Nvidia just put another $50 billion on the table. The board added that to the existing buyback authorization, bringing the total to a fresh $150 billion, right alongside another monster earnings report. MSN has the full breakdown, but the short version is: the AI boom is still paying Nvidia's bills, and Nvidia is using those bills to pay shareholders.

Revenue and profits keep climbing because data center demand refuses to cool off. It's the same story with a bigger number attached every quarter, and this quarter's number came with a buyback strapped to it.

So what does a $150 billion buyback authorization actually tell you? A few things, and none of them are subtle.

First, management thinks the stock is worth buying at these levels. Companies don't announce one of the largest buyback programs around because they think shares are expensive. Warren would tell you buybacks are just management admitting they can't find better uses for the cash, and on a normal day I'd argue with him. But $150 billion is a billboard. It says the people who know the order book best aren't flinching.

Second, it says the cash flow is real. Buybacks get authorized off actual money coming in the door, and Nvidia's money problem right now is figuring out where to put it. Which is a very good problem to have. Not every AI story can say that, and I've flagged plenty that can't, from the dot-com-style valuation signals I keep watching to the companies burning billions with nothing to show for it.

Third, and this is the part nobody says out loud: buybacks shrink the share count, which pumps earnings per share, which makes the stock look cheaper on paper without the business doing anything different. It's shareholder-friendly, but it's also arithmetic. Don't confuse the two effects.

Now the honest part, because that's the deal here. I don't own Nvidia directly. My exposure is through index funds, which means I own plenty of it whether I like it or not, and it's one of the biggest weights in the S&P 500. If Nvidia stumbles, your "diversified" index fund feels it, and that's a risk most people are carrying without ever thinking about it.

And Nvidia is not a sure thing. Customer concentration is wild, the hyperscalers building their own chips isn't a rumor anymore, and the valuation already prices in years of flawless execution. The buyback softens the math but it doesn't erase any of that.

So what would I do? The same boring thing I always do. Keep the automated index fund buys running, let Nvidia be the biggest slice of a very large pie, and resist the urge to make it a bigger slice because a $150 billion headline feels exciting. Individual stocks, even great ones, are toppings. I learned that the hard way in my twenties with a position I'd rather not describe in detail.

If you already own Nvidia, the buyback is a reason to relax, not a reason to double down. If you don't, this doesn't change the calculation. The AI trade has been very good to a lot of people, and it will keep being good right up until the quarter it isn't.

Warren thinks I'm too calm about the concentration. I think he's too calm about missing it. We're both right, again.