Oracle Stock Is 54% Below Its High. The Last Time It Fell That Far, Getting Back Took More Than a Decade.

Oracle (NYSE:ORCL) closed at $158.78 on Friday, 54% below its record high of $345.72. The stock set that high on Sept. 10 of last year -- one year to the day before its next earnings report, scheduled for Thursday, Sept. 10.

Shares have been in a hole this deep before. Measured on monthly closing prices, adjusted for stock splits, the stock peaked at $45.47 in August 2000 and fell 83% to $7.86 by September 2002. It didn't close a month above the 2000 level until June 2017, nearly 17 years after the peak.

Before that history scares anyone out of the stock, though, it's worth looking at what made the last recovery so slow. It wasn't the business.

Image source: Getty Images.

The 17-year round trip

Daily closing prices tell the same story. The stock's best close of 2000 was $46.31, on Sept. 1 of that year. Shares came within 8 cents of that level in December 2014, faded, and finally closed higher on June 21, 2017. Later that day, Oracle reported its fiscal 2017 results.

And the fall itself was enormous. A stock that drops 83% has to rise nearly 480% just to get back to even.

Today's decline has been shallower, but not by as much as the 54% figure suggests. At its low of $114.50 in late July, the stock was down 67% from its high.

What took so long?

To me, the striking part of the dot-com episode is what didn't happen. Demand didn't collapse.

Oracle's revenue rose 7% in fiscal 2001 (the company's fiscal years end May 31), then fell about 12% in fiscal 2002 as customers cut back on technology spending. That was the worst single year. In fact, operating income was higher in fiscal 2002, at about $3.6 billion, than it had been in fiscal 2000.

The problem was the starting price. When the stock peaked in August 2000, Oracle was on its way to earning a split-adjusted $0.44 per share in fiscal 2001, the year then underway. The peak price was more than 100 times those earnings.

However, even a growing business can spend years catching up to a price like that.

By fiscal 2017, revenue had more than tripled from fiscal 2000's $10.2 billion to $37.7 billion, and earnings per share had grown about five times, to $2.21. By the time the stock crossed its old peak in June 2017, it traded at about 21 times earnings.

In other words, the stock didn't so much recover as grow into its old price.

This time, the spending is Oracle's own

The two declines do share one thing: heavy spending sits at the center of both.

But in 2000, the spending at risk belonged to Oracle's customers, and when they pulled back, revenue dipped. Today the heavy spending is Oracle's own -- and demand is the strong part of the story.

Highlighting that demand, Oracle's fiscal 2026 revenue climbed 17% year over year, to $67.4 billion. Cloud infrastructure revenue did even better, rising 77% to $18.1 billion. Remaining performance obligations (contracted future revenue that hasn't shown up in results yet) ended the year at $638 billion, up from $138 billion a year before. And management expects more, confirming a fiscal 2027 revenue target of about $90 billion, or growth of about 34%.

Paying for that demand is the hard part. Oracle's operations produced a record $32 billion of cash in fiscal 2026, 54% more than the year before. But capital expenditures jumped 162%, to $55.7 billion, leaving free cash flow at negative $23.7 billion.

The comparison to 2000 breaks down at the valuation, though, and in shareholders' favor. Shares have a price-to-earnings ratio of about 27. Management's fiscal 2027 guidance calls for $8.05 of non-GAAP (adjusted) earnings per share, so buyers today are paying about 20 times those expected earnings. That could prove expensive if the margins on Oracle's artificial intelligence (AI) contracts disappoint, but it's nothing like the triple-digit price-to-earnings multiple of 2000.

Ultimately, the 17-year round trip is arguably a lesson about starting prices more than a warning about Oracle. Still, I wouldn't buy the stock today. A company that spent about $24 billion more cash than its operations brought in last fiscal year carries real risk if AI demand cools.

But for investors who already own the stock, the dot-com comparison seems like a weak reason to sell. The last long wait began at more than 100 times earnings. Today's starting point is nowhere near that.

Daniel Sparks and his clients do not have positions in any of the stocks mentioned. The Motley Fool has positions in and recommends Oracle. The Motley Fool has a disclosure policy.