Artificial intelligence (AI) is certainly having its moment. And if we are fully honest with ourselves, this isn't the first time; in the past decade, we have seen other instances of AI becoming a popular buzzword in investing circles. What does seem different this time is the real-world impact, as AI tools are starting to show up in many of the software tools used by millions of people every day, and in generative AI like OpenAI's ChatGPT and Google's Bard, where nearly anyone can interact with and potentially benefit from their capabilities. 

We've also seen dozens of stocks touting AI as a core part of their future rocket higher. And while more than a few of those have spent a lot more pitching AI to investors as their future than actually building cash-generating AI tools, Nvidia (NVDA 2.15%) is nearly certain to be one of the leaders -- and leading profit-makers -- in the AI future. Despite this, and the likelihood that its revenue and profits are going to continue climbing for many years to come, I made the hard decision to sell the majority of my Nvidia stock this week.

It had become such a significant concern for me that I even made the decision to sell before the company reported earnings on May 24. Depending on their situation, this very well might be the right decision for a lot of other people, too. 

Pulling the flowers?

Both Motley Fool co-founder David Gardner and investing legend Peter Lynch have emphasized the importance of letting your winners run and buying more of them as part of a market-beating investment strategy. Based on both of their track records, focusing on winning investors, and "averaging up" -- buying more of a stock that's gained in value -- is a better approach than "averaging down" and adding to your losers. Both investors have described it as akin to "watering the weeds and pulling the flowers." 

On the surface, selling Nvidia is "pulling the flowers." My oldest shares were purchased in June 2017, and the stock has gained a remarkable 735% since then. That's good for almost 43% annualized returns! Considering my aforementioned expectations that Nvidia, one of the most important semiconductor companies on earth, will keep growing its revenue and profits for many years to come, what's my thinking here? 

In short, Nvidia had become a pretty large portion of my net wealth. Sometimes we have to sell off those winners to start preserving future spending power. 

Growth at any price = big downside risk

This is an excellent example of when it's OK to love a company but not fall in love with a stock. In the case of Nvidia's stock, today's price has shot far beyond shouting distance of a fair price. AI stocks have become caught up in a bit of a mania, and that's pushed Nvidia's stock price close to levels we last saw in late 2022, when most tech and growth stock valuations had reached extremes that we are still far below 18 months later. 

Nvidia shares now trade for almost 140 times operating cash flow and 206 times free cash flow. For context, Nvidia's stock has traded for an average of 39 and 47 times operating cash and free cash flow over the past 10 years: 

NVDA Price to CFO Per Share (TTM) Chart

NVDA Price to CFO Per Share (TTM) data by YCharts

That means today's price is more than three times the average operating cash flow multiple, and more than four times the average free-cash-flow multiple over the past decade. That's a hard premium to justify, and if things don't go perfect, investors will almost certainly sell faster than you can say "AI took my job." 

Why not just hold? Why sell?

Excellent questions -- probably the most important ones to ask yourself, because I realize that selling now probably smacks of market timing, a very un-Foolish way to invest. But let's not ignore the "Motley" part of The Fool's name. In my case, it's important to look at the downside risk if Nvidia's future results don't earn today's price premium, and the implications for me and my family. As a result of the recent run, Nvidia had become a large enough portion of my total wealth that it was time to sell. 

Could the stock keep going up? Sure, I wouldn't be surprised if it did. After all, AI is still all the rage and shares are still below Nvidia's high from 2021. I'm not trying to catch a top here; but I am looking to preserve some of the wealth that Nvidia has helped me build. And since my core holdings are in a retirement account, there are no tax consequences to further complicate this decision. 

My answer to these questions: Why not both?

While I have sold most of my Nvidia shares, I've held a small position. That's because I do expect shares will eventually be worth far more than today's price; whether it's a month, a year, or a decade from now is far less certain because the company has a lot of growing to do to "earn" this price at a more reasonable valuation. But I still have a couple of decades before I reach retirement age, so I can hold a little bit for a long time, and can buy more shares if Mr. Market gives me a better price. 

Making the decision that's right for you

As Warren Buffett famously said, "our favorite holding period is forever." Too much attention is paid to "forever" and not enough to "favorite." As much as it would just be easier to hold all those Nvidia shares and hope for the best outcome, hope isn't a good investing strategy. For me and my family, taking some returns I wasn't expecting to earn so quickly just makes sense; it's been too big of a winner, and the stakes are too large at this point for me to not reduce my exposure.

With better opportunities to put those funds to work building wealth, it's the right move for me.