CoreWeave sells computing power to AI companies. It owns racks of Nvidia chips, leases them out by the hour, and has used borrowed money to build a lot of them very fast. This week The Verge ran a piece arguing that CoreWeave’s debt is becoming the whole AI industry’s problem, and honestly, the headline undersells it.
Most retail investors will read that story as news about a stock. I read it as a lesson about leverage, and it applies to your portfolio whether you own a single share of CoreWeave or none.
Here’s the thing. Debt itself isn’t the villain. Companies borrow to build factories, and plenty of great businesses carry heavy loans. The question is always the same: does the income arriving each month comfortably cover the payments going out?
CoreWeave’s version of that question is uncomfortable. Its customers are a handful of very large AI labs. Its main asset is hardware that loses value as newer chips ship. And the loans were taken out against a future that may or may not show up on schedule.
That isn’t a new pattern. The company’s financing trouble was visible from the start: it reportedly went looking for another $1.5 billion in debt after its IPO disappointed, having already piled up $12.9 billion of borrowing to build data centers. That sounds like a company with momentum until you notice how much of it was borrowed before the public ever saw a quarterly number.
Two sentences cover it. Revenue is a promise. Debt is a contract.
Most of us are careful about how much of our portfolio goes into one coin or one stock. We’ll say “never more than five percent” and mean it. Then we buy shares in a company that has taken a far bigger bet with money it doesn’t own, and call that diversification.
A simple habit fixes this. Before you buy, find three numbers:
• Total debt against annual operating cash flow. If paying it off would take more than a few years of real cash, that’s a flag.
• Interest cost against operating profit. If interest eats most of the profit, one bad quarter breaks the model.
• Customer concentration. Two or three buyers accounting for most of the revenue means one phone call can change everything.
None of that needs a finance degree. It needs ten minutes and a willingness to be bored. The reason it matters is that asset managers have been warning about exactly this setup. Research from Man Group on the debt-fuelled AI capex cycle and off-balance-sheet financing describes how leverage builds quietly when spending runs ahead of earnings. You don’t have to agree with every conclusion to take the mechanism seriously.
I’d go further. Reading a balance sheet is a better predictor of whether you’ll sleep well than reading a price chart. Charts tell you what happened. Debt schedules tell you what has to happen next.
This is where it gets personal, because retail investors take on leverage constantly without calling it that. Margin loans. Leveraged ETFs. A home-equity line used to fund a trading account. Credit card balances carried while waiting for a position to recover.
The pattern is old. Research from the Bank for International Settlements on the rising influence of retail investors found that surges in retail margin debt have tended to be followed by market declines. Retail leverage tends to peak close to the top. Not always, but often enough to treat as a warning light.
I’ve watched this play out with friends. One of them borrowed against his flat to buy a basket of AI stocks in early 2026. The stocks did fine. His stress did not, because every dip became a question about whether he could still make the mortgage payment. He sold a good position at a bad time, purely to calm himself down. That is what leverage does. It doesn’t just multiply returns, it shortens your patience.
The Atlantic Council makes a related point about how retail margin debt and leveraged ETFs connect to systemic risk when markets turn volatile. Individual choices add up to something larger than any single account.
The fix isn’t to avoid risk. It’s to build walls between pots of money, so a loss in one can’t spill into another.
Living money covers rent, food, tax, insurance and the emergency buffer. It never goes anywhere volatile. Not a little, not just this once.
Long-term money sits in boring, diversified holdings and gets left alone. This is where the HODL logic from the power of long-term investing belongs.
Speculative money is the small slice you can afford to see drop to zero without changing how you live. Early-stage crypto, a single high-variance stock, a concentrated AI bet. Size it so that losing all of it is an annoyance, not an event.
Some readers also keep a small discretionary entertainment budget, and the same walls apply. If you’re in Europe and you set money aside for online casinos, treat it as spending, not investing, and check which operators are properly vetted before you deposit anything. For EU listings, one place to start looking is https://dotesports.com/eu/casinos, but do your own checks on any operator before you deposit. Gambling involves risk, so only use money you’ve already decided you can lose.
Then back to the portfolio, because that’s the point of all this. The discipline is identical in both places: decide the amount in advance, fund it from money that isn’t doing anything else, and never top it up from the living pot to chase a loss.
This is the routine I use now, and it takes about the time of a coffee.
1. Whose money is it? If any part is borrowed, cut the position size roughly in half.
2. What happens if this goes to zero? If the honest answer involves missed bills, the position is too large.
3. Who owes whom? Check the company’s debt against its cash flow, not its share price.
4. What would make me sell? Write the exit rule down before buying. Rules written during a drawdown are written by panic.
None of these checks is clever. That’s the point. Overconfidence loves clever, and it hates a checklist.
I’m not predicting CoreWeave fails. It might grow into every dollar it owes, and the buyers of its compute might keep paying. Plenty of heavily indebted companies have worked out fine.
But notice what the story does to the sector. When one large borrower wobbles, lenders reprice risk for everyone connected to it, and suddenly a debt problem at one firm becomes a funding problem across an industry. Leverage doesn’t stay inside the company that took it on.
The same is true for you. Your leverage doesn’t stay inside your trading account either. It follows you to the dinner table, the mortgage statement, the decisions you make when you’re tired. Keep it small, keep it visible, and keep it away from money you need.
How can a retail investor check how much debt a company has? Open the latest quarterly filing and find the balance sheet. Compare total debt to annual operating cash flow, then check interest expense against operating profit. Free tools like company investor pages and regulator filing databases show all of it. It takes ten minutes.
Is using margin or leveraged ETFs ever a good idea? For most retail investors, rarely. Leverage multiplies losses as quickly as gains and shortens your patience, which causes forced selling at bad moments. If you use it, cap it at a small slice of speculative money and never borrow against assets you need to live on.
What does it mean to separate speculative money from living money? Living money covers bills, tax and your emergency buffer, and stays in safe holdings. Speculative money is a small amount you can lose completely without changing your life. Keeping them in separate accounts stops a bad bet from becoming a household problem.
Why does one company’s debt affect a whole sector? Lenders and investors reprice risk across connected firms when a major borrower looks stretched. Funding gets costlier for competitors and suppliers, even healthy ones. That chain reaction is why heavily financed sectors tend to move together on bad news.
How much of my portfolio should go into a single high-risk position? There’s no universal number, but many investors cap one speculative holding at around five percent of the total. The better test is practical: if losing it all would force you to skip bills or sell something else in a hurry, it’s too large.
Photo: iStock
