Cathie Wood's ARK funds are some of the most prominent voices on Wall Street when it comes to investing in cutting-edge companies. Whether it's biotech or artificial intelligence (AI), the ARK funds have made their stances known. One of their favorite stocks is Tesla (TSLA 0.25%), which they believe could be worth $2,000 per share by 2027 -- indicating upside of roughly 650% from today's prices. However, multiple ARK funds have been doing something odd with Tesla stock lately: selling.

It seems odd that anyone would sell a stock with a price target indicating 51% compounded annual growth over the next five years, as that would be a stellar investment. But that's what ARK has been doing.

So should you follow Cathie Wood's lead and sell a leading AI stock like Tesla? Or is this a buying opportunity? Let's find out.

Investors must look beyond the headlines

Probably one of the most alarming notes about this transaction is the breadth of it. ARK hasn't bought any Tesla shares since April, and has sold shares from its ARK Innovation ETF (ARKK 0.45%), ARK Autonomous Technology & Robotics ETF (ARKQ 0.39%), and ARK Next Generation Internet ETF (ARKW 0.40%). Across all of the transactions, ARK sold over 677,000 shares, or about $173 million worth.

However, Tesla has had a strong run since ARK quit buying in April, as the stock is up 65% since May 1. As a result, the sale of Tesla stock shouldn't be examined through the lens of "ARK has lost its faith in Tesla." Instead, investors should consider it a move to take some gains off the table.

For the three funds listed above that own Tesla stock, Tesla is still a large holding.

Fund Weight Rank Portfolio Weighting
ARK Innovation ETF 1st 10.47%
ARK Autonomous Technology & Robotics ETF 1st 13.56%
ARK Next Generation Internet ETF 5th 6.15%

Data source: ARK.

So even though Cathie Wood made headlines by selling Tesla stock, it was an effort to rebalance the portfolio and redeploy the gains elsewhere.

But this action should get investors thinking: "Should I be doing the same thing?"

Is it time to sell some AI stocks?

A few things come to mind when assessing whether you should sell a stock (especially one wrapped up in a trendy field like AI).

First is the sleep test. Has your position gotten so large that you can't sleep at night because you're worried about what the market might do the next day? If that's the case, your position has gotten too large, and it's time to sell some to take risk off the table. There's no hard and fast number here, as each person's risk tolerance differs. It's around 15% to 20% for me, and this threshold has caused me to sell some stock lately.

Second is expectations. There's no way to tell how much of an impact AI will have on businesses three years out. This also weaves into stock valuations, as a company may have a massive impact but not live up to the hype for years. Take the dot-com crash in the early 2000s. Everyone knew the internet was a game changer and would shape the world. As a result, almost every stock associated with the internet was bid up. While the internet changed how everyone lived, almost every stock from that era crashed and never recovered, despite how prevalent the internet became.

As a result, I look at how a company is valued to see if the expectations are too high. If they are too lofty, I might trim my position if it's had a nice gain, but I won't sell all of it.

This is likely how many Tesla shareholders feel -- it is a highly valued stock, but investors still believe in the business.

I think Cathie Wood and the ARK funds are wise to take some profits from Tesla but maintain their high concentration in the stock. This is a valuable lesson I learned personally from the tech crash of 2022, and I'm putting those lessons learned into action in 2023 by trimming some positions that have likely risen too far, too fast. But if the valuation of a stock like Tesla drops, Cathie Wood and her team will undoubtedly be ready to add more shares, just like myself.