Every month we take five stories that actually moved the economy and break down the mechanism behind them. Not just what happened, but why it happened the way it did.
July was a month where conviction got punished, a government's ledger did something interesting, a mountain pass reopened after six years, hiring stopped following the old rules, and fifty billion dollars sat frozen because of a war ten thousand kilometers away. Let's get into it.
Story one: the fund that got the future right, and still lost everything
Leopold Aschenbrenner was twenty four. He used to work at OpenAI. In 2024 he wrote a 165 page essay arguing that AI was going to keep scaling, fast, and that almost nobody was positioned for it. He raised 225 million dollars to bet on that essay. By June this year, his fund was up 439 percent for the year. Assets peaked near 45 billion dollars.
Then in July, AI infrastructure stocks turned. And because Aschenbrenner was running his positions on close to four times leverage, a correction that would have dented an ordinary portfolio wiped his out. His prime brokers issued margin calls. He didn't have a choice. He sold his entire public stock portfolio, longs and shorts, in a single block, overnight, to Citadel.
Here's the part worth sitting with. He was right. AI infrastructure probably is a massive investment theme. That's not the lesson. The lesson is that leverage doesn't care if your thesis is correct. It only cares whether you can survive the next four hours. A portfolio that can't absorb a sharp drawdown isn't a bold bet, it's a countdown timer. Aschenbrenner still holds a private stake in Anthropic worth around five billion dollars, and he kept that, because private holdings can't be marked to market every day and margin called. Public ones can. That single distinction is the whole story.
Story two: India's tax collections just told you something the headlines won't
Gross GST collections in July came in at 2.11 lakh crore rupees, up 15.4 percent from last year. That number alone is boring. Here's what's not boring. Import related revenue jumped almost 29 percent, and domestic collections held up too.
Rewind a year. The government cut GST rates on hundreds of everyday items, the kind of move that on paper should shrink tax revenue, because you're charging less per transaction. Instead, revenue went up. Why? Because cheaper goods mean more people buy them, and a wider base of transactions can raise more money than a narrow base taxed more heavily. It's the same logic that sits behind every argument about tax cuts paying for themselves, except this time the data actually backs it up. That's a genuinely rare thing to see play out in real numbers instead of a textbook diagram.
Story three: a trade route that shut for six years just reopened
Shipki La is a mountain pass in Himachal Pradesh that used to connect India and Tibet. It's been shut since 2020. In July, it reopened.
It's not a headline about dollar figures. It's a headline about what a closed border actually costs. For six years, traders on both sides who used to move wool, herbs, and household goods through that pass had no route. When a government reopens something like this, it's a small, deliberate signal, that a relationship which had gone cold is thawing, even if just a little, even if just in one place on a map. Watch this space. Small reopenings tend to come before bigger ones.
Story four: Indian startups are cutting jobs, and it isn't a funding story
Flipkart cut around 500 roles. Zupee cut about 200. Across Indian startups, roughly 1,700 jobs disappeared in the first quarter alone. In past cycles, that pattern meant one thing: funding dried up, companies panicked, and they cut costs to survive.
This time is different, and that's what makes it worth a story instead of a stat. These companies aren't broke. They're still investing, still growing. They're cutting because they now believe a smaller team, supported by AI tools, can produce the same output, or more. That's not a downturn story. That's a structural one. For the first time, companies are shrinking their headcount on purpose, in good times, because the math of what one employee produces has actually changed. That's a much bigger deal than a bad quarter.
Story five: fifty billion dollars of IPOs, on hold because of a war
India was getting ready for a wave of major public listings this year, worth an estimated 50 billion dollars, once tensions in the Middle East seemed to be easing in June. Then, in early July, the US ended the ceasefire with Iran. And that pipeline froze.
For context on how far behind India already is: the US has raised 128 billion dollars this year across 72 listings. Hong Kong has raised 27 billion across 84. India has raised just 4 billion. A listing calendar isn't really a calendar. It's a bet that the macro backdrop will hold steady long enough for a company to go public without its valuation getting caught in a shock. When that backdrop includes an active war that can restart with one geopolitical decision, that bet gets a lot harder to make. Companies don't list into uncertainty. They wait it out. And waiting has a cost too, the cost of capital not raised, of growth not funded, sitting there while the war headlines decide the timeline instead of the company.
One thread running under all of them
Leverage punished conviction. A tax cut raised more tax. A mountain pass measured a thaw in relations. A hiring slowdown revealed a structural shift instead of a downturn. And a war ten thousand kilometers away is quietly deciding when Indian companies get to go public.
None of these are really about the numbers. They're about what the numbers were hiding. That's what we try to do here every month, pull the mechanism out from under the headline.
