It’s time to take the pulse of Wall Street. JPMorgan, Goldman Sachs, Citigroup and Wells Fargo are set to report results tomorrow, as America’s big banks kick off the first wave of the third quarter earnings season. Morgan Stanley and Bank of America are scheduled to follow on Wednesday.

They have a challenge ahead of them: meeting an awfully high bar set by worried investors who have dumped shares in recent weeks even though the economy and credit quality remain on sound footing.

Theory vs. Practice

Lately, investors are Dostoevsky-level pessimistic about lenders in today’s high yield environment. The bank sector is among the worst-performing on markets, with the iShares S&P U.S. Banks UCITS ETF down 7.7% in the last month. Compare that to the S&P 500 and Nasdaq Composite, where the mood has been more like an upbeat Mitch Albom paperback: Both hit all-time highs last week, driven by tech (and not much else).

Higher interest rate environments are generally seen as good for banks, but that’s not always the case. “In theory, higher rates = higher loan interest = greater bank profit,” explained analysts at broker CMC Markets. “In practice, if rates rise too quickly and remain high for too long, customers stop taking out mortgages and investment loans, and existing borrowers struggle to repay. A bank that fails to benefit from a higher margin starts losing volume and must set aside provisions for bad debts.” Those fears have been compounded by the risk of inflation and rising energy prices, which already have US consumers expressing record levels of frustration. While household spending has been incredibly resilient, any slowdown would impact banks’ balance sheets by hurting their fee incomes.

It’s not just consumers in the crosshairs. JPMorgan CEO Jamie Dimon has spent months warning that companies may be unprepared for the one-two punch of higher prices and higher borrowing costs. “I think there’s a risk that inflation is sticky, that rates continue to go up and so, for corporations, whether you’re leveraged or not, when you have to [refinance] or borrow money are you prepared for higher credit spreads?” he stressed in a podcast interview last week. With that in mind, this week’s earnings will put capital markets activity in the spotlight:

• Heading into earnings season, Oppenheimer analysts said they expect “plain vanilla” banking activity — loans, credit and costs — to come in just fine. But they warned of capital markets income at the low end of the market’s expectations and, expecting a dealmaking slowdown, slashed their aggregate investment banking revenue estimate by 9.5%.
• Raymond James analysts expect sound fundamentals, too, and are more optimistic about quarterly earnings beats (Oppenheimer expects only JPMorgan to top the Street’s expectations). But they emphasized that executives will have to demonstrate “credible and achievable” 2027 targets that account for higher interest rates and their impacts to quell investor anxiety.

Where There’s Smoke: Evidence in support of Dimon’s warning has already materialized in credit markets. JPMorgan reported earlier this month that the value of deeply distressed US loans rose to $65 billion in the third quarter, up from $40 billion a year ago and the highest in more than six years.

The arrival of Reflection is cause for self-reflection for the AI frontier labs. The Nvidia-backed US startup said last week that its soon-to-be-released open-weight model Beam rivals the best of the best in the open-source field. It’s big news for domestic AI users who don’t want to pay for costly top-flight closed models from frontier labs but also don’t want to rely on an open-weight ecosystem that hails from China.

Made in the USA

Beam’s launch is a long time coming. Founded in 2024 by former Google DeepMind researchers Misha Laskin and Ioannis ​Antonoglou, it emerged from stealth last year saying it planned to build powerful open-weight tools intended for business users, with an emphasis on vibe coding and agents. It had massive institutional backing out the gate. Nvidia, a champion of open-source AI, was among the investors in a $2 billion funding round last year that valued Reflection at $8 billion; the company has held talks this year for another fundraising round that would value it at $25 billion, The Wall Street Journal previously reported. It’s also been at the center of the White House’s AI diplomacy plans as it aims to reduce ally reliance on Chinese AI; in March, the US government brokered a deal that will see Reflection aid South Korea in developing its own sovereign AI ecosystem.

Last week, Reflection said Beam compares favorably on advanced reasoning benchmarks to GLM-5.2, the flagship open-weight model by Chinese player Z.AI. That would be better than any US open-weight player, though a tick behind the top Chinese models from the likes of DeepSeek and Moonshot. Crucially, Reflection says Beam requires comparatively less computing power than its Chinese rivals. Those rivals, by the way, are not slowing down

• Bloomberg reported last week that DeepSeek is nearing at least 80 billion yuan ($12 billion) in a new fundraising round, ahead of a planned IPO to come sometime next year. The firm had initially sought just 50 billion yuan for the fundraising round.
• Moonshot, meanwhile, closed a fundraising round that valued it at $50 billion, according to a Bloomberg report. The company is also en route to an early 2027 IPO.

Cheap Shot: Still, the worm may be turning on cheaper open-weight models, which have gained prominence and market share this year in the US. One recent study spanning 6,800 tasks from researchers at Sanford, Carnegie Melon, Berkeley and Microsoft found that cheaper models ended up costing more than more expensive models for one-third of tasks because they took longer to perform and sometimes ended in failure. It’s a lesson we thought no one still had to learn: giving the intern too many responsibilities just creates more work for everyone.

Financial influencers, or finfluencers, are moving markets. A study of 15 paid Discord servers found that finfluencers’ legions of followers temporarily pushed up prices when they rushed to buy recommended options.

The users can pay hundreds of dollars monthly for access to Discord servers led by admins that offer investing recommendations like “BTO $EBAY 72c 05/09 @0.30.” That’s shorthand for buying an eBay call option set to expire in two days. Subscribers act on advice like this within minutes, the academic study found, causing a ripple in prices that can last for a couple of hours. The eBay option rose 30% within five to seven minutes of being recommended on the server.

The study found that most trades in the chat were speculative options with median expirations of three days. Subscribers analyzed in the research responded the strongest to the most speculative recommendations.

Shifting Trust

WallStreetBets, the famous Reddit forum, introduced a new generation of traders that actively participated in markets, often with high-risk trades. These retail traders followed each other’s advice and together, they moved markets, catalyzing the steep climbs GameStop and AMC shares experienced in 2021.

Finfluencers evolved from that growing sentiment — that anyone can offer financial advice worth following and anyone can grow wealth with active investing. As they amassed followings online, some started offering paid subscriptions for access to exclusive advice, no Series 7 needed:

• About two in 100 influencers offering financial advice were qualified to do so in a study of 1,700 posts with more than 692 million combined views, consultancy Legalaes found. Nearly a third of the posts were misleading. There are exceptions: Arguably the most famous influencer, GameStop ride-or-die Keith Gill AKA Roaring Kitty, was a licensed broker when he started his channels (that status was revoked later on).
• Offering financial advice, especially in a paid server, is legally murky territory. But so far, it’s largely flown under regulators’ radar.

Cherry-Picking: A financial influencer running a $100 a month Discord server, who spoke to The Daily Upside on the condition of anonymity, said he frequently sees accounts market themselves by pointing to rare large wins. A screenshot of a big win brings in new subscribers, even if most subscribers won’t see those kinds of returns. From his POV, investors should see super short-term recommendations (like the eBay call) as high-risk gimmicks, and ultimately, red flags.