SUCCESS

Average mortgage rates today

Mortgage Type
Label
Rate
APR

30-Year Fixed
Most Popular
6.53%
6.57%

30-Year FHA
Lower Credit
6.06%
7.28%

30-Year VA
Military
6.13%
6.29%

30-Year Jumbo
High Balance
6.66%
6.68%

15-Year Fixed
Shorter Term
5.84%
5.91%

7/6 ARM
Shorter Term
6.14%
6.22%

HELOC
Home Equity
8.09%
8.09%

Home Equity Loan
Home Equity
8.14%
8.14%

Updated on 08/26/2026

Rate data provided by RateUpdate.com. Displayed by Mortgage Research Center, LLC, NMLS# 1907, Equal Housing Opportunity, Payments do not include taxes or insurance premiums. Actual payments will be greater with taxes and insurance included. Rate and Product details

The average 30-year rate fell for the second straight day, averaging 6.57% on Wednesday. Well-qualified borrowers who are ready to buy a home should be able to find and lock in a competitive rate.
Key mortgage rate averages:

The 30-year fixed-rate mortgage averaged 6.57% APR
The 30-year fixed-rate FHA mortgage averaged 7.28% APR
The 30-year fixed-rate VA mortgage averaged 6.29% APR
The 30-year fixed-rate jumbo mortgage averaged 6.68% APR
The 15-year fixed-rate mortgage averaged 5.91% APR
The 7/6 adjustable-rate mortgage averaged 6.22% APR
The rate on a HELOC averaged 8.09% APR
The rate on a home equity loan averaged 8.14% APR

Mortgage rate trends
Rates continue to hover in the mid-6% range and are likely to remain there in the near term. Inflation remains an economic concern, as does the rising national debt. Uncertainty over these issues, among others, is keeping yields on the 10-year Treasury note (and other long-term bonds) elevated, which in turn influences mortgage rates.
Most housing economists now expect rates to remain near their current level through the end of the year. While borrowing conditions remain challenging for many buyers, those who are financially prepared can take advantage of improving market conditions. Heading into the fall, inventory is rising, home sellers are more flexible when negotiating a sale and there is less competition from other buyers.
Which loan is best for you?
When shopping for a mortgage, you may be offered several loan options that will fulfill different needs. Here’s a rundown of the most common loan types you’ll find, and who they work best for.
30-year conventional mortgage: Conventional loans work best for borrowers who have a credit score above 620, have saved enough to make a down payment of at least 3% and are looking for flexibility in the type of property being purchased.
30-year Federal Housing Administration (FHA) mortgage: FHA loans are good for first-time homebuyers, borrowers with less-than-perfect credit scores or those with a high debt-to-income ratio.
30-year U.S. Department of Veterans Affairs (VA) loan: Specifically designed for active duty and retired service members, members of the National Guard and Reserves, and surviving spouses. Offers 0% down loan options, competitive rates and accepts less-than-perfect credit scores.
30-year jumbo loan: Good for homebuyers purchasing property that is priced above the Federal Housing Finance Agency (FHFA) conforming loan limit. In 2026, that limit is $832,750 in most of the U.S. but increases to $1,249,125 in high-cost areas.
15-year fixed-rate loan: Borrowers who prefer a shorter loan term and can afford to make higher monthly payments will pay less overall interest with a 15-year mortgage and pay off the loan faster.
7/6 adjustable rate loan: Good for a buyer who wants to lock in a favorable interest rate for a set period of time and either plans on selling the home before the interest rate starts, is willing to make a higher monthly payment once the rate becomes variable or is open to refinancing the loan.
Home equity line of credit (HELOC): A good option for a homeowner who wants to access the equity they’ve accumulated in their home and have an open line of credit to use as needed.
Home equity loan: Another option for a homeowner who wants to access their home equity and have the financial capacity to take on a second mortgage.

How mortgage rates affect affordability
The rate on your mortgage can make a big difference in how much home you can afford and the size of your monthly payments. That’s true whether buying your primary residence, an investment property or refinancing an existing loan.
Here’s an example. If you bought a $250,000 home and made a 20% down payment of $50,000, you would end up with a starting loan balance of $200,000. On a $200,000 home loan with a fixed rate for 30 years, here’s what you would pay:

At a 3% interest rate = $843 in monthly payment (not including taxes, insurance, or HOA fees)
At a 4% interest rate = $955 in monthly payment (not including taxes, insurance, or HOA fees)
At a 6% interest rate = $1,199 in monthly payment (not including taxes, insurance, or HOA fees)
At an 8% interest rate = $1,468 in monthly payment (not including taxes, insurance, or HOA fees)

Experimenting with a mortgage calculator allows you to find out how much a lower rate or other changes could impact what you pay. A home affordability calculator can also estimate the maximum loan amount you may qualify for based on your income, debt-to-income ratio, mortgage interest rate and other variables. The Consumer Financial Protection Bureau can also provide a range of rates offered by lenders in each state.

Current mortgage rates FAQs
What is a 30-year mortgage rate right now?
The average rate on a 30-year fixed-rate mortgage is 6.6% as of August 25, according to Money’s rate data. Other rate surveys show 30-year rates averaging over 6.6%.
Can you get a 4% mortgage rate?
No, not under current market conditions. A 30-year fixed-rate loan is averaging in the mid-to-6% range as of August 25.
Will we ever see a 3% mortgage rate again?
Mortgage rates are unlikely to fall below 3% in the near term unless a severe economic downturn occurs. However, rates averaged in the mid-3% range before the pandemic, so a return to that range at some point in the future is not out of the question.
How much is a $300,000 mortgage at 7%?
The monthly payment on a 30-year, $300,000 conventional mortgage at 7% is $1,995.91, excluding taxes, insurance and HOA fees. Your actual payment will vary depending on your credit score, down payment, lender and location, among other factors.

Visa’s open-source security harness now finds the vulnerability, writes the fix, and turns an adversarial panel on its own patch before any human reviews it. The whole loop ships on by default. A plain scan of the Visa Vulnerability Agentic Harness runs all 11 stages and edits source files in the target repo unless the operator caps it at detection.The announcement Thursday pairs the release with an expansion of the Visa Consulting & Analytics advisory practice. Visa is shipping that default 18 days after Tenet Security demonstrated GhostJacking on the DEF CON 34 main stage, an attack chain in which an agent read an attacker’s payload out of a log file and rewrote DNS with a valid credential. Two days earlier, Steve Wilson, Chief AI and Product Officer at Exabeam and project co-lead for the OWASP Top 10 for LLM Applications, made the case in VentureBeat for the opposite default. “The first thing I’d do is put an authorization gate outside the model,” Wilson said in written responses. “The agent can propose the exact DNS change, but it cannot grant itself the authority to make it.”The bottleneck moved, so Visa moved the pipelineRajat Taneja, Visa’s president of technology, rejects the premise that the default is a risk decision and calls it the product. “The bottleneck has moved,” Taneja told VentureBeat in an exclusive interview. “AI is finding vulnerabilities faster than humans can in the history of our technology industry. The new bottleneck is fixing and proving we have fixed things.”VVAH grew out of Visa’s participation in Anthropic’s Project Glasswing, where the company aimed Claude Mythos at the network behind billions of daily transactions and watched the model chain minor weaknesses into working exploits, a hunt VentureBeat covered in July. “VVAH initially was completely only using Mythos, and that’s when all of us, as part of Project Glasswing, realized the power of this new class of models that does semantic reasoning,” Taneja said.The harness went to GitHub in June and has climbed from 595 stars and 97 forks on July 20 to more than 2,300 stars and 300 forks as of August 25, with a clone-to-visitor ratio Taneja put near 9%. “We have got some very high-profile companies that have started using this harness,” he said.Why give it away? Taneja’s answer starts with Visa’s technology DNA and a harness built “to protect Visa and our ecosystem.” The reason he leaned on hardest was obligation, “to do good by doing right” for “companies who may not have the same level of investments or knowledge in cybersecurity.”Contribution runs one way. The repo states it is not currently accepting external code contributions, so the harness that edits adopters’ source takes no code into its own.Thursday’s release extends the pipeline past the report. “We’re going from discover, verify, and report, and then fix it, to discover it, verify it, remediate it, validate it, and iterate it,” Taneja said. “If a fix doesn’t negate the exploit, then there should be a structured, automated feedback that preserves the learnings from the first run and then enhances it.” Underneath that loop, the release refactors scanning around an abstract syntax tree call graph that maps subroutine calls and the traversal paths an attacker could reach. Taneja argued the change cuts token counts while delivering “better reasoning, context, and better exploitability analysis.” On top sits MTTA observability across the stages, what he called a window pane, plus real-time progress views. “A pretty good step function,” he said of the release.One metric, three definitionsMean Time to Adapt, the metric Visa invented alongside the harness, gets a shorter definition in this release. The short form is the time between discovery and resolution of attack paths, with some resolutions, Visa claims, shrinking from weeks to hours. Visa published a wider construction in June, and the Project Glasswing white paper tracks MTTA along three dimensions that include inventory freshness, exploitable paths per release, and validation cycle time. The repo carries a third, elapsed time from AI-discovered exploitability to a validated fix in production. Board slides will quote the shortest interval. Ask for all three, because a resolution count that skips validation is what MTTA was invented to replace.Taneja ranks MTTA as “the most strategically important metric” because it shifts the focus from scanning to how fast an enterprise adapts. His shorthand is blunter. “It’s not the finding. It’s the fixing that matters,” he said.The default and the gateWilson’s argument went past naming the gate. “We have to remember that security rules written inside prompts may shape the model’s behavior, but they are still suggestions to the model, not enforceable security controls,” he wrote. He also priced the control honestly. “The tradeoff is that the agent loses the ability to improvise arbitrary, high-impact infrastructure changes on its own, while retaining autonomous investigation and routine, bounded remediation,” Wilson said.The harness ships no approval step between patch and edited file. Where the human sits was the first question VentureBeat put to Visa in writing.The company’s own June white paper sets the bar. “AI agents are identities” sits among its 12 non-negotiable practices, requiring scoped permissions, least privilege, audit trails, and IAM governance for every agent that modifies a system. VVAH’s shipped default is that agent.”A lot of the traditional systems that are used today are basically signal providers,” Taneja told VentureBeat. “They are telemetry, and then it’s a lot of human analysis, and your SOC and your security and incident response teams doing a lot of the heavy lifting when they respond,” and that, he said, cannot work at this scale. He pointed to the Hugging Face incident and “other frontier models escaping sandboxes to do things more autonomously” as the preview. “We have seen the trailer of this movie,” Taneja said, and “every company in the world should prepare and rethink their architecture.”What the harness automates is the adversarial step. Before a fix counts as validated, the panel scores whether the patch negates the exploit, Taneja’s test for done, with failed fixes feeding the next attempt, the iterate step Taneja described. Stage 11 itself runs read-only, per the README, and VVAH does not compile, build, or run tests against the patched tree. Taneja calls that wrapper “the governance architecture on top of that,” and chaining findings into working exploits takes threat modeling and business context, which is why he argued “the harness with a model is far more effective than somebody using the model by itself.”Visa answers the gate questionVentureBeat put its questions to Visa in writing after the interview, and the answers arrived before publication. On why remediation ships on, the response repeated the bottleneck argument, then narrowed the scope. “VVAH is meant for authorized operators running against code they own, and in a controlled environment,” the company said in written responses.The approval question drew the most specific answer. “VVAH is a harness, not a merge tool,” Visa wrote. “Stage 10 writes candidate fixes to a working copy of the repo. Stage 11 then runs an adversarial validation panel that scores each fix and returns one of three verdicts: validated, validation failed or needs review. None of these bypasses your normal build, test, and code review flow.” Humans, the company wrote, are “the gate in three places. Before running the tool. When reviewing the patches. And before anything gets merged.” “The final call on any fix stays with the security and engineering teams. In an enterprise, trust and auditability are not optional. The default flow is built around that.”Set beside Wilson’s standard, the architecture lands close to his line and the sequence does not. Wilson’s gate clears an action before it happens. The default’s human gates open before the run and after the write. The attack is the automated part, and the three human gates sit outside the model, the boundary Wilson drew. “Our goal is to help security teams work at AI speed, not to replace them,” Visa wrote. “VVAH does the repetitive parts. It finds issues, tests whether they are real, and proposes fixes. Before a fix gets to a human, an adversarial validation panel at stage eleven tries to break it. That way the human is spending time on decisions that need judgment, not on triaging noise.”Client zero got direct confirmation. “VVAH runs against Visa code today,” Visa wrote, and Taneja had volunteered the posture on the call. “We designed this and we were using it for ourselves, and we were client zero,” he said, adding “only when we saw the impact and the positive effect of what we were finding, we said every company would need this.” What adopters value, Visa says, is context. VVAH pulls in CMDB data, threat models, and business risk, and where “most tools stop at findings,” it “tries to answer, ‘which of these should you fix first, given how your business runs.'”Model choice becomes a per-stage decisionMulti-model orchestration is the other substantive change. “Mythos has a very high recall, but the Opus model has very high precision,” Taneja said. “On stage one I want to use this model. On stage two I want to use this model,” is how Taneja framed the per-stage setup, with newer GPT releases in the ensemble and open-weight models where pricing stings, all through configuration rather than code changes. “The whole is greater than the sum of the parts,” as he put it. The harness was model-agnostic from day one, he added, and the evolution moved that choice into configuration, with prompt tuning and caching shared underneath. One boundary moved. In June, applying a fix required Anthropic backends, and OpenAI-compatible backends ran report-only. The current README extends remediation and validation to OpenAI-compatible and open-weight models through a shared model-agnostic runtime, with no single provider as a hard dependency, and the default routing for both stages stays Anthropic.That flexibility lands on a market already churning. VentureBeat’s Q2 2026 Pulse research found 59% of enterprises plan to adopt or switch agent security tooling within the year, and 82% still rely on provider-native controls as the primary layer. Visa said Thursday it is contributing VVAH to Nvidia’s Open Secure AI Alliance as a model-agnostic framework and collaborating in Project Lightwell, the $5 billion IBM and Red Hat effort to harden open-source components.Before turning fix mode onDecisionWhat to establish firstRun postureStart with –stop-after s9 and read the SARIF output before any run that can write to source files.Approval gateMap Visa’s three human gates onto the pipeline, at run, at patch review, and at merge, and name who holds each.Validation scopeStage 11 verdicts score the fix. Build, test, and code review stay in the team’s own flow, per Visa, so keep an exploit re-test before merge.Repository scopeThe tool runs with elevated privilege, per its own README. Fence which repos the harness can reach, and run scans in an ephemeral environment with scoped credentials, no production secrets, and network limited to the target repo and model endpoint. Write access to production code is the GhostJacking exposure class, an agent acting on data it read. Per the README’s egress warning, any role routed through the SDK, OpenAI, or DeepAgents backends sends prompt data to that provider’s endpoint.Model rolesAssign models per stage deliberately. Recall and precision differ by model, per Taneja, the fix stages carry the highest blast radius, and the README states precision and recall figures are not yet published, so measure your own.Consulting is the other half of Thursday’s announcement. Visa Consulting & Analytics is adding executive workshops, a VVAH-informed maturity assessment scored on a NIST one-to-five scale, and a cyber risk prioritization roadmap. “We were getting a lot of calls. Hey, can you help?” Taneja said, and the practice “became very important to handhold and help those who are using it.” Carl Rutstein, global head of Visa Consulting & Analytics, framed it the same way. “Finding vulnerabilities is no longer the hardest part. Speed to remediation is the new battleground.”

Presented by EDB As enterprises give AI agents more autonomy — the ability to plan, decide, and act across systems without a human approving each step — a hard question moves to the center of every architecture review: When an agent tries to complete an action that it was never authorized to do, what actually stops it?These are your agents, running on your models, touching your data in your infrastructure — and the responsibility for what they do sits with you. That responsibility can’t be met in hindsight or with a set of abstract policies that live on paper but not in practice. Agents need rules in the context of the moment, because they don’t exercise overriding judgment of their own actions.Consider a simple rule: Never open the car door. Followed literally, an agent could never get in or out of the car at all. But if you change the context (the car has just crashed, there’s a fire, someone is hurt and needs to get out), then the rule you actually want is the opposite. Context in the moment is everything. We are asking agents to do intelligent things; that requires intelligent rules.The instinct is to add guardrails around the agent: instructions, policies, and monitoring layered above the model. Those mechanisms matter, but they share a structural limit: The car-door rule is plausible right up until the moment you actually have to decide whether to open the door. Controls at the agent layer are only as reliable as the agent’s output is predictable, and autonomy is precisely the property that makes that output hard to predict. Governance that depends on reviewing an action before it happens cannot keep pace with a system that acts in milliseconds, across many systems at once.Governance has to become executable, and enforced where agents actually do their work: at the operational data layer, in the context, and exactly at the moment it is happening. The data layer is the enforcement pointAgents create value by touching data. They query it, retrieve it, transform it, and increasingly act on it. A policy that says an agent should not reach a certain class of data is meaningful only if the system can deny that access at the moment the agent requests it. Additionally, a principle that says AI must be auditable is meaningful only if the organization can reconstruct what the agent did, what data it touched, which user it acted for, and what resulted. When governance lives at the data layer, it holds regardless of how the agent was built or how it behaves, because the control is a property of the database itself, not a promise made by the agent.Agent behavior may be probabilistic. Governance cannot beThe enterprise should not rely on a model choosing to follow policy. The policy has to be enforced by the system. That is the difference between hoping an actor stays in bounds and constructing bounds it cannot cross to begin with.The controls that make this real are ones many enterprises already run at the data layer: role- and attribute-based access, row- and column-level security, classification and masking, policy as code, and complete audit trails. What agents change is not the mechanism, but who the mechanism has to recognize. Identity management has to treat the agent as a principal in its own right, with its own identity and a purpose declared when the session opens. Once purpose is bound to identity, the policy engine can evaluate it the same way it evaluates role or department today, and the record of what happened can capture not just who acted and what they touched, but what they declared they were there to do.In practice, this resolves into nine controls, grouped under three imperatives:Enforce itRole- and attribute-based access control enforced at query time, for agents as well as usersDynamic column masking driven by the same policy pathAgent identity as a first-class principal, with declared purpose bound at session start and the acting user preservedSee it and prove itClassification and tagging that drives policySession-level audit logging that records which agent acted, for which user, and under what declared purposeLineage across pipelines, so a result can be traced back to the request that produced itUnify and hardenCentralized, portable policy managementEncryption at rest and in transitConsistent enforcement across on-prem, cloud, and sovereign or air-gapped environments“Declared purpose is what makes the difference. It becomes an attribute the access layer already understands, evaluated in the same policy path as role and row-level security. The enforcement mechanism does not change. What changes is that the agent’s purpose is part of what it evaluates, and part of what the record proves afterward,” says Priyanka Jain, VP, product management, data & AI governance, EDB. Wherever you are in your AI adoption journey, enforcement at the data layer is what lets you move faster rather than slower. The controls are already in the database. The difference is that agents now have to pass through them.A digital leash, not a locked doorThe goal is not to stop agents from doing useful work. It is to define how far an agent can go, what it can touch, what it can change, what requires escalation, and how the organization can reconstruct events if something goes wrong. Governed this way, agents are identified, scoped, monitored, and auditable. The enterprise can adopt them faster, because security, risk, and leadership teams trust the operating model underneath.Open, sovereign, and enforceable at the sourceBuilt on open source Postgres, this open foundation keeps enterprises in control of where their data lives, who can reach it, and under what policy, without ceding governance to a layer they don’t own or can’t inspect. For regulated industries, that combination of data sovereignty and source-level enforcement isn’t a nice-to-have; it’s the precondition for putting agents into production at all.Agentic systems will keep getting more capable and more autonomous. That is a reason to be deliberate about where control lives, not a reason to slow down. The enterprises that enforce governance at the data layer can move aggressively on AI, because the thing protecting their data is more than just wishful thinking. EDB Postgres AI is an open, enterprise-grade sovereign data and AI platform that unifies transactional, analytical, and AI workloads — with governance enforced where the data lives. For the full framework, see EDB’s white paper Governing Agentic AI at Enterprise Speed.Max Romanenko is Chief Technology Officer at EDB.Sponsored articles are content produced by a company that is either paying for the post or has a business relationship with VentureBeat, and they’re always clearly marked. For more information, contact sales@venturebeat.com.

Nvidia had previously invested $235 million in Hugging Face during its Series D funding round in August 2023.

Forbes has opened nominations for America’s Top AI attorneys

A number of Wall Street banks, including Jefferies, raised their price targets for the cybersecurity’s stock following its record-breaking results.

Nvidia reported earnings on Aug. 26. Wall Street expected it to beat. Bank of America was not particularly interested in that part of the story.

The firm’s note, published ahead of the report, was really about one thing: whether the market had correctly priced the scale of what Nvidia has been doing with its balance sheet.

The answer, according to Bank of America, was no. The results Nvidia delivered made that argument harder to dismiss.

Bank of America’s Buy rating and $350 Nvidia price target

In a note shared with TheStreet on Aug. 25, analyst Vivek Arya reiterated a Buy rating and a $350 price target on Nvidia, implying roughly 64% upside from where the stock was trading that day.

The note’s title says everything about where the firm directed investor attention: “Balance sheet disclosures could speak louder than EPS beat.”

ALSO READ: NVIDIA Corp. Q2 2027 Earnings: Live Updates of $NVDA Earnings Call, Forecast

Nvidia delivered revenue of $96.2 billion for the second fiscal quarter, up 106% year over year and well above consensus expectations of roughly $92 billion, according to Nvidia’s official earnings release. Data center revenue came in at $89 billion, up 117% year over year. Gross margin held at 75%.

“AI has reached its inflection point. It’s doing useful work. Its tokens are productive and profitable. Now, compute is revenue,” Jensen Huang said on the earnings call. “And demand is accelerating.”

Nvidia also returned roughly $26 billion to shareholders in the quarter through buybacks and dividends, with approximately $99 billion remaining under its buyback authorization.

What the market had not priced properly, Bank of America argued, was the scale of what Nvidia had committed financially to keep the AI ecosystem running.

Nvidia’s $300 billion AI capital commitments and balance sheet risk

Nvidia is no longer just selling chips. It is increasingly financing the companies that buy them.

Bank of America estimated total capital commitments of approximately $300 billion, split between roughly $70 billion in direct equity investments and roughly $230 billion in residual value guarantees and backstops.

The equity side covers much of the AI supply chain, according to CNBC. The largest single check was $30 billion for OpenAI. Beyond that, Nvidia has invested in Anthropic, Safe Superintelligence, Intel, CoreWeave, Nebius, Lumentum, Coherent, Marvell, Synopsys, Nokia, Corning, and others.

More Nvidia:

Nvidia just made a move Wall Street wasn’t ready for

Nvidia just locked down deal that changes AI race

Nvidia stock is doing something it hasn’t done in years

The backstop commitments are the more unusual piece. A $105 billion backstop for SB Energy and a $125 billion backstop for a special-purpose vehicle involving six financial firms were both announced in August, as TheStreet reported.

Against those numbers, Bank of America estimated Nvidia could generate approximately $469 billion in free cash flow across the next two calendar years. The total committed capital represents roughly 64% of that figure.

In a scenario where AI demand stays strong, Nvidia may never have to absorb the full economic cost of those commitments at all.

The worst-case scenario, including purchase commitments and cloud service agreements, puts total obligations at approximately $500 billion. Even that figure represents only about 10% of Nvidia’s enterprise value.

Nvidia valuation de-rating and AMD comparison

Nvidia’s forward earnings multiple had collapsed by roughly 44% from its five-year historical median, putting it at less than half the forward multiple of AMD right now.

The most plausible explanation is that investors were treating the balance-sheet commitments as a tail risk and discounting accordingly.

Bank of America’s counter was that the market had de-rated Nvidia further than the actual worst-case math justified. The $96.2 billion revenue print and the $108 billion Q3 guidance give that counter argument more support than it had before the results landed, as TheStreet reported.

In a note shared with TheStreet on Aug. 25, analyst Vivek Arya reiterated a Buy rating and a $350 price target on Nvidia.Marcin/Getty Images

The Nvidia buyback case and the Apple comparison

The second catalyst Arya identified was buybacks. The comparison he drew was Apple.

Over more than a decade, Apple returned the vast majority of its free cash flow and retired roughly 43% of its shares outstanding. That buyback program put a floor under the stock and helped lift its valuation multiple from a low single-digit figure to the mid-twenties.

Nvidia currently returns roughly a third of its free cash flow to shareholders. Bank of America believes it could increase that to between half and three-quarters.

The $26 billion returned in Q2 alone, with $99 billion still in the buyback authorization, suggests the company already has the firepower to move in that direction.

By next year, Nvidia could be generating close to a billion dollars in free cash flow every single day. A larger buyback would give investors a second reason to hold the stock beyond the AI thesis alone.

What Nvidia’s earnings mean for its customer mix and investors

One data point the earnings call confirmed was the customer mix story. Amazon Web Services announced it will buy 2 million Nvidia GPUs and adopt the company’s new Vera CPU. That kind of hyperscaler commitment reinforces Nvidia’s customer breadth argument directly.

Nvidia’s CFO Colette Kress said capital expenditure among the top five hyperscalers is expected to increase to $1.3 trillion next year from $800 billion in 2026.

That is the demand backdrop Bank of America was betting on when it reiterated its Buy rating. The earnings report confirmed the bet was reasonable.

What happens to Nvidia’s balance-sheet commitments as that spending cycle plays out remains the more important question.

Related: Goldman Sachs spots huge twist ahead of Nvidia’s earnings

TheStreet aims to feature only the best products and services. If you buy something via one of our links, we may earn a commission.

Why we love this deal

After a nice hot shower or a soothing bubble bath, a relaxed, good mood can easily be ruined when you step out and into a grungy, damp towel. Although towels certainly don’t need consistent replacing, they do need to go in the trash when the fabric starts to pill and wear, when they don’t fully dry between showers, and when the color starts to fade. Not only is it more hygienic for you, but it also just makes getting clean far more enjoyable when there’s a warm, fluffy towel waiting for you when you get out. With the Redkiss 6-Piece Towel Set, you can get color, quality, and variety, all for 33% off.

The Redkiss 6-Piece Towel Set, which includes towels in three different sizes, is at a 30-day low price of just $20 at Amazon. This limited-time sale can save you $10 and give you a fresh batch of soft, thick towels to use in your bathrooms without paying a fortune. 

Redkiss 6-Piece Towel Set, $20 (was $30) at Amazon

Courtesy of Amazon

Shop at Amazon

Why do shoppers love it?

Made of microfiber coral fleece, these towels are exceptionally soft and durable. The double-sided brush has a textured, velvety feel that’s comfortable and gentle on even the most delicate skin. What’s even more key, however, is the material’s high water absorption. The fabric’s dense, split-fiber structure is able to absorb up to seven times its weight in water and dries you off much faster than natural fibers like cotton or standard microfibers. These towels are specifically engineered with advanced microfiber technology to absorb moisture twice as fast as standard cotton towels. 

Available in 10 colors, these microfiber towel sets include two bath towels, two hand towels, and two washcloths, giving you the perfect matching set for use in the shower and out of it. The fabric resists shredding and pilling, even after multiple washes, and the color is fade-resistant, so you don’t have to worry about bright towels becoming dull and drab after an extended period of use. The color also won’t transfer to your skin or clothes. 

Related: Amazon’s top-rated bed sheets that are as ‘soft as a cloud’ are only $13 for a limited time

Great for at-home use or on-the-go help at the gym or while traveling, these hotel-style versatile towels come in this variety pack, or you can opt for a pack of four bath towels to better suit your needs. 

Details to know

Material: Microfiber coral fleece.

Includes: The set includes two bath towels, two hand towels, and two washcloths. 

Colors: 10.

Care: Machine wash with cold water and a mild detergent. Tumble dry with low heat. 

Amazingly fluffy and soft, these towels are lightweight but super absorbent. Shoppers say they don’t leave you feeling humid or damp after drying off the way some towels do, and they maintain color and that fluffy feel very well even after multiple washes. They are also generously sized, especially the bath towels. You don’t have to worry about them being too short or not wrapping fully around your body. “They’re the best towels I’ve ever purchased,” one shopper said. 

Shop more deals 

Olanly Microfiber Bath Mat, $10 (was $15) at Amazon

Infinitee Xclusives 100% Ring-Spun Cotton Bath Towels (Pack of 4), $37 (was $49) at Amazon

American Soft Linen Luxury Turkish Towels (6-Piece), $38 (was $45) at Amazon

For only $20, you can wrap yourself in one of the fluffy towels from the Redkiss 6-Piece Towel Set and have that post-shower euphoria feeling continue even once you’re out from under the hot water. 

Planning for retirement isn’t just about saving money. It also includes putting a plan in place to ensure you don’t blow through that money too quickly once you save goodbye to the workforce.
Personal finance guru Dave Ramsey refers to each person as the CEO of their retirement. Developing strong money habits as soon as you can will help ensure that you don’t hurt your retirement savings once you’re in your 50s, 60s and beyond. Here are three behaviors Ramsey says could set you back.

Must Read

10 Smart Ways Seniors Are Earning Extra Money
Your Social Security Number Isn’t ‘Suspended’ — That Phone Call Is Probably a Scam
Silver Is Up 70% Since This Time Last Year — and These Gold IRA Companies are Handing Out Up to $25,000 of It

1. Treating debt payments as ‘normal’
Debt can be a key piece of a financial journey, such as a mortgage to buy a house or student loans to fund your education. But building up high-interest debt like credit card debt — and not focusing on paying off your debt in general — can chip away at your savings.
Ramsey is very debt-averse. He says that people should avoid debt as much as possible, and pay it off aggressively should they accrue it. Debt can make sense for some people’s plans, but a piece of advice you can glean from Ramsey’s approach is to not view debt payments as just a normal part of your budget. You shouldn’t get used to making the payments so much so that you aren’t focused on paying that debt off.
Ramsey believes people should aggressively pay off debt and that retiring with any of this type of debt can ruin their golden years. He says the best way to approach debt if you have it is to pay it off as quickly as possible, and become debt-free before retiring. That way, you have fewer expenses to worry about and are more prepared for any surprises.
2. Lifestyle creep with no written plan
Costs tend to climb over time, but some retirees may still be shocked by rising expenses that take place during their golden years. Home upgrades, frequent travel and impulse spending can increase monthly expenses if you aren’t careful, and some people spend so much money during retirement that their nest eggs get stretched too thin.
Ramsey regularly suggests creating a detailed budget, living below your means to avoid lifestyle creep and avoiding reckless spending. Every unplanned dollar you spend is another dollar that can’t work toward your retirement and compound in a portfolio.

Must Read

Retirees Are Doing These 10 Things to Add to Their Monthly Income
Your Social Security Number Isn’t ‘Suspended’ — That Phone Call Is Probably a Scam
Warren Buffett on Market Volatility — and 3 Ways You Can Take Advantage

3. Procrastinating your savings and having false confidence
Retirement planning is a long-term process, and saving it for right before you’re ready to step back from work can leave you financially vulnerable deep into your golden years. Expecting to rely solely on Social Security with no strategy for developing supplemental income, not maxing out your retirement accounts and getting deeper into debt can have consequences.
Ramsey says that saving for retirement shouldn’t be complicated, but it should be consistent.
He suggests saving at least 15% of your gross income — that is, your income before any taxes are taken out. One of the best moves you can do to achieve this is opening up a high-yield savings account, where the rates far exceed traditional banks. See Money’s best high-yield savings accounts to compare options.

My colleagues recently published the latest installment of Morningstar’s semiannual Active/Passive Barometer report. If you’re not familiar with it, the Barometer is a running tally of how many active funds have topped an average of passive funds in their category over various past periods. Here’s the money table from the most-recently published report. In general, the results are sobering for active funds. The longer the time period, the lower the odds of success—and those odds are very long indeed in a number of categories. Rattle AroundOne of the things you’ll also notice is that short-term success rates rattle around. To illustrate, here’s a comparison from the report showing how many active funds topped their benchmark over the years ended June 30, 2025 and June 30, 2026.Granted, it’s just one year. But one of the things that stood out was the way success rates improved in seven of the nine US equity categories, especially the small- and mid-cap peer groups. What caused that? Substance vs. Style It appears to stem mainly from differences between the active funds’ and benchmarks’ portfolio holdings. That is, the active funds’ style differed from the passives’ and those differences worked to the funds’ advantage over the year ended June 30, 2026.To illustrate, here’s a breakdown of the average active fund’s style exposures by Morningstar Category. I’ve used “Home” to denote the average weight in stocks that fell in the Morningstar Style Box region matching the fund’s category (for example, the average large-growth fund’s weight in stocks that plot in the large-growth box), “Neighboring” for stocks in regions bordering the “Home” region (for example, large-blend and mid-growth neighbor large-growth), and “Other” for stocks splayed elsewhere in the style box. It’s an interesting breakdown. The average active fund in five of the six large- and mid-cap categories has less in the “Home” style than the passive composite. That’s about what you’d expect—passive funds are usually thought to be more “style-pure” than actives, which enjoy more latitude to range around. What’s kind of surprising, though, is the average active fund in the other three categories has more exposure to the “Home” style than the passives. Why is that? The short answer is it seems to reflect a difference between how some of the leading small-cap index funds define “small-cap” and how Morningstar defines it for style-box purposes. For instance, the $183 billion Vanguard Small-Cap Index Fund tracks a benchmark that defines small-caps as stocks that account for 85% to 98% of cumulative US stock market capitalization, whereas the style box defines it as stocks that account for 90% to 100% of US market cap. In other words, the passive composite doesn’t dip as deep into small cap as many active funds do. Tilts and Success RatesThat quirk aside, the point is these differences can help explain why active fund success rates might have fluctuated higher or lower over a given period. To illustrate, here’s how the nine styles performed over the year ended June 30, 2026. I used the asset-weighted average return of the passive funds in each Morningstar category to proxy for each style’s return.We can approximate the aggregate impact of an active fund’s style tilts by taking the difference between its “Home” style’s return (such as the large-growth style’s return for a large-growth fund) and the returns of other styles to which it’s exposed (large blend, mid-growth, et cetera). Then, we multiply those differences by the fund’s over- or underweighting to those styles when compared with the passive composite. Here’s an example for the average active mid-cap value fund:StyleOver-/ UnderweightingStyle-Relative Performance (vs. Passive Mid-Value)ImpactLarge Value-8.6%4.4%-0.4%Large Blend-0.7%2.8%0.0%Large Growth0.7%5.4%0.0%Mid-Cap Blend-4.1%3.8%-0.2%Mid-Cap Growth1.9%-5.8%-0.1%Small Value10.6%10.0%1.1%Small Blend10.7%16.0%1.7%Small Growth2.5%15.5%0.4%Total2.5%In summary, the average active mid-cap value fund had heavier exposure to the small-cap styles and, correspondingly, less weight in large- and mid-caps. That was a boon to performance because small-caps outgained the passive mid-cap value average to a far greater extent than the large- and mid-cap styles did. A simple attribution like this can yield a sense of how much benefit an active fund might have derived, or the penalty it incurred, versus the index from its style tilts. Here’s how it looked for the average active fund in all nine categories:Active US stock funds reaped rewards from their style tilts in seven of the nine categories, with active small- and mid-cap funds deriving the most significant benefits. That largely lines up with the year-over-year changes in success rates for active funds in these categories. Given this is just a single year-over-year comparison, I decided to expand the analysis to assess the relationship between payoffs from style tilts and changes in active-fund success rates over four additional 12-month periods: June 2021 to June 2022, June 2022 to June 2023, June 2023 to June 2024, and June 2024 to June 2025. Here’s a scatterplot showing the relationship for all nine US equity categories over the five total year-over-year periods: It’s not a perfect fit, but the direction of success rates—higher or lower—corresponded with the payoff from style tilts more than 70% of the time. (The attribution I’ve walked through is better known as “Dunn’s Law,” a concept my former colleague John Rekenthaler has been writing about for years, most recently in this piece he wrote around a decade ago.)Micro and MomoOf course, the style box doesn’t tell the whole story. There are other factors besides size and valuation, or at least some dimensions the style box doesn’t break out. Take micro-cap stocks, for instance. The style box doesn’t separately classify micro-caps. Rather, it lumps them in with all other small-caps. Thus, two funds could have similar small-cap stakes, but one might have a much larger micro-cap weighting than the other. When micro-caps and larger small-caps perform alike, that difference doesn’t matter much. But when they diverge, as they did over the year ended June 30, 2026, it can be very consequential.(My colleague Dan Culloton recently published a nice rundown of the recent rally in micro-caps and some other factors, which you can find here.)To assess how that affected trends in active-fund success rates, I tallied up every active US stock fund’s micro-cap weighting as of June 30, 2025, and then compared it to the micro-cap stake of the passive composite for its category. Then I assigned every active fund to a bucket based on the magnitude of its micro-cap over- or underweighting. Finally, I counted up the number of active funds in each bucket that beat the passive composite over the year ended June 30, 2026.In summary, the more overweight an active fund was in micro-cap stocks, the likelier it was to beat its passive composite, and the opposite for active funds that were underweight. There’s also the momentum factor to consider. Price momentum, which isn’t incorporated into two-factor style-box configuration, holds that investors derive a risk “premium” from buying stocks that have risen in price while shorting those that have fallen. Morningstar has built its own proprietary factor model and ranks each fund based on an array of factors including momentum. Given this, I was able to compile every active US equity fund’s momentum factor ranking over the year ended June 30, 2026. Using those rankings, I assigned all of the funds to quintiles and then measured the percentage of funds in each quintile that outperformed their passive composite.As with micro-cap stocks, the more exposure to momentum an active fund had in the 12 months ended June 30, 2026, the likelier it was to succeed, owing to the fact that momentum had a very strong year. Taken together, these factors were very important to active US equity fund success rates over this period, as shown below. Active US stock funds with lots of micro-cap and momentum exposure succeeded at far higher rates than the norm. TakeawaysSignal vs. Noise: If it’s not already clear, short-term fluctuations in active-fund success rates tell you more about holdings differences than they do manager skill. To be fair to managers who are picking stocks bottom-up in a way that aligns to a particular style, any successful payoff owes to the choices they made. But it is also hard to separate noise from signal over such short intervals, with headwinds quickly becoming tailwinds and vice versa.Heroes and Goats: Though the stylistic trends worked in more active funds’ favor over the year ended June 30, 2026, that doesn’t make them heroes. And by the same token, when their favored styles were out of vogue in past periods when they slumped, it didn’t mean they were goats. Smart or Dumb? Along those lines, it’s silly to ascribe any of these changes in short-term success rates to manager acumen. No question, fund managers are some of the most talented, highly trained, and accomplished individuals in finance. But it’s not like they flake out in one 12-month period and then come to their senses the next.Death to the Stockpicker’s Market: It just might be the hoariest cliché in investing circles—the stockpicker’s market. John Rekenthaler once called the adage “as common as flies at a picnic.” He was being kind. As the kids say, it’s cringe, and the industry can’t rid it from its vocabulary soon enough.Switched OnHere are other things I’m reading, watching, and listening to:Beware “innovative” ETFsSEC charges SPV manager with fraudYeah, hockey ETFs“Everything Looked Like Me”ObsessionPhoebe Bridgers’ Lost WeekendDon’t Be a StrangerI love hearing from you. Have some feedback? An angle for an article? Email me at jeffrey.ptak@morningstar.com. If you’re so inclined, you can also follow me on Twitter/X at @syouth1, and I do some odds-and-ends writing on a Substack called Basis Pointing.