In his own words, it’s been “an intense 7 days” for Elon Musk. Last Friday, Tesla demonstrated a new prototype product, a robot called Optimus. On Sunday, Tesla announced that it had delivered a record number of new vehicles in the third quarter. And on Wednesday, a rocket built by SpaceX, one of Musk’s other companies, completed a successful launch from Cape Canaveral, carrying astronauts to the International Space Station.

At the same time, a cloud of controversy always seems to trail Musk. For most of this year, he has been in litigation with Twitter, which he offered to buy in April before reversing course. And this week, Musk proposed a peace plan for Ukraine that unleashed vitriol from around the world.

Not surprisingly, Tesla, as well as its stock, have been similarly unpredictable. Almost since its inception, Tesla has been a sort of Rorschach test for investors—myself included. One conversation, back in 2014, stands out in my mind. After Tesla’s shares had jumped 344% the year before, I became skeptical. The shares seemed grossly overpriced. A friend, meanwhile, felt that the company was onto something and that its share price made sense.

Since that time, Tesla’s shares have gained another 2,000%, so it’s fair to say my assessment was incorrect. In recounting this story, I’m happy to eat crow because I believe it conveys key lessons—not just about Tesla but about the stock market in general. These are the five ways I misjudged Tesla:

Growth rate. The way most investment analysts value a stock is to extrapolate the company’s current revenue and profit figures into the future. From that, a projected future stock price can be calculated. Suppose a company is earning $5 per share today, and I think it’ll earn $7 in three years. And let’s assume the company’s current price/earnings ratio of 15 will remain constant. That would allow me to project a future stock price of $105 ($7 x 15).

For a mature company, that’s a generally reasonable approach. But for companies that aren’t yet profitable, it’s much harder because you have to guess when they’ll turn the corner. That was the situation in 2014. Tesla was growing rapidly, but it was losing money. It wasn’t until 2020, in fact, that it produced its first full-year profit.

Investment analysts trying to value unprofitable companies have two less-than-ideal options. The first is to extrapolate from current numbers, as described above. In 2014, for example, Tesla delivered 32,000 cars, up from about 22,000 the year before. An analyst might project future growth at a similar pace. The second approach would be the “blue sky” method, whereby an investor would simply use his imagination to paint a picture of the future.

Each approach carries a potential flaw. With the extrapolation method, the risk is that an estimate may end up being too low. That was the mistake I made. But with the second approach—the blue sky method—the risk is that an estimate could end up being too high. Neither is a great result. And we’re really only guessing when we choose which way to go.

Valuation. Not only did Tesla grow faster than expected, but the stock market also put a higher value on those profits too. I expected that when Tesla turned profitable, its price/earnings ratio would be more in line with other fast-growing companies. But sometimes investors are willing to give a company a pass, allowing its P/E ratio to remain in the stratosphere for years. Even today, after a 30% year-to-date decline, Tesla’s shares trade at 47 times projected earnings. That’s about three times higher than the overall market average and double that of a more typical fast-growing company.

Plans. Suppose you or I had been able to sit down with Elon Musk in 2014, and suppose he had shared with us all of the plans he had for the company. Would that have helped in making a more accurate projection for Tesla’s shares? I’m not sure it would have. That’s because Tesla has a track record of promoting products years before they are released. For that reason, even if we’d received information directly from the CEO, it’s unlikely that would have led to a more accurate forecast of Tesla’s growth.

Competition. For years, Tesla skeptics have argued that competition would inevitably slow its growth. That was part of my concern back in 2014. After all, an electric car is simpler to manufacture than a gasoline-powered one, so the world’s major auto makers faced no real technical hurdle in challenging Tesla.

In fact, an oft-cited rule of thumb in business is that “fast followers” tend to fare better than first movers in new industries. Google wasn’t the first to create a search engine, Apple wasn’t the first to make cell phones, and Facebook wasn’t the first to create a social network. Toyota, it was widely assumed, could put Tesla in its place if it chose.

What no one counted on, though, was that Toyota would go in the opposite direction. As recently as last week, Toyota CEO Akio Toyoda reiterated his lukewarm view of electric vehicles. “EVs are just going to take longer to become mainstream than [the] media would like us to believe,” Toyoda said. He cited a lack of infrastructure, among other concerns. Whether Toyoda is right or wrong on this, the fact is that he’s allowed Tesla more time with less competition to gain customers and build an infrastructure of its own.

Unforeseen factors. In assessing a company’s prospects, the factors discussed above are usually the most significant. But they’re not the only ones. What else could have gone wrong for Tesla? Consider just a few potholes the company has avoided:

• As noted above, Musk has been unpredictable, both professionally and personally. That could have led to his departure from Tesla—either for political reasons or of his own volition. At age 51, with 10 children and as the world’s wealthiest person, that wouldn’t have been unheard of.
• The company could have made a strategic error. Today, Tesla’s original high-end models—the Model S and Model X, which each cost more than $100,000—account for less than 5% of the company’s sales. If the company hadn’t developed new products, this suggests how much slower growth might have been.
• Numerous other factors, too, could have slowed Tesla’s growth. Gas prices might have been lower, dampening interest in electric cars. Lithium might have been in tighter supply than it already is. Another company might have had a breakthrough with battery technology. A bug in Tesla’s self-driving technology might have caused a catastrophic lawsuit. The list is nearly endless. In hindsight, Tesla’s success looks inevitable. But onetime competitors like Fisker—now bankrupt—prove that success wasn’t guaranteed.

What can you learn from this story? As I’ve noted many times before, stock-picking is immensely challenging. That’s why, at the risk of sounding like a broken record, I always recommend index funds. But it’s also instructive, I think, to see exactly why this is the case. While Tesla may be unique, it’s also representative of the challenges stock-pickers always face. Look at any company, and you’ll observe strengths and weaknesses. Unfortunately for investors, though, what we can observe represents just a fraction of what can potentially, over time, impact a company. That, in a nutshell, is why stock-pickers have such a challenging time.

This story also helps explain another piece of research I’ve mentioned in the past: Hendrik Bessembinder has found that just 4% of all stocks account for all of the stock market’s historical outperformance relative to bonds. This strikes many people as an almost unbelievable statistic. But Tesla’s performance—a gain of about 11,000% over the past ten years—helps explain Bessembinder’s finding. That too is another reason to own index funds. Even if you dismiss a high-flier the way I did, at least you’ll benefit from owning some stake in it.