Makes a Yield Look Better Dividend yield is the annual payout divided by the stock’s price. If the price falls and the payout remains the same, the yield goes up on its own, even though shareholders get nothing extra. A struggling company can end up looking generous. A dividend becomes unsustainable when earnings, free cash flow or the company’s financial position can no longer comfortably pay for it. Check all three before you trust any yield.

Wendy’s (WEN): Two Cuts and a Turnaround Plan Still to Come Wendy’s now pays $0.07 per quarter, or $0.28 a year. At a share price of $6.12, that works out to a yield of about 4.58%. That is a high yield from a brand everyone knows. Be careful with stock screens that use the past 12 months of payments ($0.49). They show a bigger number than a new buyer will actually collect.

Here, the warning signs turned out to be real. The quarterly dividend went from $0.25 to $0.14 in 2025, then to $0.07 when the company reported results on Aug. 7, 2026. In the second quarter, U.S. same-restaurant sales fell 7% as traffic dropped 12.5%. Net income fell 40.8%, and management withdrew its full-year guidance. On the earnings call, CEO Bob Wright said: “Traffic is down, our value proposition has slipped, and franchisee economics are under pressure.” Net debt is 5.0 times earnings before interest, taxes, depreciation and amortization (EBITDA). The company also needs to refinance about $430 million of debt due in March 2028.

There is a case for the smaller payout. It equals about 42% of trailing diluted earnings per share (EPS) of $0.66, and first-half free cash flow rose to $120.3 million. Still, the chief financial officer said the new rate runs “slightly above” the company’s usual payout of 50% to 60% of adjusted net income. Things to watch: traffic, and the turnaround plan due with the next quarterly report.

Dine Brands Global (DIN): Profits No Longer Cover the Dividend Dine Brands Global (NYSE:DIN), which has Applebee’s and IHOP, pays $0.19 a quarter, or $0.76 a year. At $29.95, that is a yield of about 2.54%. The low yield is itself the warning. The payout was $0.51 until the company cut it, starting with the December 2025 payment. That was a 62.7% reduction. The stock is down 57.33% over five years.

Even after the cut, the dividend is hard to cover. It equals about 123% of trailing GAAP EPS of $0.62, meaning the company pays out more than it makes. Second-quarter net income fell to $4.3 million from $13.8 million. Adjusted free cash flow for the first half was just $3.7 million, compared with $48.7 million a year earlier. Capital spending rose to $23.2 million from $9.3 million. Shareholders’ equity is negative $292 million. Net interest costs rose to $22.0 million, and cash fell 49.8% to $97.5 million. On top of that, the board approved a new $100 million buyback, which competes with the dividend for very little spare cash.

Adjusted EPS was $1.16 in the second quarter, and management expects remodel spending to “ease as the program advances.” For coverage to recover, Applebee’s same-restaurant sales (down 1.8%) need to turn positive, and the company-owned restaurants need to reach break-even.

Domino’s Pizza (DPZ): A Fallen Stock With a Dividend That Still Holds Up Domino’s Pizza (NASDAQ:DPZ) pays $1.99 a quarter, or $7.96 a year. At $297.62, that is a yield of about 2.67%. The stock is down 29.48% over the past year, which drove the yield up. That makes it a useful test case: trap, or a strong company that investors have soured on?

The numbers say the dividend looks sound. It equals about 45% of trailing EPS of $17.65. In 2025, free cash flow came to $671.5 million, which covered $236.9 million in dividends about 2.8 times. The quarterly payout has gone up every year, from $1.21 in 2023 to $1.99 now.

The risks are real but further down the road. Second-quarter U.S. same-store sales rose just 0.1%, and EPS of $4.07 missed expectations of $4.17. Shareholders’ equity is negative $3.98 billion, with $4.77 billion in securitized notes (debt backed by the company’s franchise fees). Domino’s also spent $156.2 million on buybacks in the second quarter, so it has room to scale those back before the dividend comes into question. The thing to watch is whether flat sales growth holds up long enough to strain the debt.

How to Tell a Yield Opportunity From a Yield Warning A dividend cut usually pulls the stock down too, so a high yield is never a reason to buy on its own. A yield that signals opportunity is covered by earnings and free cash flow with room to spare, on a stock that fell because sentiment turned while the business remains healthy. A yield that signals stress comes with a payout above earnings, cash used up by capital spending or buybacks, and debt rising while traffic falls (we counted seven of these warning signs and put all of them in a free report: Dividend Traps).

Contact [email protected] for any questions or corrections.

Investors seeking to cash in on the global nuclear energy resurgence have focused on small modular reactor (SMR) upstarts Oklo (OKLO -0.75%) and NuScale. However, neither company can run one of their future reactors without fuel. That's where Centrus Energy (LEU +0.11%) comes into play, as it supplies the enriched uranium fuel that nuclear reactors need. Centrus recently announced a planned joint venture with Oklo to support growing demand for its specialized nuclear fuel.

Here's a closer look at this lower-risk way to invest in the country's nuclear energy renaissance.

Image source: Getty Images.

Fueling the nuclear renaissanceCentrus Energy supplies nuclear fuel and services to the nuclear power industry. It pioneered the production of High-Assay, Low-Enriched Uranium (HALEU), required for most U.S. reactors to achieve smaller designs (i.e., SMRs). The company's Piketon, Ohio, facility was the first one in the country to receive a license from the U.S. Nuclear Regulatory Commission to produce HALEU.

Several SMR developers, including Oklo, have signed agreements with Centrus to purchase nuclear fuel. Oklo initially signed a letter of intent to purchase fuel for its Aurora Powerhouse Deployment in Ohio earlier this year. The planned 1.2 gigawatt clean energy campus will supply power to Meta Platforms to support its data centers in the region. Oklo and Centrus more recently agreed to pursue a joint venture focused on HALEU deconversion services and advancing related fuel-cycle technologies and supply chains.

Today's Change

(

0.11

%) $

0.16

Current Price

$

139.27

Robust demand is driving real growthOklo and NuScale don't currently generate commercial power, let alone any real revenue or earnings. Centrus Energy, on the other hand, is producing strong, growing financial results. During the second quarter, the company reported $176.1 million of revenue, a 14% year-over-year increase. Meanwhile, it posted $16.8 million of net income during the period.

The company ended the quarter with a $4.5 billion future revenue backlog extending through 2040. That includes $3 billion of HALEU and Low-Enriched Uranium (LEU). It recently signed a $900 million HALEU Enrichment award with the U.S. Department of Energy. Centrus has signed several additional deals since the quarter ended, including a multi-year HALEU contract with Antares Nuclear, a long-term HALEU supply partnership with Radiant, and a HALEU supply deal with X-energy. This growing backlog supports the company's plans to rapidly expand its production capacity to meet future demand, as several deals include pre-payments to help fund the expansion.

Why try to pick an SMR winner when you can invest in the company fueling them allThere's a lot of debate about which nuclear energy stock is the best to buy to capitalize on the country's nuclear renaissance. However, instead of betting on which pre-revenue SMR start-up will win, you can invest in the company that will fuel their reactors. That makes Centrus Energy a lower-risk way to capitalize on the country's nuclear energy resurgence, as it should win no matter which SMR developer emerges as the eventual leader.

Matt DiLallo has positions in Meta Platforms and has the following options: long December 2028 $650 calls on Meta Platforms and short December 2028 $660 calls on Meta Platforms. The Motley Fool has positions in and recommends Meta Platforms. The Motley Fool recommends NuScale Power. The Motley Fool has a disclosure policy.

Amazon just pledged over a billion dollars near its data centers without purchasing a single piece of hardware, and the reason reveals a threat to its $496 billion backlog that no amount of chips can solve.

This post may contain links from our sponsors and affiliates, and Flywheel Publishing may receive compensation for actions taken through them.

Headline Number: $1 Billion for the Neighbors On October 2, 2026, Amazon (NASDAQ:AMZN | AMZN Price Prediction) attached a dollar figure to something no chip plant can make: permission to build. Amazon Web Services committed more than $1 billion over the next five years to the communities around its U.S. data centers. The spending flows through a program called Built Together. The money goes to job training, water preservation and energy affordability. Those are the same three complaints that fill council meetings whenever a new server farm is proposed. None of it buys a GPU.

The pledge comes as local pushback grows. Amazon says more than 100 data center moratoriums are under consideration across the country. AWS chief executive Matt Garman announced the program in a post that compared data center construction to the building of the interstate highway system. He warned that if those bans pass, “the US could be writing its own losing ticket to this race, and the consequences would last generations.”

What It Means for the AWS Buildout Amazon is, in effect, buying community acceptance, and its own economics explain why. Capital expenditures reached $54.21 billion in Q2 alone, up 68.44% year over year. The community pledge is a small line item spread over five years, and its purpose is to protect that much larger capital spending.

Timing is what makes local opposition expensive. Chief Executive Andy Jassy spoke on July 30. On the earnings call, he said “data center capital is spent starting two years before we can put servers into them to start monetizing.” Once a facility opens, Amazon can earn from it for 30-plus years. If a bans stalls a site in its first year, the money spent so far earns nothing while the project waits.

The demand is already committed. AWS backlog was at $496 billion, growing triple digits year over year. Management said most of the AWS capacity planned for 2027 is already reserved. Most AI capacity is signed for at least five-year terms. Each of those contracts needs a building, a grid connection and a town willing to host it.

Residents’ concerns rest on data. After 15 years of nearly flat use, U.S. electricity demand has grown 2.1% per year on average over the last five years. The federal Energy Information Administration lists data center server energy use as a major factor in demand going forward. The average individual data center load doubled from 150 to 300 megawatts between 2023 and 2024. The energy affordability part of Built Together addresses directly to that pressure.

This is also Amazon’s third recent $1 billion commitment that goes to something other than chips. In Q2, Amazon announced a $1 billion investment in AWS Forward Deployed Engineering, with early customers including the NBA, NFL, and Southwest Airlines (NYSE:LUV). It also committed up to $1 billion in cloud credits to help the U.S. Intelligence Community move to the cloud.

How Shares Responded to the $1 Billion Pledge On the day of the announcement, shares closed at $251.52, up 1.33% from $248.23 in a single session. The stock is up 8.97% year to date and down 1.36% over the past month. Recent coverage has focused on competitive concerns, which Barron’s described as “overstated”.

Bull Case For Amazon, the limit on growth is how fast it can add capacity, and each new site needs local approval. AWS revenue reached $42.23 billion in Q2, up 37%, its fastest growth in 18 quarters. The unit’s operating margin was 39.4%. Amazon’s AI and chips businesses each passed $25 billion annual run rates, and each is growing at triple-digit rates. Amazon Bedrock customers spent more in Q2 than all prior quarters combined.

Amazon also has the cash to keep the program going. Operating cash flow reached $45.39 billion in the quarter, up 39.59%. Management remains on track to double its power capacity by the end of 2027 versus 2025. Jassy discussed AWS. He said the unit could “very possibly be a trillion-dollar annual revenue business for us in time.” The main risk remains cash generation: trailing free cash flow is -$7.6 billion, and the pledge adds another cost. Still, keeping sites on schedule protects a $496 billion backlog.

What Amazon’s Next Earnings Report Must Show Built Together turns local approval into a budgeted cost of doing business, much like land or power. Amazon’s next quarterly report will be measured against guidance of $197.0 billion to $202.0 billion in net sales and $22.5 billion to $26.5 billion in operating income. That compares with $17.4 billion a year earlier. Investors should watch whether AWS growth holds up as new sites come online, and if the pledge keeps those sites on schedule, $1 billion in community spending will have protected hundreds of billions in signed revenue.

Contact [email protected] for any questions or corrections.

Amazon is quietly restructuring who actually owns the AI hardware powering its data centers, and the investors taking the other side of that deal may not realize what depreciates fastest in Silicon Valley.

This post may contain links from our sponsors and affiliates, and Flywheel Publishing may receive compensation for actions taken through them.

Amazon (NASDAQ:AMZN | AMZN Price Prediction) is in talks to move $8 billion of NVIDIA (NASDAQ:NVDA) Grace Blackwell chips into a special purpose vehicle, the Financial Times reported. The vehicle would sell debt to outside investors, buy the chips, and lease them back to Amazon. The hardware stays where it is. It is already installed in data centers across at least five states, and Amazon would keep up to 10% of the equity.

Amazon spent $54.2 billion on capital projects in the second quarter, 68% more than a year earlier. Andy Jassy committed to roughly $200 billion in spending for 2026. Free cash flow over the past 12 months turned negative at -$7.6 billion. Long-term debt rose to $119.1 billion from $65.6 billion a year earlier. A leaseback lets Amazon keep running the chips while protecting its credit rating and moving the obligation into lease disclosures.

Investors Would Own Tech’s Fastest-Aging Asset The deal is smart for Amazon and weak for whoever buys the debt. Airlines do sale-leasebacks on planes that fly for decades. Grace Blackwell is already one generation behind Vera Rubin. NVIDIA says Vera Rubin delivers 35X lower token costs than Grace Blackwell Ultra. NVIDIA estimates about $25 billion per gigawatt of revenue for Grace Blackwell and roughly $40 billion for Vera Rubin. Each new launch makes the collateral worth less.

On the July earnings call, Jassy said servers take “a little less than three years” to break even and last at least five to six years. Amazon keeps the early, profitable years. Lenders are left with later years, when a chip two generations old must be refinanced or resold.

NVIDIA Is Building Similar Financing Structures NVIDIA is building similar structures. NVIDIA reported $279 billion in supply commitments and guarantee obligations limited at $108.5 billion for AI cloud and data center partners. It signed up six asset managers and banks to raise more than $500 billion of outside capital for AI infrastructure. Management said: “We know some will call this circular financing.” NVIDIA expects the AI labs it supports to account for roughly a quarter of its business next year.

Demand for computing power remains intense. NVIDIA says it can supply about 70% of what customers want. Having chips that lose value quickly carries risk. Amazon is choosing cash over hardware ownership. Buyers want what the chips produce and are passing on the risk of having them to someone else.

What to Watch in Amazon’s Next Two Earnings Reports Three signals will show whether this becomes the template. First, whether Amazon’s third-quarter earnings report reveals the vehicle and how much its lease obligations grow. Amazon has guided to $22.5 billion to $26.5 billion in operating income for the quarter. Second, the interest rate investors demand on the debt.

A wide spread would mean lenders are pricing in an aging risk that stock investors have ignored. Third, whether other cloud giants announce similar deals within two quarters. Should they do so, more of AI’s hardware bill will sit off the balance sheets investors read, and lenders will be left holding chips that NVIDIA is working to make obsolete.

Contact [email protected] for any questions or corrections.

McDonald's Today

$232.06 +0.23 (+0.10%)

As of 10/2/2026 03:59 PM Eastern

This is a fair market value price provided by Massive. Learn more.

$229.61▼

$341.753.21%

18.85

$300.64

September delivered mixed signals for McDonald’s NYSE: MCD shareholders, including a landmark dividend achievement, before the stock opened October by falling to a fresh 52-week low.

On Sept. 17, the board of the fast food restaurant chain declared a quarterly cash dividend of $1.93 per share, marking a 4% increase and the company’s 50th consecutive year of raising its payout.

Get McDonald's alerts:

According to Executive Vice President and Global Chief Financial Officer Ian Borden, the “milestone is the result of decades-long commitment to maintaining [McDonald’s] financial discipline and rewarding [its] shareholders."

But lately, those shareholders haven’t been feeling all that rewarded. After joining the elite Dividend Kings club, the stock hit a fresh 52-week low of $229.61 on Oct. 1, reaching prices not seen since September 2022.

Now, the nearly $164 billion market cap company is hoping a multi-year strategic plan can win back consumers and serve as the catalyst to restore investor confidence.

McDonald’s > NEXT: The Fast Food Giant’s Comeback PlanThe fast food industry’s recent struggles have been well-documented. In August, Restaurant Business Magazine’s Jonathan Maze wrote that while McDonald’s same-store sales increased 0.8% in Q2, traffic to its restaurants declined.

Affordability continues to be the main challenge. Maze noted that “McDonald’s lost sales from lower-income consumers totaled $310 million last quarter alone.”

That, among other factors, has tempered growth expectations for the industry at large. According to consultancy firm Grand View Research, the global fast food market is forecast to undergo a 5% compound annual growth rate (CAGR) between 2022 and 2029.

Domestically, the outlook is worse. The U.S. fast food market is expected to see a CAGR of just 3.9% between 2022 and 2028.

McDonald’s is already planning to address its own growth and productivity challenges. On June 1, the company launched a new multiyear strategic plan called McDonald's > NEXT, which it hopes will usher in its next era of productivity and growth.

The $8.5 billion plan calls for revamping employee training programs for the chain’s approximately 2 million workers, investing billions into upgrading franchisees’ equipment and technology—including advanced back-of-house operational systems and AI-enabled tools—improving dining room spaces, and aiming for higher food quality.

Of that dollar amount, McDonald’s has earmarked $5 billion to be spent through 2030 on a combination of rent relief and capital support. In his Q2 earnings call comments, CEO Chris Kempczinski said the new growth strategy has a singular objective: to make McDonald’s the first choice for more customers, more often.

Despite Ambitious Plans, McDonald’s Still Faces Structural ChallengesOn Sept. 23 at its investor day, the company laid out plans for two connected tools that are critical components of its McDonald’s > NEXT initiative.

The first is a pricing engine that combines transactional data and local market data to make restaurant- and item-level price recommendations while ultimately leaving pricing decisions up to franchisees.

The second is a customer-engagement engine that will personalize experiences through offers and rewards to encourage its existing 220 million 90-day active loyalty members to visit more frequently and upgrade to higher-priced items.

McDonald’s hopes that the initiative will result in a 1.5% gain in market share in both the chicken and beverage categories by 2030, while maintaining its leadership position in beef and expanding operating margins to the low-to-mid 50% range by 2030, with a 2.5% increase in gross restaurant-level efficiency. The company currently enjoys high margins, and how much its new strategic plan can increase them is yet to be determined.

Another perhaps more pressing issue is the aforementioned disconnect between pricing and lower-income customers. AI-informed pricing recommendations alone will not sufficiently address record-high ground beef prices, ongoing inflation, and the subsequent fallout for fast food chains.

After decades of establishing brand loyalty through value menu offerings, McDonald’s has been trying to rebuild its value proposition as lower-income consumers pull back. The cost of a Big Mac in 2000 was approximately $2.24, or $4.27 in today’s dollars. Yet the average price of a Big Mac in 2026 is closer to $7, while a Big Mac combo meal with fries and a drink can run as high as $18.

In doing so, McDonald’s has been able to squeeze more profit per transaction from middle- and higher-income consumers who have been willing to absorb those price hikes, while abandoning the lower-income staple of its customer base.

McDonald’s Needs a Fix, and AI Might Not Be the AnswerIf McDonald’s > NEXT is the company’s answer to its recent troubles, it has some heavy lifting to do. In Q2, revenue growth slowed to 3.74% after registering 9.42% in Q1 and 9.71% in Q4 2025. Meanwhile, net income growth slowed to 4.84%.

98th Percentile

Moderate Buy

29.6% Upside

Healthy

Strong

0.24 N/A

6.54%

See Full Analysis

McDonald’s also pushed its target of reaching 50,000 global locations to 2028 from 2027, attributing the delay to inflationary development costs and a pressured consumer environment.

The company’s financial health, as measured by TradeSmith, has been in the red zone for over five months, and while the stock’s forward price-to-earnings ratio of 17.99 may suggest it is undervalued to some investors, recent price action remains bearish.

Short interest stands at just 1.73% of the float, but in dollar terms, that equates to $3.08 billion worth of MCD, which remains near multi-year highs. At the same time, corporate insider buying has dried up, with no insider purchases reported over the past 12 months. Institutional activity, however, has remained net positive. Institutional buying totaled $16.40 billion from 2,170 buyers over the trailing 12 months, compared with $8.78 billion in outflows from 1,741 sellers, while insider buying stood at zero over that period. Shares are down around 24% year to date, including a more than 32% slide from their all-time intraday high of $341.75 on March 2.

Should You Invest $1,000 in McDonald's Right Now?Before you consider McDonald's, you'll want to hear this.

MarketBeat keeps track of Wall Street's top-rated and best performing research analysts and the stocks they recommend to their clients on a daily basis. MarketBeat has identified the five stocks that top analysts are quietly whispering to their clients to buy now before the broader market catches on... and McDonald's wasn't on the list.

While McDonald's currently has a Moderate Buy rating among analysts, top-rated analysts believe these five stocks are better buys.

View The Five Stocks Here

The space race is growing fast, and you don’t have to have gotten in early on SpaceX to profit. This report shows seven space stocks you can buy today that may grow as rockets, satellites, defense, space internet, and new space technology become more important.

Get This Free Report

Continue following MarketBeat

Add MarketBeat as your preferred source on Google to see our latest stories in your feed.

SummaryMicron generated $54.2 billion in Q4 revenue, an 87% gross margin, and $33.2 billion of free cash flow.Twenty-six strategic agreements now carry $32 billion of customer commitments, with $12.7 billion of deposits already received.Customer deposits effectively help finance Micron’s expansion, while government incentives further reduce the capital shareholders must ultimately fund.MU trades near 6.1x FY1 non-GAAP earnings despite $150 billion RPO and supply constraints expected through 2028. jxfzsy/iStock via Getty Images

The market is still treating Micron Technology (MU) on the premise of how sustainable its extraordinary memory pricing will prove to be. I'm less inclined to believe that this approach to the stock is most

2.37K Followers

Analyst’s Disclosure: I/we have a beneficial long position in the shares of MU either through stock ownership, options, or other derivatives. 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.

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.

Broadcom just agreed to lend billions to the same AI lab it sells chips to, lease equipment to, and may soon own a stake in. That four-way entanglement raises a question shareholders cannot yet answer: whose demand is actually driving…

This post may contain links from our sponsors and affiliates, and Flywheel Publishing may receive compensation for actions taken through them.

Broadcom’s $42 Billion Number Makes the Supplier the Lender Broadcom (NASDAQ:AVGO | AVGO Price Prediction) has agreed to lend Anthropic up to $42 billion, according to the AI lab’s IPO filing, as reported by Reuters on October 1, 2026. The money pays for Anthropic’s infrastructure spending, and much of that spending goes to Broadcom hardware. The debt can also convert into Anthropic shares. Broadcom now plays four roles in this relationship: chip supplier, equipment lessor, lender, and possible future shareholder.

What $42 Billion Shows About Broadcom’s AI Growth The facility could cover about one-third of Anthropic’s $125.2 billion five-year TPU compute commitment. That commitment supports Broadcom’s AI plan. On the September earnings call, management said it expects Anthropic to become its largest XPU customer in 2027 and to hold that spot in 2028. Broadcom is installing one gigawatt of Ironwood chips for Anthropic in 2026. Anthropic is expected to installing five gigawatts of TPU v8i chips in 2027, and Broadcom says it has “clear line of sight” to an extra 10 gigawatts in 2028.

Those installations support Broadcom’s AI revenue outlook of about $115 billion in fiscal 2027 and $230 billion in fiscal 2028. The $42 billion facility tops the $23.975 billion in cash Broadcom held at the end of its third quarter.

The loan adds to earlier financing support. In June, Broadcom set up a financing platform with Apollo Global Management (NYSE:APO) and Blackstone (NYSE:BX). Its first $35 billion tranche closed that month to pay for Anthropic’s one-gigawatt installations. Management said at the time that third parties would insure those assets and that Broadcom would provide no direct financing, though it might offer “modest residual value guarantees.” Broadcom lending directly to Anthropic is a much deeper commitment.

How Broadcom Stock Reacted to the News Barron’s reported on October 1 that Broadcom stock rose as the two companies became more closely tied. On the next trading day, October 2, shares gained 3.35%, from $343.64 to $355.14. CNBC listed Broadcom among that day’s biggest noon movers.

Over a longer stretch, the stock has gone almost nowhere. Shares changed -0.39% from September 3 through October 2, and they remain well below the 52-week high of $493.28.

Bear Case: Broadcom Is Financing Its Own Demand Broadcom’s operating results are strong. Third-quarter revenue reached $29.59 billion, up 85.5% from a year earlier, and free cash flow was $13.665 billion. The concern is how much of the future revenue depends on Broadcom’s own credit.

Anthropic’s filing flags the conflict. Anthropic itself warns that Broadcom’s two roles as supplier and lender create “potential conflicts of interest” involving pricing and hardware decisions. Default terms work against both sides. Anthropic put cash into a restricted account for Broadcom’s benefit in April 2026 and may have to add more. If Anthropic misses payments or other obligations, much of what it owes on its leases could come due at once, and its access to the loan could be limited at the same time. Exposure is hard to measure. On the call, management would not give a maximum figure for future guarantees. It said future deals would have “unique features” tailored to each lab. An analyst noted the first tranche carried maximum exposure of about $29 billion. The customer list is short. Broadcom has only six XPU customers, and its filings list customer concentration and “significant indebtedness” as risks. Gross fixed-rate debt stands at $59.6 billion. Reuters described the deal as part of a broader industry pattern, comparing it with NVIDIA (NASDAQ:NVDA) using its balance sheet to support chip sales. Shareholders end up carrying credit risk tied to a single AI lab, while the stock still trades at 45x trailing earnings.

Bottom Line: Anthropic’s IPO Is the Key Test The $42 billion figure shows how much of Broadcom’s path to $230 billion in fiscal 2028 AI revenue depends on one customer that Broadcom is also helping to pay for. The next major event is Anthropic’s IPO. Management said the listing could change Anthropic’s credit profile, and it would also show what Broadcom’s possible equity stake is worth. Before that, Broadcom has guided fourth-quarter revenue to about $34.80 billion and AI semiconductor revenue to $21.70 billion. Long-term holders will need to judge whether that growth comes from customers’ own spending or from Broadcom’s lending.

Contact [email protected] for any questions or corrections.

Dutch Bros (BROS +0.15%) stock fell 21% in September, according to data provided by S&P Global Market Intelligence. The company is still feeling negative market sentiment after some disappointing earnings updates in August.

Growing painsDutch Bros is a relatively small coffee shop chain that has ambitious expansion plans. It owns 1,177 stores as of the end of the second quarter, and it plans to reach 2,029 stores by 2029, nearly doubling today's count.

Image source: Dutch Bros.

It consistently reports robust growth, and the second quarter was fantastic, which you probably wouldn't realize based on the market's reaction. The company is still growing fast and becoming more profitable, and it's generating consumer loyalty as it expands across the country.

Revenue increased an impressive 32.5% to $550.9 million in the quarter, and company-owned same-shop sales were up 8.3%. Company-owned same-shop transactions were up 3.4%, indicating that the increase didn't come entirely from price hikes. Systemwide metrics were somewhat lower, at 5.8% and 1.7%, respectively, and Dutch Bros is transitioning from its original franchise model to new company-owned stores only.

So what didn't the market like? There's concern about rising costs, which is why negative market sentiment persisted into September, with inflation remaining strong.

The market also didn't like management's guidance. Although it slightly raised full-year guidance for total revenue and same-store sales, it significantly raised guidance for capital expenditures (capex), from a midpoint of $280 million to a midpoint of $360 million. Management explained that average capex per new store is still $1.4 million, and the higher total reflects aggressive expansion.

After the earnings release, the company announced that a deal to acquire a small restaurant chain that would have made it easy to add 65 new stores fell through, feeding into the negativity about whether Dutch Bros would be able to expand as fast as it's aiming for.

An opportunity to buy on the dip?Dutch Bros has short-term plans to double store count and long-term plans to reach 7,000 stores. These stores are attracting fans and sales, and it's easy to see how it could reward investors despite the short-term drop.

Premium Feature

Moneyball Superscore

78/100

Today's Change

(

0.15

%) $

0.06

Current Price

$

38.67

Dutch Bros' P/E ratio has fallen dramatically along with the price, and it trades at 54 times trailing 12-month sales. That's still expensive, and it demonstrates how much the market expects from it.

The good news is that it's the cheapest P/E ratio ever. The bad news is, it's rich enough that the stock can still fall on any missteps.

I think Dutch Bros is an excellent stock with tons of long-term potential. If you can buy now and hold for at least five years, it's a good time to buy.

Canary’s amended S-1 provides new operational details for its proposed spot PEPE ETF. The trust would hold PEPE directly, with BitGo providing custody and Cboe BZX as the proposed exchange. PEPE’s 10 largest wallets held about 41% of circulating supply in January, although several are exchange addresses. The filing advances the product but leaves its ticker, launch timing and eventual investor demand unresolved. Canary Capital is moving ahead with one of the more unusual experiments in the U.S. crypto ETF market: putting PEPE inside a conventional exchange-traded fund.

The asset manager filed Pre-Effective Amendment No. 1 to the Canary PEPE ETF registration statement on Oct. 2, adding operational details to its original April filing. The proposed trust would hold PEPE directly and seek to reflect the token’s market price after expenses.

PEPE was trading near $0.000004272 on Oct. 4, down about 2.33% over seven days. The filing therefore comes without the immediate price surge that often accompanies speculative crypto announcements, putting more attention on the product itself and whether brokerage access can create a new source of demand.

Bloomberg ETF analyst Eric Balchunas viewed the amendment as another indication that issuers are becoming more comfortable testing the boundaries of the crypto ETF market.

Canary just amended their Pepe ETF filing… arguably another sign crypto winter has ended. You just weren’t going to even think about launching a Pepe ETF a few months ago. Feels safer to test market now. https://t.co/7UBxit0m2y

— Eric Balchunas (@EricBalchunas) October 2, 2026

His observation points to how far product development has moved: a memecoin ETF that might previously have looked commercially implausible is receiving enough attention for Canary to continue building out the registration.

Canary Defines the Infrastructure Behind a PEPE ETF The amended prospectus provides a clearer picture of how PEPE would fit inside a traditional fund structure.
Canary intends to hold PEPE directly, rather than use futures or other derivatives to replicate its performance.

Shares are proposed for listing on Cboe BZX, while the fund would calculate its net asset value using the CoinDesk PEPE Benchmark Rate 60m NY Rate, a time-weighted benchmark derived from eligible PEPE trading venues.

The institutional infrastructure is also taking shape. BitGo Bank & Trust is named as crypto custodian, U.S. Bank would custody cash, U.S. Bancorp Fund Services would act as transfer agent and CSC Delaware Trust Company would serve as trustee.

Creation and redemption would occur in 10,000-share baskets, with the prospectus allowing authorized participants to transact using cash or PEPE.

One provision is specific to PEPE’s blockchain infrastructure. Because the token operates on Ethereum, the trust may initially hold up to 5% of its assets in ETH to cover network transaction fees. The ETH would be used operationally rather than as a separate investment position. After that reserve is depleted, the sponsor would pay network fees and seek reimbursement from the trust.

It is a small part of the product, but one that exposes the mechanics involved in placing an ERC-20 token inside an ETF. Investors may interact with ordinary brokerage infrastructure, while movement of the underlying PEPE still requires Ethereum transactions.

PEPE’s Largest Wallets Hold 41% of Supply The filing also provides information that matters beyond the ETF structure itself.

According to Canary, the 10 largest PEPE wallet addresses controlled approximately 41% of circulating supply as of January 2026. That figure initially suggests heavy concentration, but it needs qualification because several of the largest addresses belong to centralized exchanges holding assets for customers.

An exchange omnibus wallet can represent thousands of beneficial owners, making it different from a single whale controlling the same amount of PEPE.

Concentration nevertheless remains relevant. Large movements from major wallets can affect available liquidity, particularly when tokens are transferred onto exchanges.

PEPE also differs from many venture-backed crypto assets because it does not have a conventional vesting schedule creating predictable waves of future token unlocks. Substantially all of its supply is already circulating, shifting attention from future issuance toward ownership distribution and existing holder behavior.

Canary’s prospectus is similarly direct about the token’s underlying demand. PEPE has limited transactional utility, while its market value is closely connected with the Pepe the Frog meme, community engagement and online sentiment.

The ETF can change how investors obtain exposure to those characteristics. It cannot remove them.

The Wrapper Does Not Change PEPE’s Risk Profile That distinction is important when comparing the proposal with established crypto ETFs.

Bitcoin has increasingly entered portfolios through regulated funds, corporate treasury holdings and macro investment strategies. PEPE remains much more dependent on speculative demand, trading liquidity and cultural relevance.

Canary is not proposing an active strategy to offset those risks.

The trust would not sell because PEPE appears expensive, accumulate because the token looks undervalued or hedge its exposure during a downturn. Investors would effectively receive PEPE’s market performance after the fund’s expenses and liabilities.

The prospectus accordingly warns that shareholders could lose all or substantially all of their investment.

For some investors, an ETF could still solve a practical problem. Exposure would become available through a conventional securities account without independently maintaining an Ethereum wallet, safeguarding private keys or trading directly on a crypto exchange.

That is an access improvement, not a change in the economics of the underlying asset.

The SEC Has Not Approved the PEPE ETF The Oct. 2 amendment also needs to be separated from an approval decision.

Canary filed a pre-effective amendment, meaning the registration statement remains part of an ongoing process. The prospectus states that the securities cannot be sold until the registration becomes effective.

Several pieces of the proposed fund are now defined, including custody, pricing and creation-redemption mechanics. Others remain unresolved. Most visibly, the ticker field is still blank, and the filing does not establish a confirmed launch date.

That makes the amendment relevant without turning it into an approval signal.

Canary is spending resources to advance the product and has assembled much of the infrastructure necessary to operate it. Whether the fund ultimately reaches investors depends on what happens next in the registration and listing process.

PEPE ETF Faces a Different Test Than Bitcoin Regulatory progress would still leave Canary with a commercial question.

Bitcoin demonstrated that substantial demand existed for crypto exposure through U.S.-listed ETFs. That result does not mean the same demand exists for every digital asset with a large trading community.

Memecoins are particularly difficult to assess.

PEPE can generate significant spot-market activity while remaining readily available to crypto-native traders through existing exchanges. An ETF therefore has to attract investors who want exposure to the token but prefer, or require, a traditional brokerage structure.

That is a narrower proposition than simply proving that PEPE has an active market.

The current price adds useful context. At roughly $0.00000427, PEPE remains a highly sentiment-sensitive asset, and wrapping it inside an ETF would not create cash flows, alter its token supply or reduce its dependence on market demand.

Canary has now provided much of the plumbing needed to test whether that distinction matters to investors. The next meaningful milestones are more concrete: an effective registration, final ticker and fee, a trading date and, if the fund launches, its initial creations and assets under management.

Those numbers would show whether the expansion of crypto ETFs into memecoins is producing a new investment market or simply a broader menu of products.

Alexander Zdravkov is a crypto journalist and market analyst specializing in digital assets, blockchain technology, and the rapidly evolving cryptocurrency market. He entered the industry more than four years ago, drawn to the speed of innovation in crypto and the potential of blockchain-based financial systems to reshape traditional markets. His work focuses on understanding the forces behind crypto market movements, from Bitcoin and altcoin cycles to institutional adoption, regulation, exchange-traded products, stablecoins, DeFi, and emerging blockchain ecosystems. Rather than simply following price action, Alexander looks at liquidity, investor positioning, market structure, and macroeconomic conditions to identify developing trends and explain what they could mean for digital assets. With a degree in International Relations, he brings a broader perspective to cryptocurrency coverage by connecting developments in the sector with monetary policy, geopolitics, regulation, and global capital flows. This allows him to examine crypto not as an isolated market, but as an increasingly interconnected part of the global financial system. Over the course of more than five years in financial journalism, Alexander has written over 5,000 articles covering cryptocurrencies, blockchain, traditional markets, and the global economy. His reporting spans Bitcoin, Ethereum and altcoins, DeFi protocols, institutional crypto adoption, regulation, market infrastructure, security incidents, and major developments shaping the wider digital asset industry. His approach is centered on giving readers context behind the numbers and headlines, explaining not only what is happening in crypto markets, but why it matters and what could come next.

Full Profile

Hyperliquid is transforming Circle’s reserve yield into a direct source of HYPE buyback demand. This transformation will change the way the protocol’s revenue returns to its token holders.

In essence, this agreement directs all of the AQA yield generated by an existing $6.07 billion treasury to be returned directly to Hyperliquid.

Using a 3% interest assumption, this results in approximately $498,580 in daily purchase power. Assuming each Hype token costs $88, at the above rate the flows can burn approximately 5,684 tokens per day or increase the burn by 25% relative to its current average.

Source: X Furthermore, because the mechanism has been designed to scale directly with the size of Circle’s reserves, it allows the purchase power for burns to increase as more assets are deposited.

Therefore, Hyperliquid has created a new source of recurring demand for HYPE. In doing so, it strengthens HYPE’s overall tokenomics, creating greater opportunity for burning HYPE.

AQAv2 yield begins HYPE buybacks That proposed buyback mechanism is transitioning from conceptual to implementation. Hyperliquid’s System Interest Address has confirmed receipt of the first AQAv2 Reserve Payment, thus proving actual revenue flowing into the system.

Of the $14.58 million in USD Coin [USDC] generated from this source, only $2 has currently entered the Assistance Fund. However, there is still $14,580,777.21 in USDC awaiting to be transferred from the reserve account to the assistance fund.

Source: Hyperliquid Ecosystem The significance of this distinction lies within the fact that the reserve yields have been successfully recorded. Therefore, once these funds enter the Assistance Fund, they may then be converted by Hyperliquid into buyback activities.

Hence, the pending transfer is what needs to occur in order for the reserve income to become realized HYPE deflation.

Tate moves HYPE to Binance Meanwhile, Tate’s latest move has shifted focus off of Hyperliqd’s buybacks towards possible sources of HYPE liquidity. He transferred 20,950 HYPE, worth approximately $1.87 million, to Binance.

Despite that move, this action in itself doesn’t confirm that the HYPE was sold. Therefore, the remaining 63,550 HYPE will likely gain even greater importance. This remaining balance is approximately $5.62 million and is still out of exchange.

Thus, there are many options left open with regard to future deposit possibilities should selling become an attractive option.

Source: Lookonchain According to Lookonchain data, Tate realized over $7.24 million in profit with the sale of HYPE. This is indicative of the size of profits that exist at this time for people who have purchased HYPE at levels below $4.48.

Thus, future exchange transactions may counteract some of the demand created by Hyperliquid’s stated intention to repurchase HYPE. In turn, this would increase sell-side pressure, making wallet movements increasingly important for HYPE’s price going forward.

Final Summary

The International Monetary Fund approved the immediate disbursement of about $138 million to El Salvador under the country’s $1.4 billion financing program, despite some performance criteria being missed.

The IMF’s Executive Board completed the second and third reviews under El Salvador’s 40-month Extended Fund Facility (EFF) arrangement on Thursday, it said in a press release.

While certain performance criteria were not met, including on the Bitcoin accumulation front, the IMF granted waivers “based on strong corrective measures and renewed commitments.”

The agency noted that El Salvador has made progress in financial sector reforms, fiscal transparency, AML/CFT reforms and the transfer of majority ownership and control of the government Chivo Bitcoin wallet to a private operator.

“Efforts will continue to reduce the state’s involvement in Bitcoin-related activities, strengthen crypto‑asset regulation and governance, and enhance transparency regarding public-sector crypto‑asset holdings,” the IMF said, adding: “No further Bitcoin accumulation is envisaged beyond the documented donations.”

Cointelegraph reported on Sept. 4 that El Salvador used no public resources to accumulate Bitcoin after the first review of its IMF financing program in June 2025, according to the lender. 

The IMF said documents supplied by Salvadoran authorities verified that the accumulation came from private donations. According to the lender, the increase in El Salvador’s holdings therefore did not reflect additional Bitcoin purchases financed with government resources. 

The IMF also said majority ownership and operational control of the Chivo wallet had been transferred to a private operator, while the government retained a minority stake and custodial responsibilities.

The explanation addressed how El Salvador’s holdings grew after the first review, when the country said in November 2025 that it had acquired 1,090 BTC worth $100 million, renewing questions about compliance with its $1.4 billion IMF program. 

Magazine: Exchanges reporting crypto gains to IRS becomes tax nightmare

Cointelegraph is committed to independent, transparent journalism. This news article is produced in accordance with Cointelegraph’s Editorial Policy and aims to provide accurate and timely information. Readers are encouraged to verify information independently.

Ripple, a leading blockchain company focused on enterprise crypto solutions, has introduced selective confidentiality directly into the XRP Ledger (XRPL) protocol. This development aims to address privacy concerns for institutional users without relying on a separate or private version of XRPL. Instead, Ripple is embedding confidentiality at the core protocol level to combine public transaction settlement with private financial information.

Selective confidentiality on the public ledgerThe official documentation for XRPL’s Confidential Transfers outlines the use of EC-ElGamal encryption and zero-knowledge proofs as foundational technologies. These cryptographic methods help conceal Multi-Purpose Token balances and transfer amounts from being visible on the public blockchain, while still allowing verifiers to confirm transaction validity.

With this structure, validators can approve that transactions meet system requirements without accessing the underlying transaction amounts. At the same time, issuers have the capability to designate specific auditors who can decrypt balances for compliance and regulatory checks, ensuring responsible oversight in financial operations.

This balance provides banks and regulated institutions with a model that preserves the transparency benefits of blockchain—such as public settlement—while maintaining privacy over sensitive financial details.

Mini dictionary: EC-ElGamal encryption is a public-key cryptosystem considered secure for both confidentiality and supporting zero-knowledge proofs; zero-knowledge proofs allow validators to confirm the validity of a statement without revealing any other information.

Institutional features and regulatory complianceThe XRPL 3.3.0 protocol upgrade introduced ConfidentialTransfer for amendment voting, alongside other new features aimed at institutional requirements such as Batch processing and Permission Delegation. These updates reinforce XRPL’s strategic push toward serving regulated entities and facilitating compliance in decentralized finance frameworks.

Developers are also testing permissioned liquidity pools that limit Automated Market Maker (AMM) participation to credentialed users. When paired with features such as Credentials, Permissioned Domains, and confidential token transfers, the system is increasingly positioned as regulated financial infrastructure—distinct from anonymous open DeFi models but still built on a public blockchain foundation.

Potential use cases for confidential transfersRipple has highlighted confidential collateral as a key use case for institutions. In this model, a financial entity can provide proof that sufficient collateral exists to meet regulatory requirements without revealing the actual size of the client’s position, trading counterparties, or specific strategies.

Ripple’s institutional DeFi roadmap references applications including proof of reserves, private settlements, and compliance verification, all of which may benefit from integrating zero-knowledge technology into public blockchain systems.

This approach underscores Ripple’s emphasis on compliance with privacy, aligning with growing demand from banks and regulated firms seeking both transparency and confidentiality in their blockchain transactions.

Most chipmakers struggle to sustain dividends through brutal industry downturns, but three semiconductor companies have quietly built the cash flow machines and balance sheet discipline to keep raising payouts no matter where the cycle goes.

This post may contain links from our sponsors and affiliates, and Flywheel Publishing may receive compensation for actions taken through them.

Chipmakers have a reputation for wild earnings swings and huge factory bills, both of which can strain a steady dividend. That makes the few semiconductor companies that keep raising their payouts worth a closer look, because every check they mail says the business throws off cash through the whole cycle. The clearest example came when Texas Instruments (NASDAQ:TXN | TXN Price Prediction) increased its dividend 7% to $1.52 per share, marking 23 consecutive years of increases. The three names below are Broadcom (NASDAQ:AVGO), Texas Instruments and Qualcomm (NASDAQ:QCOM). Their yields are modest. Their cash flows are what make them notable for income investors, and each one has a different kind of cycle risk.

Broadcom: Record Free Cash Flow Dwarfs the Dividend Bill Broadcom has the lowest yield in this group at 0.76%, based on an annualized dividend of $2.60 per share and a recent price of $355.14. The dividend is lowest next to the share price, but it is very well covered.

Dividend safety: Broadcom produced $13.7 billion of free cash flow in its fiscal third quarter, equal to 46% of revenue. It paid $3.1 billion in dividends over the same period. Measured against trailing earnings per share of $7.83, the annual dividend takes about 33% of profits. The balance sheet is moving the right way. Cash ended the quarter at $24 billion, management repaid $5.6 billion of long-term debt during the quarter, and it retired another $1.5 billion of notes after the quarter closed. The remaining $59.6 billion of fixed-rate debt has an average coupon of 4%.

Track record: Broadcom has paid a dividend every quarter since December 2010, and the amount has stepped up many times. The two most recent increases took the quarterly payout from $0.53 to $0.59 to $0.65. The last raise was declared on December 9, 2025, so income investors should keep an eye on the fiscal fourth-quarter earnings report scheduled for December 9, 2026.

Bull case: Broadcom designs custom AI accelerators for hyperscalers including Google, Meta and OpenAI, sells the Ethernet switches that connect them, and runs an infrastructure software business with a 94% gross margin. AI semiconductor revenue reached $16.7 billion in the quarter, up 221% year over year. Management guided fourth-quarter revenue to about $34.8 billion and expects fiscal 2027 AI revenue of around $115 billion (that kind of early-mover setup is exactly what we studied in our free Next Nvidia playbook). Hock Tan, the chief executive on the call, put it this way: “Q3 demand was simply hot and we’re just getting started.” The stock trades at about 19 times forward earnings, so investors get that growth plus a growing payout.

Risk: This is a cycle tied to a handful of customers. Broadcom’s AI outlook depends on six XPU customers, and management said installations still depend on land, power and data center shells being ready. Should leading AI labs pull back on spending, Broadcom’s cash growth could stall quickly, even though the dividend would still be covered by a wide margin.

Texas Instruments: 23 Straight Years of Raises and a Cash Flow Rebound Texas Instruments yields about 2.07%, based on its new annualized dividend of $6.08 and its stock trading at $293.80. The newly raised $1.52 quarterly payment goes out on November 10 to shareholders of record on October 30.

Dividend safety: TI has the best dividend-growth record of the three. The quarterly payout went from $1.24 in 2023 to $1.30, $1.36, $1.42 and now $1.52, with one raise each autumn. Coverage is lower than at Broadcom. Trailing dividends of $5.62 per share against trailing EPS of $6.58 work out to a payout ratio near 85%. Free cash flow is what has changed. Trailing 12-month free cash flow reached $6.5 billion, up from $1.8 billion a year earlier, and TI returned $5.8 billion to owners over that stretch. The balance sheet holds $7 billion of cash and short-term investments against $14 billion of debt.

Bull case: TI makes analog and embedded processing chips. These are low-cost, long-lived parts that go into factory equipment, cars and power systems. Second-quarter revenue rose 22.82% to $5.463 billion, and data center revenue doubled year over year. For income investors, capital spending matters most. Quarterly capex fell to $514 million from $1,305 million a year earlier now that the heavy 300mm factory work is done, and that freed-up cash can go to shareholders. Management guided third-quarter revenue to $5.65 billion to $6.15 billion. The stock is up 72.21% year to date, which tells you investors already expect the recovery to last.

Risk: If the cycle turns, there is little buffer. The around 85% payout ratio leaves little room if industrial demand fades, and trailing free cash flow includes $1.6 billion of CHIPS Act incentives that will not repeat forever. Management also plans to fund the Silicon Labs acquisition with cash and new debt. A downturn would test how far that 23-year run can stretch. At about 28 times forward earnings, investors are paying up for the recovery.

Qualcomm: Licensing Cash and a Massive Buyback Back Up the Payout Qualcomm yields 1.93% on an annualized dividend of $3.68, with the latest quarterly check at $0.92. Its free cash flow yield is 6.49%, which means the dividend uses only about 30% of the free cash flow the business produces.

Dividend safety: The balance sheet is conservative. Debt to equity is 0.77, net debt is just 0.61 times EBITDA, with interest coverage at 18.61x. Qualcomm returned $2.3 billion to shareholders in its fiscal third quarter, including $1.4 billion of buybacks, and it has a $20 billion repurchase authorization in place. Because so much of the payout goes to buybacks, management could slow repurchases in a downturn and leave the dividend alone. The record shows a raise every year from 2020 through 2026, most recently from $0.89 to $0.92. Over the full history, the quarterly payout has grown from $0.05 in 2003.

Bull case: Qualcomm sells Snapdragon processors for phones, cars and connected devices, and it takes in patent royalties through its QTL licensing arm. That licensing business posted a 69% EBT margin, a high-margin source of cash that does not depend on building chips. The company also renewed its global patent license agreement with Apple in September. Automotive revenue rose 61% to $1.588 billion, its 23rd straight quarter of double-digit growth. Management is targeting $40B in non-handset revenue by fiscal 2029, helped by data center work that includes a new multi-generational collaboration with Amazon.

Risk: Qualcomm faces a handset cycle hitting at the same time as rising costs. Handset revenue fell 20% to $5.086 billion, and net income dropped 24.91% as memory, wafer and packaging costs rose. Management expects Apple product revenue to fall about 50% from the September quarter to the December quarter, and it guided chip-segment margins down to 23% to 25%. Non-GAAP EPS of $2.21 missed the $2.22 consensus estimate. Earnings coverage is weaker than cash coverage right now, with the stock at about 36 times trailing earnings.

Three Chipmakers, Three Ways to Get Paid Through the Cycle Broadcom has the widest coverage and the lowest yield, and AI cash flow is funding both the dividend and debt paydown. Texas Instruments has the strongest raise record and the highest yield of the three, but also the thinnest payout buffer, so its upcoming third-quarter report matters most for the dividend. Qualcomm sits in the middle: free cash flow covers the dividend comfortably, licensing royalties add a steady stream of cash, and a big buyback can be cut back before the dividend is touched. For income investors who want chip exposure, all three dividends look well supported by cash flow at this point in the cycle.

Contact [email protected] for any questions or corrections.

TLDR: The Cronos token burn removes 228 million CRO from the community pool. Community approval lifts this program to 428 million tokens.  All Ult and Cronos Launch revenue will fund CRO buybacks and monthly burns. Published transaction hashes allow public verification.  Staking rewards keep their current terms, with Strategic Reserve support maintaining payouts as emissions decline under existing policy.  CRO trades near $0.0669 after a roughly 1.56% daily gain. Future purchase amounts depend on product revenue and execution costs.  The Cronos token burn removed 228 million CRO from the community pool after voters approved two governance proposals. Cronos Network announced the decision on October 3, bringing burns under the community program to 428 million tokens. 

CRO traded near $0.0669, gaining approximately 1.56% over 24 hours, according to Coingecko market data. The destroyed tokens carried an estimated market value of about $15 million. 

The approved framework directs all revenue from Ult and Cronos Launch toward open market purchases and monthly burns. Cronos will publish transaction hashes, while staking rewards retain their existing terms and funding support through the Strategic Reserve. 

Cronos (CRO) Price Cronos token burn lifts community total to 428 million Proposal 36 authorized the latest community pool transfer to a burn address on Cronos POS. Four previous rounds each removed 50 million CRO, making this fifth round considerably larger. 

The Cronos token burn therefore exceeded the combined size of those earlier rounds by 28 million tokens. The 428 million total describes this community initiative, rather than every historical destruction of CRO.

The immediate operation used tokens already held by the community pool. It did not require purchasing the entire 228 million CRO from exchanges, an important distinction when assessing market demand. 

Future CRO buybacks introduce a separate source of purchases because they use product revenue to acquire tokens. Their size will vary with platform earnings and the market price when transactions execute.

Although the Cronos token burn reduces available supply, it does not establish a guaranteed price increase. Liquidity, demand, reserve distributions, and broader market conditions still influence how CRO trades.

Coingecko listed daily trading volume near $5.82 million and market capitalization around $3.31 billion. These figures show that the burn announcement arrived alongside relatively limited turnover compared with the token valuation. 

That leaves a practical distinction between governance approval and sustained trading interest. A smaller token supply alone cannot reveal whether demand will grow enough to support a lasting advance.

CRO Buybacks Tie Monthly Burns to Revenue From Trading Proposal 37 commits 100% of Ult and Cronos Launch revenue to token purchases and destruction. Cronos Labs says existing capital will cover operations, infrastructure, and growth spending instead of those product receipts. 

Both tokenomics proposals have passed.

228,000,000 CRO has been burned from the community pool, bringing the total to 428,000,000 CRO.

100% of Ult and Cronos Launch revenue will now buy CRO on the open market and burn it monthly, with every transaction hash published.

Staking… pic.twitter.com/vE5MBp9HeB

— Cronos Network (@CronosNetwork) October 3, 2026

The Cronos token burn framework links supply reduction directly to activity on both products. Ult generates trading revenue, while Cronos Launch collects fees associated with tokens created and traded through its launchpad.

Monthly burns provide a schedule for removing purchased tokens, and published transaction hashes allow independent verification. Community members can check execution amounts rather than relying solely on announcements about future commitments. 

Under the Cronos token burn plan, published records can help compare revenue commitments with the tokens actually purchased and destroyed. A transaction record verifies execution, while separate revenue reporting would explain whether purchases reflect all earnings from the two platforms.

The revenue percentage is fixed, but the purchase budget is not. As an illustration, $100,000 would acquire about 1.5 million CRO near $0.0669 before fees and execution costs.

That calculation highlights why product revenue remains a central measure for future CRO buybacks. More usage can generate larger purchases, although lower revenue would reduce the number of tokens acquired.

Staking rewards retain their current parameters, including existing lock periods and bonus structures. The Strategic Reserve will support payouts as emissions decline, keeping reward funding separate from revenue committed to monthly burns. 

The Cronos token burn does not change the review process for potential Crypto.com listings. The published strategy says qualifying tokens enter an ongoing assessment, with Crypto.com controlling criteria and final decisions. 

Cronos Launch opened on September 15, followed by Ult on September 17. The September governance document listed the purchasing contract as being developed. Subsequent execution hashes will show the amounts bought and burned.

FX strategists at Westpac still expect a November RBA hike, but Goldman Sachs and Rabobank favour patience, leaving AUD/USD sensitive to the next inflation report. Australia's softer-than-expected underlying inflation has weakened market bets on another November interest-rate rise without shifting Westpac's forecast, exposing a disagreement that matters for the Australian Dollar.

Latest — Exchange Rates:

Australian Dollar to Dollar (AUD/USD): 0.695707 (+0.40%)

Pound to Australian Dollar (GBP/AUD): 1.903035 (-0.09%)

Euro to Australian Dollar (EUR/AUD): 1.617382 (-0.37%)

August's trimmed-mean inflation of 0.2% on the month and 3.6% annually matched Westpac's expectations and remains consistent with its forecast for a 0.9% quarterly rise in the third quarter.

Goldman Sachs expects an extended pause, while Rabobank believes any further tightening is more likely to wait until early 2027.

AUD/USD closed Friday, 2 October, near US$0.6957, up 0.40% on the day, after losing 3.15% in September.

The recovery accompanied a broader Dollar decline following the weaker-than-expected US jobs report, which reduced the immediate pressure for another Federal Reserve hike.

Markets cut November odds, but Westpac holds its call Westpac's 1 October Morning Report recorded a clear change in Australian rate expectations after the CPI release.

"Following a slight downside surprise on inflation figures yesterday, near-term pricing for the RBA has recalibrated, with a follow-up rate hike in November now seen as a circa 25% chance versus a circa 40% chance previously."

Those are the report's dated market probabilities, rather than Westpac's own forecast.

Its economists saw little reason to alter their inflation projections.

"In Australia, the August CPI was spot on Westpac's forecast, with headline inflation up 0.4%mth/4.0%yr and trimmed mean inflation up 0.2%mth/3.6%yr, though the latter was slightly below market expectations."

The softer surprise should not be confused with falling annual headline inflation, which rose to 4.0% from July's 3.5%, while the annual trimmed mean was unchanged.

In its analysis of the latest CPI figures, Westpac analysts retained its November hike call, warning that businesses passing on higher energy costs posed upside risks to underlying inflation later this year.

Its 0.9% third-quarter trimmed-mean forecast would represent an acceleration from 0.8% in the second quarter.

How much restraint is already in place? The RBA's September increase took the cash rate to 4.60%, following a cumulative percentage point of tightening during 2026.

The disagreement centres on whether those increases need more time to restrain demand or whether persistent inflation requires another move in November.

Goldman's assessment after the September decision favoured waiting.

"Our base case is that the RBA holds policy for an extended period until a gradual easing cycle commences from 2H2027."

That means cuts starting in the second half of next year, although Goldman still regarded further tightening as a material risk, including in November if inflation surprised higher.

Rabobank's October outlook gives greater weight to the danger of excessive restraint.

"While we do not rule out the possibility of further rate hikes, we think the RBA is going to be feeling its way and will be careful that any further tightening does not tip the economy into recession or cause a disorderly spike in the unemployment rate."

The bank highlighted weaker new-home construction inflation and a 0.4% monthly fall in market services prices excluding volatile items.

"Overall, we believe that the August inflation data supports our view that the RBA will not be in a rush to raise the cash rate in November and that any further lift in the cash rate is likely to be delayed until early next year."

For AUD/USD, fewer expected Australian hikes remove one potential source of yield support, although Friday's rally shows how weaker US data can offset that pressure.

The next Australian CPI release on 28 October will include September-quarter figures, allowing investors to test Westpac's 0.9% forecast before the November decision.

McDonald's is handing income investors bigger checks each year while Domino's holds its payout flat, and the reason behind that gap reveals which pizza-and-burger giant faces a trickier road ahead.

This post may contain links from our sponsors and affiliates, and Flywheel Publishing may receive compensation for actions taken through them.

Domino’s Pizza (NASDAQ:DPZ | DPZ Price Prediction) and McDonald’s (NYSE:MCD) reported second-quarter results this summer. Their dividends now move at different speeds. McDonald’s raised its payout again. Domino’s has held its payout flat since its February increase. Income investors are inspecting the yields.

Pizza Orders Climbed While Burger Traffic Slipped Domino’s U.S. same-store sales grew just 0.1%, compared with 3.4% a year ago. Order counts rose in both delivery and carryout, but the lower average check caused the slowdown. CFO Sandeep Reddy blamed a Premium Series that underperformed and promised, “We had a one-quarter blip on ticket. We’re not going to have another blip.”

McDonald’s grew global comparable sales 1.3%. Every segment posted positive comps, but U.S. restaurant traffic fell and sales declined in China and France. CEO Chris Kempczinski named Skye Anderson president of McDonald’s USA to “bring focus and urgency” to its biggest market.

Dividend Lens Domino’s McDonald’s Quarterly Payout $1.99, flat $1.93, up from $1.86 Forward Yield About 2.67% About 3.12% 2025 Free Cash Flow $671.5M $7.186B Cash Engine Supply chain plus royalties Rent plus royalties Domino’s Flat Year Matches Its Usual Pattern Domino’s raised its quarterly dividend from $0.20 in 2013 to $1.99 in 2026, raising once yearly then holding flat. On the balance sheet, stockholders’ equity is at negative $3.98B due to debt and buybacks, and cash fell 39.6% to $164.8M. Supply chain revenue of $731.7M provides regular cash as franchisees buy dough weekly.

McDonald’s makes rent as a landlord, a recurring income stream. It paid about $1.3B in dividends in the second quarter alone, while it also funds $3.7 to $3.9B in capital spending. One risk: management expects interest expense to rise 4% to 6% this year.

February Will Show Whether Domino’s Is Pausing or Stalling Domino’s announced its last increase in February 2026. A skipped increase next February combined with lower average checks and cash-draining buybacks would signal a real problem. Management cut its U.S. store-opening outlook to about 175 as franchisees make less. At McDonald’s, watch whether U.S. traffic recovers under its new president.

Regular Rent Versus Faster Dividend Growth McDonald’s offers regular, more predictable income. Its yield is higher, raises come yearly, and rent supports them. The tradeoff is slower dividend growth.

Domino’s offers faster dividend growth alongside more debt. Its payout rose from $0.20 in 2013 to $1.99 in 2026. A ticket recovery would strengthen the case. Another lower quarter followed by a skipped increase would change the outlook.

Contact [email protected] for any questions or corrections.

Shares of Adobe (ADBE -1.49%) have had a tough 2026. The stock ended last year at about $350 and trades near $240 as I write, down about 31%.

The software company has kept buying its own shares on the way down. In April, its board approved a new $25 billion stock buyback authorization, meant partly to "reduce share count over time." It runs through April 30, 2030.

And as Adobe's fiscal third quarter of 2026 ended Aug. 28, $24.55 billion of it was still unused.

Compared with Adobe's market value, the amount is large. With about 389.2 million shares outstanding in mid-September, Adobe is worth about $93 billion at today's stock price. The remaining authorization could buy back about 26% of it -- about 102 million shares.

Sure, an authorization isn't a promise. Adobe has around three and a half years to use it, and nothing forces the company to spend it all. But I think the company's pace this year shows it plans to put plenty of it to work.

Image source: Getty Images.

A steady, cash-funded paceIn the fiscal year's first three quarters, Adobe's buybacks averaged around $2.3 billion a quarter. That came to $6.82 billion, for 26.1 million shares, all bought on the open market.

And as the stock dropped, the same money bought more shares. The 9.5 million shares Adobe repurchased in fiscal Q3 cost an average of around $235 each, versus about $261 over all three quarters.

Notably, the company paid for it with cash from its operations. Adobe generated $7.65 billion of cash from operations during the same stretch, up around 11% year over year, and buybacks used close to 90% of it. Its balance sheet had $5.64 billion of cash and short-term investments at quarter-end, versus around $6.4 billion of debt.

This pace leaves room to spare. Spreading the unspent $24.55 billion evenly through April 2030 comes out to around $7 billion a year, about what Adobe spent in just those three quarters. If the company keeps buying at its recent rate instead, it could use up the authorization sometime in 2029.

Buybacks are doing about half the workAdobe's diluted share count (the number used to calculate earnings per share) was around 395 million in the fiscal third quarter. That compares to 402 million in the fiscal second quarter and 424 million in the year-ago quarter.

Management's fourth-quarter targets assume around 389 million. In other words, the count's dropping about 2% a quarter, and steadily.

Not every repurchased share sticks, though. Shares outstanding dropped by around 22 million in the nine months, fewer than the 26.1 million Adobe bought, because employee stock awards added some back.

Showing what the shrinking count means for shareholders, Adobe's non-GAAP (adjusted) net income rose around 8% year over year in fiscal Q3 to $2.42 billion. Adjusted earnings per share climbed 15% to $6.13.

On a GAAP basis, net income grew just 3%, but earnings per share climbed 11% to $4.62. In short, fewer shares drove about half of the quarter's adjusted per-share growth.

If Adobe spent all $24.55 billion at roughly $240 a share and its earnings stayed flat, retiring about 26% of its shares would raise earnings per share by over a third.

That won't happen all at once, and the price Adobe pays will move. A lower stock price would stretch the cash further, and a higher one would shrink it.

Premium Feature

Moneyball Superscore

81/100

Today's Change

(

-1.49

%) $

-3.59

Current Price

$

237.69

Is the buyback a reason to buy?At roughly $240, Adobe shares cost about 13 times earnings. Using the midpoint of management's fiscal 2026 target for adjusted earnings per share ($24.45 to $24.50), the price is about 10 times adjusted earnings. When shares reached their 52-week high of $363.70, that same target worked out to almost 15 times adjusted earnings.

I'd say buying back stock at about 10 times adjusted earnings is one of the better uses of Adobe's cash right now.

True, a buyback can't fix slowing growth. Adobe's total annualized recurring revenue (ARR), its gauge of the subscription business on a yearly basis, rose 11.2% year over year in fiscal Q3, and management's full-year target is 10.2%. And the market's discount probably reflects worries that artificial intelligence tools might hurt Adobe's pricing power. But as long as the subscription business keeps growing, a falling share count should help earnings per share outpace the business itself.

All in all, I think the buyback makes a cheap stock more interesting. Adobe stock looks worth considering at this price, but I'd want the subscription business to keep growing at roughly its current pace.

Customers at Robinhood Markets (HOOD +1.43%) can buy and sell many of the most popular stocks and exchange-traded funds (ETFs) from Sunday at 8 p.m. ET until Friday at 8 p.m. ET on the broker's 24 Hour Market. That adds up to 120 hours a week, and it leaves the weekend as the only time when stocks are off-limits.

At its HOOD Summit in Houston on Sept. 29, the firm said it plans to fill the gap by letting customers trade a curated list of U.S. stocks and ETFs around the clock, weekends included. Robinhood expects to launch it early next year, pending regulatory review.

Adding weekends would bring the 24 Hour Market to all 168 hours in a week, or about 40% more time to trade. But I don't think 40% more hours means anywhere near 40% more revenue.

Image source: The Motley Fool.

Stock trading is a smaller business than it looksRobinhood's equities revenue comes mainly from payment for order flow (fees that trading firms pay the broker to handle its customers' stock orders).

Showing how fast this line is rising, equities revenue almost doubled year over year to $129 million in the second quarter, according to Robinhood's quarterly filing. That was up from $82 million in the first quarter, too. So equities made up around 17% of transaction-based revenue, compared with about 13% three months earlier and 12% a year before. Average stock-trading volume per trader grew 56%, and the number of customers making stock trades climbed 13%.

But against total revenue of $1.31 billion (up 32% year over year), equities supplied just 10%.

And Robinhood's take is small. Customers traded a record $956 billion in stock in the quarter, up 85% from a year earlier, so the company earned around $1.35 of equities revenue per $10,000 traded.

This growth has slowed a little since then. In August, customers traded 68% more stock by dollar value than a year earlier, according to Robinhood's monthly operating data.

How much might weekends add?Robinhood doesn't say in its quarterly reports how much of its stock volume already trades in the 24 Hour Market, so any estimate here is rough.

At the second quarter's pace, equities revenue comes to around $516 million a year. If weekend hours boosted that by the full 40% the extra time represents, the rise would be about $206 million, or about 4% of Robinhood's yearly revenue pace of $5.2 billion.

I'd say the company probably won't get near that ceiling. Weekend trading will start with a limited list of stocks and ETFs. And some Saturday trades could just be trades a customer would've made on Monday morning anyway, which shifts revenue around instead of adding to it.

A 5% to 10% boost to equities revenue would add about $26 million to $52 million a year -- 1% of sales at most.

Sure, weekend hours could also keep active traders on the app on days when stocks have been off-limits, and so could the artificial intelligence (AI) agents Robinhood introduced at the summit to trade for customers. Any payoff there is tougher to measure.

Cheap to startThe summit's other launches tie into bigger revenue lines. But the tech company is pricing them to win customers first.

Perpetual futures, which are leveraged crypto contracts that never expire, are starting to roll out to eligible U.S. customers, with up to 10x leverage on Bitcoin and Ether and 3x on six other coins. Robinhood plans to charge just 0.01% a trade on them through the end of the year. Over time, they might help a shrinking crypto business, whose second-quarter revenue dropped 38% from a year earlier, to $100 million.

Earnings contracts, exchange-listed binary options linked to company-specific performance metrics, are expected to launch in October, pending regulatory approval. Robinhood won't charge contract fees for them through year-end, either.

Premium Feature

Moneyball Superscore

81/100

Today's Change

(

1.43

%) $

1.59

Current Price

$

112.74

Also starting in October, Robinhood is adding 35% more options trading hours for its most popular symbols. The same caution about hours applies. But options are Robinhood's largest transaction line, at $342 million last quarter, so the move is arguably the bigger opportunity.

Ultimately, weekend trading looks more like a convenience for Robinhood's most active customers than a big new source of revenue.

At about $112 as of this writing, Robinhood's shares cost around 38 times earnings, using analysts' forecasts for 2027.

For a growth stock with revenue climbing so fast, some premium makes sense. But the price already assumes Robinhood keeps adding new ways to make money for years to come, and weekend trading is probably one of the smaller ones.

Nu Holdings stock has pulled back sharply in recent months, falling from a high of $18.95 in January to around $13.43 today. Despite the decline, the long-term growth story remains intact, supported by the company’s expanding customer base, strong presence in Brazil, and continued push into other Latin American markets such as Mexico and Colombia.

Goldman Sachs analysts also remain optimistic about Nu’s outlook, believing the stock has further upside potential in the near term. Their bullish view suggests that the recent pullback may be more of a valuation reset than a deterioration in the company’s underlying fundamentals.

In a recent note, analysts at Goldman Sachs maintained its bullish outlook for the stock. In a note last week, they noted that the company’s entry into the United States may work out in the long term. The note argued that the company was doing so using a low-cost approach, coupled with a good customer care experience, which will help gain some market share in the country. The report added:

“We think NU’s ultra-low-cost digital approach with a strong consumer experience could allow it to successfully enter the market.”

Other analysts are largely bullish on Nu Holdings. For example, Rothschilds and Needham’s analysts set a target of $19, higher than the current $13. Wolfe Research has a target of $17, while UBS and JPMorgan Chase have targets of $18.20 and 20, respectively. 

The most recent results showed that Nu Holdings’ business continued doing well, with its number of users and customers continuing to rise. It added nearly 4 million customers in the second quarter of the year and now has 139 million globally. 

Most of these users are in Brazil, where it has close to 118 million users, a notable figure since Brazil has a population of 213 million people. Nu Holdings has also become a big neobank in Mexico, where it has 15.8 million customers. Its Colombia business has attracted over 5 million users. 

The company is now planning to launch its business in the UK, where it was said to be considering making a bid for Monzo, one of the country’s biggest neobank. It decided to walk away after Monzo insisted on a £10 billion valuation.

Nu Holdings shares chart | Source: TradingView

The daily chart shows that Nu Holdings shares have dropped by 30% from the all-time high even as its revenue and user growth accelerated. Most recently, the stock formed a down-gap after reports of its Monzo buyout emerged and then filled it after it denied the rumors.

Even so, it remains below the 50-day Exponential Moving Average (EMA), a sign that bears have prevailed. It also sits below the Strong, Pivot, Reverse level of the Murrey Math Lines tool.

Therefore, the stock will likely continue rising as bulls target the important resistance level of $15. This view will be confirmed when it jumps above the Strong Pivot Reverse level of the Murrey Math Lines at $14.

Jensen Huang, one of the co-founders of Nvidia (NVDA +1.34%), is no stranger to making bold predictions. One that he has been floating for a few years is that data center capital expenditures will reach $3 trillion to $4 trillion annually by 2030. That's a massive increase from today's levels, and he just commented that he believes that projection is still accurate.

While that projection may sound outlandish, I think there's some truth to it, because Huang is privy to conversations that the regular investor is not.

Nvidia CEO Jensen Huang. Image source: Nvidia Corporation.

Nvidia is poised to head higher with this guidance While investors should view most CEO comments through a cautionary lens, this is one situation where I think investors should be a bit more trusting. The data center build-out is in full swing, and many companies cannot access enough computing units to meet their needs. This is causing them to plan out their growth over several years. Part of this means sharing their build-out projections with suppliers like Nvidia, which can then tell their suppliers how much capacity to prepare for. This can happen years in advance of when the chips are actually needed, so with the start of 2030 being just over three years away, it's not outlandish to think that there are plans this far out.

Premium Feature

Moneyball Superscore

95/100

Today's Change

(

1.34

%) $

3.09

Current Price

$

233.95

So, I think investors can take most of what Huang says at face value.

And if he ends up being right, there could be monstrous upside.

In 2026, Nvidia expects the big five artificial intelligence (AI) hyperscalers to spend nearly $800 billion on capital expenditures. This year, Wall Street analysts estimate Nvidia's revenue will reach $411 billion. There are other companies spending big on AI outside the big five, so the actual total for this year may reach as high as $1 trillion. That would give Nvidia about a 40% share of all spending.

If it can maintain that share, Nvidia's revenue could top $1 trillion by 2030. That would easily lead to the stock doubling over that time frame, and anytime you can find a stock that has the potential to double in under five years, it's a fantastic investment opportunity.

Nvidia is still one of the top ways to invest in the AI build-out; with Huang reaffirming the strength of the AI arms race, I think it's a great buying signal for investors.

Sat 03 Oct 2026 ▪ 7 min read ▪ by Evans S.

Summarize this article with:

Bitcoin reserves on exchanges have fallen to around 2.68 million BTC, their lowest level since 2023 according to CryptoQuant data. At the beginning of 2024, they were still close to 3.2 million BTC. The decline continued despite strong price fluctuations, until the 2025 rally and then the 2026 pullback. Less Bitcoin therefore remains directly available on trading platforms. An interesting signal, but not automatically bullish.

In Brief Exchanges now hold about 2.68 million Bitcoin. Reserves were still close to 3.2 million BTC at the start of 2024. The decline reduces the immediately available supply without guaranteeing a price increase. 2.68 Million Bitcoin Remain on Exchanges The trend is not new this week. Since spring, Bitcoin reserves on exchanges had already fallen to their lowest level since 2023. Nearly 100,000 BTC then left Binance, OKX, and Gemini in less than three months.

The movement continues. According to CryptoQuant, monitored platforms now hold about 2.68 million BTC. The “Exchange Reserve” indicator counts bitcoins in wallets identified as belonging to exchanges. A drop means that some coins leave these addresses to go elsewhere: personal wallet, custody service, institution, OTC structure, or another on-chain address.

At the start of 2024, the same indicator had reached about 3.2 million BTC while Bitcoin was still trading below $70,000.

Since then, the market has experienced several phases. Bitcoin surpassed $73,000 in 2024. It then reached a peak above $126,000 in 2025, before suffering a much larger correction in 2026. Exchange reserves have continued to follow an overall downward slope.

The price returned to around $85,000 in early October. That was not enough to massively send BTC back to platforms. On October 3rd, the overall reserve remained close to 2.68 million. For a holder who does not plan to sell, leaving their bitcoins on a platform is not essential. Moving to a personal wallet also reduces risk directly linked to the exchange.

Bitcoin in 2022 The events of 2022 had already accelerated this habit. But 2026 adds other players. US spot ETFs, publicly listed companies holding Bitcoin, and institutional custody services have profoundly changed how BTC circulate. A coin leaving Binance or Coinbase does not necessarily end up in a private individual’s personal wallet.

The contrast is visible with Binance. While the overall reserve drops, the largest crypto platform experienced the opposite movement in September. Its reserves reached about 702,900 BTC on September 19, according to CryptoQuant data cited by U.Today. It was their highest level of the year.

A platform can therefore fill up while the overall market empties. This is precisely what prevents reducing the figure of 2.68 million to a simple phrase like: “everyone is buying Bitcoin.” However, the underlying trend remains clear. There are fewer BTC on monitored exchanges today than at the beginning of 2024.

Less Available Bitcoin Does Not Necessarily Mean Shortage On paper, the mechanism seems simple. When an investor wants to quickly sell Bitcoin, they generally send their BTC to an exchange. When reserves strongly increase, more coins can theoretically come to the market.

Conversely, when bitcoins leave platforms, this immediately available reserve shrinks. CryptoQuant even describes the Exchange Reserve as a direct indicator of potentially available supply on the sellers’ side. A durable drop is historically associated with shifts towards off-exchange custody and potentially lower selling pressure.

But the important word remains “potential.” This summer showed this well. In August, 28,000 BTC had returned to exchanges in less than three weeks. This rebound had erased much of the accumulated drop during previous weeks.

Bitcoin had not therefore immediately entered a big selling phase. Exchange reserves measure asset location, not their owner’s exact intention.

The case of ETFs further complicates the reading. US spot Bitcoin ETFs attracted $2.65 billion net in September. It was their second-best month since October 2025. Institutional demand clearly rebounded while overall exchange reserves kept falling.

Yet, these funds do not operate like an individual opening Binance and immediately withdrawing BTC to a Ledger.

A significant part of their operations goes through institutional custodians and over-the-counter transactions. Movements thus do not always appear as expected in simple exchange reserve statistics.

Wallet Classification Analysis companies must identify which addresses belong to Binance, Coinbase, Kraken, or other exchanges. When a platform changes its custody architecture, creates new wallets, or moves assets between different structures, reading certain data can temporarily change.

This explains why absolute levels sometimes vary from one data provider to another. The signal remains useful. It simply should not be turned into a prediction. This caution is especially important today. Bitcoin just completed a very solid third quarter. The price gained more than 40% between July and September before settling around $83,000 to $87,000 at month-end. Meanwhile, CryptoQuant also noted profit-taking and some transfers back to exchanges.

Several movements coexist. Bitcoins overall leave platforms. Some actors send them back. ETFs attract capital again. Other investors take profits after the rally. The market does not move as a whole.

ETFs and Large Holders Change BTC Circulation The reserve decrease makes more sense when viewed over several years. Bitcoin’s market structure is no longer the same as in 2020 or 2021. US spot ETFs now hold a considerable amount of BTC. Companies have also accumulated significant reserves. Strategy, the largest publicly traded corporate holder, owned 847,666 BTC at the end of September after buying 1,665 more bitcoins for around $143 million.

These bitcoins are not used to feed daily order books on exchanges.

They can remain immobilized for long. Private individuals have also resumed their purchases. In the third quarter, they accumulated about 107,000 BTC while funds and ETFs temporarily reduced their positions.

The situation changed again in September with the return of inflows into ETFs. This is probably the most interesting part of today’s market: several buyer categories can take turns without the amount of Bitcoin on exchanges rising sustainably.

This naturally fuels the idea of a “supply shock.” The scenario is known. Bitcoin has a maximum supply of 21 million coins. Some BTC are lost or inactive for years. Other bitcoins are in ETFs, company treasuries, or long-term wallets. If the quantity available on exchanges decreases while demand rapidly rises, buyers theoretically must offer higher prices to find sellers.

Maximize your Cointribune experience with our "Read to Earn" program! For every article you read, earn points and access exclusive rewards. Sign up now and start earning benefits.

Join the program

A

A

Lien copié

Evans S.

Fascinated by Bitcoin since 2017, Evariste has continuously researched the subject. While his initial interest was in trading, he now actively seeks to understand all advances centered on cryptocurrencies. As an editor, he strives to consistently deliver high-quality work that reflects the state of the sector as a whole.

Cover image via U.Today Disclaimer: The opinions expressed by our writers are their own and do not represent the views of U.Today. The financial and market information provided on U.Today is intended for informational purposes only. U.Today is not liable for any financial losses incurred while trading cryptocurrencies. Conduct your own research by contacting financial experts before making any investment decisions. We believe that all content is accurate as of the date of publication, but certain offers mentioned may no longer be available.

In the past week, DogeOS launched a public test network for Ethereum-style trading, lending and other applications using DOGE, an attempt to give Dogecoin holders more ways to use their coins beyond payments and price speculation.

Dogecoin Foundation director Timothy Stebbing highlighted this milestone in a recent X post, noting an "absolute avalanche" of new projects building on the Doge/EVM chain.

In the last week we've seen @DogeOS go live with their public testnet and an absolute avalanche of new projects building on their doge/EVM chain, and then Metallicus launched DogecoinVM which bridges Dogecoin to their Metal network.

There are more options than ever to build…

HOT Stories

— Timothy Stebbing (@tjstebbing) October 3, 2026 Describing itself as the app layer for Dogecoin, DogeOS was founded by the team behind MyDoge, a Dogecoin wallet, and allows developers to use familiar software to build applications by leveraging Ethereum compatibility.

You Might Also Like

Along similar lines, Metallicus launched DogecoinVM, which bridges Dogecoin to the Metal network. DogecoinVM is described as a DOGE-compatible ledger on the Metal Blockchain. While Dogecoin itself does not change, what changes is the speed and interoperability, making Dogecoin 300x faster with instant finality.

The recent developments, according to Stebbing, imply there are more options than ever to build amazing solutions on Dogecoin, a positive for the network.

Dogecoin eyes first 2026 daily golden crossIn a positive price development, Dogecoin is on track for a golden cross on its daily chart as the MA 50, and MA 200 are set to converge. If confirmed, this will be Dogecoin's first golden cross of 2026.

You Might Also Like

This development remains significant for Dogecoin, which last saw a golden cross more than a year ago. Dogecoin's last golden cross was in August 2025 and one month later, in September 2025, it rose to a high of $0.30.

Dogecoin was trading down on a daily and a weekly basis to $0.093 alongside the rest of the crypto market as a weaker-than-expected September jobs report altered traders' outlook for a rate hike by the Federal Reserve in October, which is now seen as extremely unlikely.

Investors are still pricing in that the Fed might hike interest rates in December. The decision on the next interest rate will be announced at the end of a two-day policy meeting on October 28.

Elon Musk told SpaceX employees in August that AI revenue would overtake every other part of the business by September. The deadline has passed, and the numbers that will confirm or contradict him are weeks away from filing.

This post may contain links from our sponsors and affiliates, and Flywheel Publishing may receive compensation for actions taken through them.

Elon Musk set a date. He released a 30-minute all-hands video on X on August 11, 2026, which Johnny Rice of The Motley Fool reported on. In it, the SpaceX (NASDAQ:SPCX | SPCX Price Prediction) chief told employees: “Probably our AI revenue, not probably, definitely, our AI revenue will exceed all other SpaceX revenue probably in September, like next month.”

Musk opened with “probably,” switched to “definitely,” then returned to “probably in September.” He stated the crossover with confidence but hedged timing. September has ended, so investors can check this forecast against reported numbers.

What the Crossover Required Johnny Rice cited SpaceX’s second-quarter report: total revenue was $7.81 billion, up 92%. Per Rice, AI contributed $2.56 billion, while Connectivity and Space brought in $5.25 billion combined. AI needed to more than double, assuming the rest of SpaceX stopped growing.

The rest of SpaceX kept growing. The Q2 earnings release filed with the SEC shows Connectivity revenue of $4.291 billion, up 66%. Starlink’s enterprise and government line rose 108%. AI revenue rose 247%, but its $367 million advertising line fell year over year.

On the August 4 earnings call, a SpaceX executive said the company had “contracted an additional $6.7 billion of cloud services revenue over a six-month period that begins ramping starting in October of this year.” That revenue starts after Musk’s deadline.

Musk’s Bigger Number Dwarfs Wall Street’s In the same video, Musk set a target of 10 gigawatts of AI computing capacity by the end of next year. He valued that capacity at roughly $30 to $50 per watt. From those two figures, he derived annual AI revenue of $300 billion to $500 billion by 2028. On the earnings call, he described the per-watt figure this way: “This is just a guess.”

Musk also said Grok will be trained on SpaceX’s internal information, including employees’ own work: “So in a way, it’ll be trained on you.”

Rice cited a much smaller outside estimate. After SpaceX signed its computing deals with Anthropic and Alphabet (NASDAQ:GOOGL), Goldman Sachs (NYSE:GS) projected $15.6 billion in AI revenue for all of 2026, according to The Motley Fool.

Three Obstacles Between the Forecast and the Filings Power. SpaceX faces public backlash over methane-burning gas turbines powering Colossus 1 and 2 (The Motley Fool). Capital. Rice noted SpaceX spent over $25 billion on AI last quarter. The filing shows $15.83 billion of the quarter’s $18.37 billion capex went to AI compute. Capacity stood at 1.4 gigawatts at Q2 end. Rice expects a funding shortfall despite $93.52 billion in cash. Musk said SpaceX expects only “a very small percentage” of NVIDIA (NASDAQ:NVDA) GPUs next year. Contracts. Either party can exit the Anthropic and Google contracts with a 90-day notice, according to The Motley Fool. If a customer leaves, SpaceX misses the revenue it built capacity for. How to Settle the September Question SpaceX has filed nothing since the August 4 8-K. The third-quarter report, expected in late October or early November, should answer the question. Check SpaceX’s investor relations site for:

Q3 AI division revenue versus Connectivity plus Space. Look for monthly or exit-rate disclosure if quarterly totals don’t break out September. Progress toward the $100 billion annualized run rate targeted for December, which Musk called “not a question mark”. Whether any customer has used its exit clause. The stock has already moved on the AI story. Shares closed at $158.96 on October 2, up 7.35% on the day and 12.97% over the past month. On October 1, SpaceX launched Google AI chips into orbit as part of its space-based data center push (CNBC). If Q3 numbers show AI still trailing Starlink, the stock will be trading on Musk’s forecast ahead of reported results.

Contact [email protected] for any questions or corrections.

Intel's (INTC -0.56%) comeback has been quite a sight to behold. As of this writing, shares of the chipmaker have exploded by more than 220% in 2026 in part due to surging demand for the company's CPUs, but also due to increased optimism about the outlook for its foundry business. And even with that incredible run-up behind it, patient investors may still want to consider buying Intel's stock.

Premium Feature

Moneyball Superscore

70/100

Today's Change

(

-0.56

%) $

-0.67

Current Price

$

119.33

Intel's turnaround is gaining momentum Intel's second quarter was one of its best in several years. Revenue jumped by 25% year over year to more than $16 billion. Intel's gross margin also improved tremendously, rising to 40.4% from 27.5%.

The bigger story here is Intel's foundry division and the turnaround happening within it. Intel Foundry's revenue was nearly $6 billion in the quarter, and its output paced 25% above the target. While the segment posted a $2.1 billion loss, the gap is narrowing. Foundry revenue from external customers was just $293 million, which means that if its turnaround is to gain steam, Intel really needs to win more business from outside clients.

The company's GAAP loss was $2.16 per share, but it's important to note that this was largely due to an escrow arrangement related to funds it was awarded through the CHIPS Act. On the books, that showed up in the form of a non-cash charge of $12.53 billion. So the quarter's loss was not reflective of operational issues.

Image source: The Motley Fool.

14A is make or break Much of Intel Foundry's future success hangs on the successful launch of Intel 14A. This is a next-generation semiconductor fabrication process node that will produce chips with even greater transistor density than 2-nanometer chips. The development of its 14A foundry is nearing completion, with risk production planned for 2027 and a high-volume mass-production target of 2028. Management's hope is that its 14A manufacturing services will position it to pose a serious challenge to Taiwan Semiconductor's dominance in cutting-edge chip manufacturing.

Premium Feature

Moneyball Superscore

97/100

Today's Change

(

2.96

%) $

13.58

Current Price

$

472.78

In a bullish sign, Intel CFO David Zinsner implied on the Q2 earnings call that the development is going very well and that conversations with potential customers have shifted from evaluating 14A to inquiring about how much capacity they can secure.

It's been reported that a slew of megacaps, including most of the "Magnificent Seven," are among Intel's prospective foundry clients. However, Intel will need to secure one of these large-scale customers sooner rather than later.

In June, Bank of America analyst Vivek Arya gave Intel stock an unusual double-upgrade from underperform to buy, citing an optimistic outlook for its foundry business and strong demand. Intel did have to engage in a $23 billion secondary stock sale in August to help fund its foundry build-out.

Intel is firing on all cylinders Intel is due to report its third-quarter earnings on Oct. 22. The company's guidance suggests revenue could be as high as $16.8 billion. One concern is that Intel's valuation is sky-high right now. The company for years maintained a forward P/E ratio in the low teens. Now that metric stands at more than 60.

Furthermore, the consensus price target among analysts is just above $116 per share, and the stock is trading above that as of this writing. There isn't much room for error with Intel's stock.

However, given the company's main revenue sources and an ever-improving foundry operation, I'm quite bullish on Intel's future -- as long as its 14A process node proves as promising as advertised. For interested investors, it may be prudent to build a position in Intel slowly rather than all at once, given the milestones the company still needs to hit over the next couple of years.

It's been a banner year for Dell Technologies (DELL +3.84%), with its stock up nearly 325% at the time of this writing.

Like many other large tech stocks, it has found its place in the artificial intelligence (AI) supply chain, specifically with its AI servers, which have been in high demand this year as graphics processing unit (GPU) clusters and data centers scale.

As such, Dell has been able to distribute more capital back to shareholders, noticeably raising its quarterly dividend by 20% earlier this year. Dell only initiated its dividend program in 2023.

Can investors trust the stock for passive income?

Image source: The Motley Fool.

Why Dell's stock has exploded this year While GPUs and other chips have been at the center of the AI trade, they need to interact with other components in a data center to obtain the capabilities needed to respond to queries from, say, a large language model.

Dell's AI-optimized servers integrate the chips, memory, and other necessary components that power them, including liquid cooling and data storage, into a full server and rack.

The company also performs quality control and then ships them to customers such as neoclouds and other large customers, including hyperscalers that operate their own data centers.

While Dell isn't the only company building and selling full-scale AI racks, it's been all hands on deck amid massive AI demand, resulting in shortages across the supply chain.

Premium Feature

Moneyball Superscore

87/100

Today's Change

(

3.84

%) $

20.78

Current Price

$

562.52

In its second quarter of fiscal year 2027, the company posted $47 billion in revenue, up 58% year over year, while diluted earnings per share of $6.34 surged 273%.

Management also lifted its full-year guidance from $167 billion to $192 billion, implying 69% annual growth.

"Demand is broadening across neoclouds, sovereigns, and enterprise customers, and our customer count has surpassed 6,500," Dell's Chief Operating Officer Jeff Clarke said on the company's most recent earnings call. "The scale and complexity of these deployments reinforce why customers choose us."

Can Dell's dividend be trusted? It's pretty easy to see why Dell was able to raise its quarterly dividend by 20% to $0.63 per common share.

The much-improved financials have also contributed to a 224% increase in adjusted free cash flow in the second quarter. More free cash flow means more money for dividends. Dell's trailing-12-month free cash flow yield of 2.37% far exceeds its trailing dividend yield of 0.43%.

The payout ratio is also not high, with $2.52 in annualized dividends per share easily covered by this quarter's profits alone. The company is also repurchasing stock, with nearly $3.8 billion of share repurchases in the most recent quarter, which dwarfs the $405 million of dividend payments.

Now, obviously, the big question or future concern for Dell's dividend is what happens if AI demand dries up.

While there's certainly no evidence of this happening right now, the hyperscalers continue to raise their future guidance for AI-related capital expenditures, which could very well top $1 trillion next year. But it's hard to imagine this level of spending continuing like this forever.

When evaluating stocks for dividends, it's key for investors to consider what a company's earnings and free cash flow could be over the long term to estimate what a sustainable dividend looks like and how much it can be raised each year.

The good news is that Dell's $2.62 projected annualized dividend is still well below the $8.68 in diluted earnings per share the company generated in fiscal year 2026, or even the $6.38 it earned in fiscal year 2025.

This tells me that even if earnings were to slow considerably, Dell's dividend would still be covered, so I see no reason why the dividend can't be increased annually moving forward.

That said, Dell's 0.43% annual yield is not particularly attractive, so if you're purely focused on passive income, there are likely more attractive dividend stocks out there. But higher-yielding dividend stocks may not be able to tap into the AI trade like Dell can, for better or worse.