America pays more in interest than defense. Long-term rates hit their highest since 2007. So last week, Scott Bessent tried to fix it himself... Rates dropped from 5.34% to 5.18% the next morning. What he called 'liquidity' was a job no Secretary has ever done: Setting interest rates on long-term debt is the Fed's job. That's why the central bank is independent from Washington. But the Fed has not cut fast enough for Washington's needs. So Bessent started doing it another way. The mechanics are simple. September 9th, Treasury announced a major expansion of its buyback program. Buybacks of 10 to 30-year bonds would at least double. From $2 billion per operation to at least $4 billion. Bessent said the goal was liquidity in older, less-traded bonds. Wall Street took that at face value for about a day. Then people started doing the math. When Treasury buys bonds, bond prices go up. When prices go up, yields go down. Every new bond the Treasury issues is priced off that yield. Lower prevailing yields mean cheaper borrowing on the next auction. For a government paying $1.3 trillion in interest, that math matters. But the Treasury does not talk about the funding side. Bessent cannot create money to fund these buybacks. Only the Fed has that power. So the Treasury raises the cash by issuing more short-term Treasury bills. Net effect: the government swaps long-term debt for short-term debt. Total debt does not go down. The annual interest bill does not go down either. But long-term bond supply gets smaller. And short-term bill supply gets bigger. Fewer long bonds in the market push long yields down. This is called operation twist. The Fed ran the same play in 2011 and 2012. But the Fed was the buyer both times. The Treasury has never done this on its own before. This is not the Treasury versus the Federal Reserve. It is the Treasury versus the bond market. The Treasury market is worth more than $30 trillion. Twenty and thirty-year bonds alone total trillions more. A $4 billion buyback nudges yields at the margin. It cannot overpower the market if investors want higher returns. And investors have real reasons to want them. Inflation is rising again on energy prices and tariffs. The federal deficit shows no sign of shrinking. Treasury has to sell trillions in new debt every year. Every buyer wants more compensation for that risk. These buyers are sometimes called bond vigilantes. When they win, yields rise regardless of what the Treasury does. So the buyback expansion is a signal, not a fix. The signal is that Washington will act when long yields spike. That is called yield curve control. Japan has run it for years and cannot get out. Every step in that direction reduces the market's role in pricing. And increases the government's role in setting it. Retail investors read the headline and moved on. Institutions read the mechanism and understood the stakes. The Federal Reserve is supposed to set interest rates. The Treasury is now quietly doing part of that job. And nothing about that fixes the underlying debt problem.

OpenAI and Anthropic are heading for two of the biggest IPOs ever. Together they're driving $725 BILLION in AI infrastructure spending this year. 50% of US businesses now pay them for AI tools. But their top 1% customers just started walking away... Here's how bad the impact of that actually is: For every other software category, revenue is spread across a normal curve. You need the top 10% or 20% to reach 80% of revenue. In enterprise AI, 1% is doing all the work. And within that 1%, the customers look almost identical. They are mostly AI startups, high-growth tech firms, and coding shops. The biggest AI customers are companies built to sell more AI. Cursor pays Anthropic to power its coding assistant. Half the AI economy runs on invoices between AI companies. When one cohort tightens, the rest tighten with them. The top spenders have already started tightening. The top 1% cut per-employee AI spend from $8,000 to $7,200. Some of that is summer seasonality. Most of it is not. Token usage actually kept rising in the same period. Businesses are using more AI while paying less for it. The gap is where the real story sits. OpenAI and Anthropic have spent the last six months cutting prices. The effective price per million tokens dropped from $1.15 to $0.68. That's a 41% collapse in six months. And enterprises are quietly walking away from the frontier models. Frontier token share fell from 53% in August to 45% last week. Companies now default employees to Sonnet and GPT-5.6 Terra instead. They save money and get 90% of the quality. Meanwhile, Big Tech is spending like enterprise AI demand keeps compounding. Big Tech will spend $725 billion on AI capex this year. That's up 77% from last year, headed toward a trillion by 2027. All of that spending assumes AI revenue keeps growing exponentially. But most of that revenue flows through OpenAI and Anthropic. And their most valuable customers just switched to the cheapest tier. Nvidia's data center revenue hit $75 billion in a single quarter. Almost all of it came from those same four hyperscalers. The entire AI trade rests on that 1% at the top. The stock market is pricing AI infrastructure at record valuations. The underlying enterprise customer is choosing the discount rack. Watch the frontier token share and the top 1% spend numbers. Those two lines break the AI trade if they keep falling.

America's national debt just crossed $40 TRILLION. Interest payments alone are on track to top $1 TRILLION this year. And last month, Treasury Secretary Scott Bessent started day-trading the bond market. He used the government's checking account to buy back its own debt. The reason he did it is stranger than it sounds: The debt didn't grow this fast because we suddenly started borrowing more. It grew this fast because fewer people wanted to lend to us. For decades, three groups reliably bought Treasury bonds. Social Security trust funds held about $8 trillion of it. Corporate pension funds held hundreds of billions more. Foreign governments parked their reserves in it. All three are shrinking at the same time. Social Security is now paying out more than it takes in. Private pensions have mostly been replaced by 401(k)s. Foreign holders are pulling back after watching what happened to Russia. China alone has cut its Treasury holdings to an 18-year low. Big companies that used to park cash in Treasuries are pivoting hard. They are issuing their own bonds to fund AI data centers. Amazon, Alphabet, Meta, and Oracle now compete with the Treasury for lenders. Fewer bidders means the Treasury has to pay higher interest rates. That's the only way to fill an auction. Higher Treasury yields push up the 10-year rate. The 10-year sets the price on mortgages and corporate loans. So voters and businesses feel it directly in their monthly payments. There's also a slower fuse burning underneath all of this. The average rate on the current debt pile is about 3.1%. But new bonds issued this year are paying between 3.8% and 5.3%. Every time old debt rolls over, the government's interest bill jumps. Which is why Bessent tried something clever last month. He used Treasury cash to buy back old long-term bonds himself. He also bought Japanese yen to prop up Japan's demand at auctions. The idea was to fake extra demand and pull yields back down. Rates dipped for one day. Then they climbed right back up. The market can see what is happening. The government is quietly pretending to be its own customer. There are only four ways this debt ever gets resolved. Grow faster than the interest, raise taxes, cut spending, or inflate. The first three are politically off the table right now. That leaves the fourth. The Fed will not let the government default on its bonds. When real buyers vanish, the Fed prints and buys the bonds itself. That saves the debt. It does not save the dollar. Long-term lenders already know how this ends. That's why they demand higher yields to compensate for future inflation. Every rate cut, buyback, and tax cut just widens the gap. Retail savers are still being told bonds are the safe part. Institutions are quietly rotating into gold, foreign debt, and hard assets.

The company behind your COVID vaccine just scored a win against cancer. Its stock went up almost 200% in a single day! One year ago, it had crashed more than 90%.. Then ONE trial added $45 BILLION to its value. Here's the real reason behind this explosion: Moderna built its entire company on one vaccine. That vaccine was for COVID. When the pandemic faded, so did the demand. Sales collapsed, and the stock followed. By 2025, it had fallen from over 400 dollars to under 30. Wall Street had mostly stopped paying attention. The company still had one thing left to prove. Its mRNA technology could work on more than one virus. For years it had been testing that idea on cancer. Melanoma is one of the deadliest forms of skin cancer. More than 230,000 people will be diagnosed with it this year. Moderna and Merck built a vaccine made for each patient's own tumor. Paired with an existing drug, it aimed to stop the cancer returning. In August, the trial results came in. The combination cut the risk of the cancer returning or spreading. It was the first positive late-stage result for an mRNA cancer treatment. That is the headline everyone read. Here is the part fewer people noticed: Merck is a 333 billion dollar company. This single drug is one of dozens in its pipeline. Moderna, before the announcement, was worth about 25 billion dollars total. For Merck, this was good news. For Moderna, this was the entire company's future arriving at once. That is why one stock moved 12 percent, and the other 177. The size of the reaction was not about the drug. It was about how much of the company depended on it. Someone was on the right side of this before the headline hit. Some investors held Moderna all the way to the bottom. They were not doing it for comfort. They were betting the technology had a future beyond one vaccine. Most people who bought Moderna during the pandemic sold long before this. They watched it collapse and walked away from the story. The ones still holding on August 19 were paid for staying. This is not really a story about one drug trial. It is about what happens when a stock is priced for zero. When a market gives up on a company, good news hits harder. There is no expectation left to disappoint. Every surprise is pure upside. That is a different game than a company like Merck. Merck was already priced for success, so success barely moved the needle. Moderna was priced for failure, so a single win changed everything. You will not see this kind of move very often. It requires a stock that has been left for dead. And news big enough to change the entire story at once. The traders who understood this were not reading the vaccine data. They were reading the balance sheet and the size of the bet. The signal was in how much room there was to be surprised. Retail investors watched the crash and moved on. The ones who stayed got paid for a bet others abandoned. That's the whole game. Surmount builds rules-based strategies that trade the filings instead of the headlines.

MongoDB was trading at $434 a share on Tuesday. Next day, it crashed 14 percent! The company had just beaten every earnings estimate on Wall Street. But the reason for crashed is much simple: MongoDB sells database software that big companies run their apps on. Its cloud product is the engine everyone watches. Sales there still grew 30 percent, to 772 million dollars. Fast by any normal standard. Management even raised its guidance for the rest of the year. But for months that engine had been speeding up. This quarter it grew at the same pace as before. Still fast, but no longer getting faster. That one change is what emptied the room. The forecast for next quarter came in softer than people hoped. And part of the profit was not really cash. The company counted stock paid to staff as earnings. That is a real cost, even when no money leaves. Strip it out, and the quarter looked thinner than the headline. None of this showed up in the number everyone quoted. It sat one layer down, in the parts nobody reads. The stock was priced for perfection before any of it landed. Almost 60 times next year's earnings. At that price, a good quarter is a letdown. You must beat the numbers and prove you are still speeding up. MongoDB proved it was big. It did not prove it was accelerating. So the people holding it for the story left. Someone was on the other side of that trade. The funds that sold into the beat locked in the high. They read the same report as everyone else. They just turned to the page retail skipped. That page cost nothing to open. It just took the patience to read past the headline. This was not a one-time fluke. The same company did this once before. In early 2025 it beat again and still fell over 20 percent. The reason then was the same as now. Soft guidance, hiding under a strong headline. The market has been repeating this for over a year. Most people keep reading only the top line. This matters far beyond one stock. In a market this expensive, good is no longer good enough. The smallest sign of slowing gets punished the same day. MongoDB was not even alone that week. Other companies beat their numbers and still got sold. A beat is only the first half of the story. The guidance and the cash are the second half. The headline shows you the first half. The filing shows you the second. You saw a beat and thought about buying the dip. The real reason it fell was one line you never opened. By the time the drop makes sense, the move is done. The traders who saw it coming were not reading the headline. They were reading the guidance line and the cash. The signal was in the filing, not the news. Retail read the beat and bought the story. Institutions read the guidance and sold it. That's the whole game. Surmount builds rules-based strategies that trade the filings instead of the headlines.

A war you stopped watching just spiked the price of bread. Wheat jumped so fast the market had to halt trading. All of it started 6000 miles from your kitchen. But the path from that war to your table is shorter than you think: Russia and Ukraine grow a huge share of the world's grain. Together they handle about a third of all wheat traded. Most of it sails out through the Black Sea. That sea is now a war zone again. Whole regions depend on that grain to eat. For some countries, cheap bread keeps the peace. Russian missiles have hit Ukraine's main grain ports. Ukrainian drones have shut down a key shipping route. Grain leaving the region fell more than 40% in a year. Drought across the north made a tight market tighter. And the grain cannot simply come from somewhere else. Russia's other ports sit too far from the fields. The fields are still there, but the exits are blocked. A crop is only worth something if it can ship. So traders panicked, and prices exploded. Wheat rose so fast it hit its daily limit. Corn touched its highest level in three years. Soybeans jumped to a record for that contract. This was not a one-day scare. Wheat is up more than 20% in just a month. Corn and soybeans have each climbed over 20% this year. Someone is on the winning side of this move. Grain farmers and traders are watching prices soar. The same shock that raises your bill pads their profit. But grain is not just about bread. It feeds almost everything else you eat. It is the flour in your pasta and your tortillas. It is the feed behind every burger and every egg. It is the corn syrup inside most of your snacks. Almost nothing in your cart escapes it. When grain climbs, the whole grocery aisle follows. We watched this exact story play out in 2022. Prices jumped worldwide after the first invasion. The market has simply hit rewind. Wheat is already up more than 50% this year. That number has not reached your receipt yet. Food prices move slowly, like a wave. The grain spikes first, months before the shelf does. By the time bread costs more, the cause is old news. You will blame the store or the season. The real reason will be a map you never checked. The headlines are locked on AI and the Fed. A grain shock does not trend on your timeline. It shows up weeks later, at the register. The traders who saw this were not reading war coverage. They were watching ships that never left port. The signal was on the water, not the news. The biggest price shocks start far from the news you follow. A port you cannot place on a map sets your food budget. Markets price the war today. Your kitchen pays for it later. Retail watched the war as a headline. The market traded it as your next grocery bill. That's the whole game.

Meta was just accused of designing Instagram to addict your kids. Making that lawsuit disappear cost up to $16.7 BILLION. A hit that big should have crushed the stock. But what happened next surprised me… To understand it, let’s start with the case itself: The accusations are about as dark as they come. Twenty nine states said the apps were built to hook kids. They said Meta collected data on children under thirteen. The company denied doing anything wrong. It even disputes that social media addiction is a real disorder. The trial was already underway in a California court. Meta's own founder was expected to take the stand. Then the company settled instead. The deal is about more than money. Meta must add daily time limits for teen accounts. It must add overnight blocks and stronger age checks. It must expand the controls that parents can use. A hit like this usually scares investors away. It hints at years of lawsuits and bad press. Big penalties are supposed to punish the stock. That is the reaction the crowd expected. But look closer at the actual numbers: Meta will take a charge of about $10 billion. Meta pulls in tens of billions in profit each year. A ten billion dollar charge is just a bad quarter. For most companies that number would be crushing. For this one, the pain barely shows up. A fine only bites if it changes how a company acts. For one this rich, it almost never does. Part of the bill is not even certain. About $5 billion depends on what YouTube and TikTok do. The rest is spread across the next ten years. None of it lands as one painful hit. The blow is softened before it even arrives. There is a second reason the stock held up. The lawsuit had been hanging over Meta for months. A settlement puts a ceiling on the damage. Before it, the loss could have been anything. After it, the number is fixed and known. Investors hate uncertainty more than they hate fines. Markets will pay up for a risk that is finally capped. This was also not Meta's first payout this month. A separate case in New Mexico cost it hundreds of millions. The headlines screamed about nearly a seventeen billion dollar penalty. That number feels enormous to a normal person. To a company this size, it barely registers. The market did the math the media skipped. One number went viral across every feed. The number that actually mattered went unspoken. Wall Street has watched this movie many times. A giant pays a record fine and moves on. The business barely breaks stride. This is how giant firms absorb their worst days. The fine that shocks the public is trivial to them. Scale turns a scandal into a line item. The size of a headline is not the size of the damage. Retail read the outrage in the headline. Institutions read the size of the check. That's the whole game. Surmount builds automated strategies that follow the data, not the noise. Start for free and let the signals lead.

Amazon, Google, and Meta borrowed $220 BILLION this year. Last year the same three borrowed $12 billion. That flood of debt did not stay in tech. Here‘s the ACTUAL impact of the AI debt you are experiencing without realizing it: Everyone treats AI as a stock market story. Who is winning vs who is losing. That framing hides where the real cost lands. Building AI is stunningly expensive. It takes data centers, chips, and huge amounts of power. The biggest tech companies cannot pay for all of it in cash. So they do what anyone short on cash does. They borrow enormous amounts of money. This year those companies sold $220 billion in bonds. The same stretch last year, it was $12 billion. That is not a small jump. That is close to seventeen times more debt. The whole market for company bonds is at a record pace. The AI names are the loudest voice in that crowd. This is happening while the government borrows record sums too. Two giant borrowers are now fighting for the same lenders. Every bond competes for the same pool of money. When these giants borrow this much, buyers get picky. They demand higher returns to soak up all that debt. That pushes interest rates up across the whole market. Government bonds get dragged higher too. Bank of America did the math on this. The borrowing helped push the ten year rate higher. The jump was about a third of one percent this year. Nomura found something even more striking. Big tech now borrows about a quarter of what the government does. That share is five times bigger than last year. Even the largest bond buyers are starting to push back. They call it indigestion in the market for tech debt. The more these firms borrow, the more they must pay. And everyone borrowing after them pays more too. The ten year rate is the price of money for everyone. Your mortgage rate closely follows it. So does your car loan and your credit line. When AI firms crowd the debt market, your loans cost more. A rate higher by a fraction still adds up fast. Spread over thirty years, it can cost real money. Think about that the next time a rate quote arrives. The boom you feel priced out of is already billing you. The headlines cover AI as a race between stocks. They rarely connect it to the rate on your house. That connection is the part that reaches regular people. And almost no one is talking about it. No bill ever arrives with the word AI on it. The cost just hides inside a slightly higher number. The biggest forces in markets hide where you are not looking. You watched the AI tickers climb and fall. The bond market quietly reset the cost of your life. That is how real money usually moves. It travels through the plumbing beneath the headlines. The loudest story is rarely the one that costs you. Retail watched who was winning the AI race. The bond market sent everyone the bill. That's the whole game. Surmount builds automated strategies that follow the data, not the noise. Start for free and let the signals lead.

US companies just posted their best profits since 2021. The stock market should be soaring, but... They just had their first down week in a month. This makes no sense on the surface. But there are two reasons hiding under the headline: First, look at how good the numbers were. S&P 500 earnings grew about 50% last quarter. That is the fastest growth since 2021. Most companies beat what analysts expected. 86% of them beat, well above the usual rate. It looked like a blowout quarter on paper. But that number hides something. Two companies did most of the heavy lifting. Strip out Alphabet and Amazon, and growth drops to 32%. Those two alone added 18 points to the total. The gains were not spread across the market. That is a lot of weight on two names. A big part came from gains on other companies they own. One booked a $98 billion gain on paper. The other booked $53 billion the same way. Those were paper gains, not real sales. It is like your house going up in value. You feel richer, but you earned nothing new. It was a one-time accounting boost. It was not money from selling more products. Real growth was there, just not fifty percent of it. So maybe the market saw through the hype? Not quite. Even the real number was strong. Revenue grew 15%, the fastest in years. Profit margins hit a record high. Even without those two giants, margins set a record. The core business was genuinely healthy. Demand was real and customers kept spending. So that is not why stocks fell. So what actually pulled them down? Two things were working against the good news. Borrowing costs have been climbing fast. Higher rates make every stock worth a little less. Stocks already trade above their long-term average. And the good news was already priced in. Investors expected a great quarter and got one. Analysts had even raised their targets going in. Even companies that beat barely moved higher. Good news everyone expects is already in the price. The best quarter in years still was not enough. Every other 50% jump this century came after a collapse. This time, earnings are already at record highs. This is the trap with headline numbers. A big number can hide a weaker story. A record quarter does not guarantee a rising stock. Even a strong story can already be priced in. The 50% looked amazing on the surface. Underneath, a third of it was two companies. Part of that was paper gains, not cash. And the rest was already expected. The headline and the real story rarely match. Great news can be a terrible time to buy. Retail saw record profits and got confused. The pros read what was actually inside them. That's the whole game. Surmount builds strategies on what the numbers really say. Start for free and trade the facts, not the headline.

A guy built his first robot with $30 in parts. This week it went public in China and soared 629%. His company is now worth more than any US robot maker. You have probably seen its robots online. They do backflips, run, and dance in viral clips. The demand for shares was almost unreal… Almost 10 million people applied to buy in. Your odds of getting shares were about 1 in 5,000. When it finally opened, the stock exploded. It jumped 629% above its offer price. That valued the company near $66 billion. It closed its first day worth around $53 billion. Now put that number in context. Unitree is worth more than Baidu. It is worth more than JD(.)com. It is worth more than any US robot company. America's biggest is Figure AI at $39 billion. Unitree blew past it in a single day. One early backer saw its stake jump about 70 times. Now look at who piled into this deal. DeepSeek, the AI company, invested. So did Tencent, Alibaba, and Ant Group. Those are the biggest names in Chinese tech. DeepSeek even agreed to build AI models with Unitree. Even AI software firms want a piece of robots. Now for the part that actually matters. For two years, the AI trade was about chips. It was all data centers and Nvidia. That was AI living inside a screen. This is something completely different. This is AI stepping into the physical world. Robots that walk, lift, and work. The industry calls it embodied AI. And the first big public bet on it is Chinese. It now competes directly with Tesla's Optimus robot. China now makes most of the world's humanoid robots. One bank just doubled its shipment forecast for this year. The debut landed during China's biggest robot conference. This is not just hype and viral clips. Unitree shipped more humanoid robots than anyone last year. Its revenue grew more than tenfold in two years. That is about $235 million in a single year. The company even turns a profit, which is rare here. Do not mistake this for a sure thing. The stock is priced for perfection. Its profit actually fell 52% last quarter. It trades at over 1,000 times earnings. That is a wild price for any company. But the direction of the money is the signal. Capital is moving from screens to machines. And most of us cannot even buy this stock. It lists on an exchange closed to outsiders. So chasing this one is not the lesson. The lesson is where the smart money is heading. The next phase of AI is physical. The winners are being chosen right now. You do not need this one stock to play it. You need a system that follows the shift. Retail chased a stock it cannot own. The shift underneath is the real trade. That's the whole game. Surmount turns big shifts like this into rules-based strategies. Start for free and invest on data, not FOMO.