Tech Stocks Are Crashing. Don’t Make the Mistake of Selling All of Them
Technology stocks are getting hit hard, and for a lot of individual investors the instinct will be to do something immediately. That usually means selling whatever hurts most. I understand it. When a stock you own is down 15%, 20%, or more in a short period, the screen starts to make the decision for you. The problem is that broad selloffs tend to mix businesses that have very different problems. Some stocks are falling because the valuation got ridiculous. Some are falling because investors are finally questioning the returns on enormous AI spending. Some are being sold simply because they sit in the same funds and portfolios as everything else. Those are not the same situations, and treating them as if they are where investors can do real damage to themselves.
The easiest thing to say right now is that the AI trade is breaking. I am not convinced that is the useful conclusion. AI demand can remain very strong while many AI-related stocks still fall. In fact, that may be exactly what is happening. The market spent a long time rewarding companies for announcing more spending, more data centers, more chips, and more capacity. At some point investors were always going to ask what return they were getting on all that capital. That question matters far more to me than whether the next quarterly number shows another big increase in AI demand.
More News from Barchart
The Market Is Finally Looking at the Bill
Meta is a good example of what has changed. The business did not suddenly become bad because investors became uncomfortable with the amount of money being committed to AI. Meta still produces enormous cash flow and has the balance sheet to fund much of that investment itself. The debate is about the return on the spending, not whether the company can find the money. That distinction is important because there are plenty of other technology companies where the financing question is much harder. If a company needs to keep borrowing or issuing equity to fund the same AI opportunity, shareholders are taking a very different risk.
This is where I think the market got lazy during the boom. AI became an explanation for almost any level of capital spending. Management teams could announce another few billion dollars of investment, and investors would often treat that as proof of future growth. But spending money is easy. Earning a return on it is harder. I have watched enough capital cycles over the years to know that the point of maximum enthusiasm is often when nobody wants to ask who will actually make money at the end of it.
A company selling scarce equipment into the AI buildout may get paid long before its customer knows whether the project will earn an acceptable return. A cloud provider with huge existing cash flows can fund mistakes that would cripple a smaller competitor. A data center operator carrying heavy debt may have extraordinary demand and still deliver disappointing returns to shareholders because too much of the economics are absorbed by interest expense, capital spending, and future financing needs. You can believe completely in the growth of AI and still decide that some of the stocks attached to it are poor investments.
That is why I would be very careful about using the word “tech” as though it describes one business. It does not. Nvidia and a leveraged infrastructure company may both benefit from AI, but they do not have the same economics. Microsoft and an early-stage software company do not have the same financing risk. A business earning enormous free cash flow today should not automatically be sold because another company in the same theme needs fresh capital to keep growing.
A Falling Stock Is Not the Same as a Broken Business
The next mistake is assuming that because something is down heavily, it has become cheap. I have never found percentage declines particularly useful on their own. Even a stock that’s down 30% from an absurd value can still be costly. And a good business can fall 20% because investors have decided they don’t want to pay the old multiple anymore, even if operationally very little has changed.
This is where investors need to go back to the company rather than the chart. I would want to know how much cash the business generates, how much it plans to spend, whether it can finance that spending internally, and what the balance sheet looks like if the expected returns take longer to arrive. I would look at margins and free cash flow before I worried about whether the stock was down another 5% that day. I would also look closely at what management is doing with its own capital. Is it still buying back stock while asking shareholders to be patient? Is it issuing shares? Is debt climbing? Is capital expenditure growing much faster than revenue? Those details tell you far more about risk than the red number on your screen.
There is another thing I watch during periods like this: how stocks react when the news is good. If a company reports a strong quarter, raises guidance, and still cannot rally, expectations may simply have been too high. That can continue for a while. On the other hand, if a stock stops making new lows even when the headlines remain ugly, that gets my attention. Markets usually stop falling before investors feel comfortable buying them.
This is also why I would not rush. Individual investors have an advantage they often throw away. Nobody is forcing them to deploy capital this week. They can watch another quarter, see another earnings report, and wait for management to provide evidence that the spending is producing something. Institutions sometimes have to remain invested because that is their mandate. You do not. Cash is not an admission that you have run out of ideas. Sometimes it is simply the price of waiting for a better one.
What I Would Separate Now
If I were going through a technology portfolio today, I would divide the holdings by financial strength before I divide them by sector. I would be more comfortable with companies that can fund their own growth, have established customers, produce meaningful free cash flow, and do not need the capital markets to remain generous. I would be much harder on companies where the investment thesis requires several years of heavy spending before the economics become visible, especially where debt or dilution is doing the financing.
That does not mean the strongest balance sheet automatically gives you the best stock. Valuation still matters. Some outstanding businesses became priced for almost flawless execution, and there is nothing wrong with allowing those valuations to come back to earth. I would rather miss the first 10% of a recovery than buy something simply because it is 25% below the high and then discover that the original price was never sensible.
The opportunity in this selloff will come from separating those two things. Some businesses are being repriced because the market has become less willing to pay for distant growth. Others are being repriced because investors are discovering that the future they were promised may require far more capital and produce much lower returns than they expected. One can create opportunity. The other can create a value trap.
After more than 35 years in markets, I have learned not to become too excited when everything is rising and not to become too frightened when everything is falling. The useful work usually starts once the market stops treating every company the same. That is where we may be heading now.
I wouldn’t sell technology just because technology is being sold. I’d go from company to company and pose a far less emotive question: is the share price declining because the business is getting worse or because investors are finally placing a fairer price on the business that is still there?
Those two situations can look exactly the same on a bad day. They usually look very different a year later.
On the date of publication, Jim Osman did not have (either directly or indirectly) positions in any of the securities mentioned in this article. All information and data in this article is solely for informational purposes. This article was originally published on Barchart.com