Editor’s Note: IBM’s 25% single-day drop last month was a “show me” quarter — the market demanded proof the AI buildout was paying off, and didn’t get it. While AI data center stocks test investors’ nerves, Jim Pearce has been quietly building a portfolio on the other side of the trade: traditional businesses using AI to widen margins, not build it. See the AI Margin Rotation briefing →

As my colleague Robert Rapier explained last week, “The AI Boom is Entering Its ‘Show Me’ Phase.” Robert makes an astute observation: “The market is becoming much more selective about which companies deserve credit for [Artificial intelligence (AI) data center] demand and which ones are merely spending enormous sums in hopes of capturing it.”

Understanding that distinction may be the key to protecting your investment portfolio from taking a big hit later this year. With equity valuations high and the global economy laboring under the weight of high energy prices, it wouldn’t take much in the way of bad news to send some tech stocks reeling.

Feeling Blue

Case in point is International Business Machines (NYSE: IBM), which saw its share price fall 25 percent in a single day when it disappointed Wall Street with its Q2 results last month. Until then, IBM was on a roll due to its hybrid cloud AI data center hardware capabilities.

That’s why I advocate using trailing stop loss orders. A trailing stop loss order is set at a price beneath the current share price of a stock. When you enter the order, the difference between the current share price of the stock and your stop loss price is converted to a percentage.

If the stock continues going up, your broker will automatically raise your stop loss price to maintain that same percentage corridor. But if the stock starts falling, the trailing stop loss price stays where it is.

For example, I recommended a stop loss price for IBM of $288 to my readers on June 2. That day, IBM closed near $329. The difference between the prices that day was about 12 percent.

Coincidentally, that is the same day that IBM’s share price peaked. Only three days later it traded below my stop loss price of $288. For readers who took my advice, they would have been closed out of that position at that price since the trailing stop price cannot go down.

IBM continued to fall, bottoming below $200 on July 23 before turning around. It got back above $225 last week, but is still a long way from my stop loss price of $288.

A Different Kind of CAT

The argument against using stop-loss orders is that the share price may dip below your stop-loss price just long enough to trigger your sell order before bouncing back. That does happen sometimes and is the price you pay for eliminating the possibility of taking a big loss.

A recent example of that is heavy equipment manufacturer Caterpillar (NYSE: CAT). Caterpillar is up more than 50 percent this year, in large part due to its on-site data center power generation business.

On June 2, I suggested a trailing stop price of $770 while it was trading around $908. That works out to a 15 percent discount to that price. Four weeks later, CAT peaked near $1,065, which raised our trailing stop price to $905.

Less than three weeks later, CAT fell below $905 on its way down to $776 on July 29. However, it opened at $922 on August 4 after releasing strong quarterly results that morning. But by the end of that day, it had dropped back down to $880.

That type of volatility is emblematic of a stock market that is fully valued. Price movements of that magnitude should take several months or longer to occur, not days or weeks. That’s why I am willing to secure my sizable profit in CAT and reinvest those proceeds in the next wave of growth stocks.

Imperfect Timing

We’ve booked some huge gains in tech stocks this year, some of which are equivalent to several years’ worth of stock market appreciation under normal circumstances. That’s why I don’t mind getting stopped out of stocks like IBM and CAT after they have skyrocketed in value.

Valuations for most AI data center stocks are based on aggressive assumptions that can change quickly. By the time you realize that a stock is in freefall, it may be too late to do anything about it. If you don’t have a disciplined process for taking profits, you might end up riding a stock all the way down.

For all I know, IBM and CAT may come roaring back in a few weeks to hit new highs. On the other hand, they may not fully recover for several more months or years.

Either way, I am satisfied with my gains in those positions. Being a successful stock market investor isn’t about being perfect; it’s about knowing when it’s time to take your profits and move on to the next opportunity.

What I’ve been doing in Personal Finance this year runs in parallel to the discipline I describe above: taking profits from overvalued AI builders before the reckoning arrives — I closed NVIDIA in January 2025 at a 975% gain, the same week the math stopped working — and redeploying into traditional businesses that don’t face a “prove it” quarter. These are companies using AI to widen their own margins, steadily, without needing to justify tens of billions in capex. The volatility IBM and Caterpillar showed this year is exactly why I made that rotation. See the AI Margin Rotation briefing →