Ever since it filed its paperwork to go public, I was interested in data warehousing company Snowflake (NYSE:SNOW). I even thought about participating in the IPO.

28 million shares were up for grabs, so I figured there'd be plenty for the taking. And Snowflake was planning to sell its shares in a range of $75 to $85. Just 48 times trailing sales -- what a bargain!

But that's when everything started to go wrong.

Words Common Mistakes written on a notepad

Image source: Getty Images.

"And I would have gotten away with it too, if it weren't for you meddling [Warren Buffett]"

First, the news broke that Warren Buffett was getting involved in Snowflake's IPO, sinking as much as $573 million worth of Berkshire Hathaway's (NYSE:BRK-A) (NYSE:BRK-B) cash into the IPO before it began trading.

News that the Oracle of Omaha was backing Snowflake's play (confirmed by data from S&P Global Market Intelligence) had an immediate effect. In no time at all, an IPO that was expected to cost about $80 jumped to $120 a share.

Snowflake had become 50% more expensive -- and it hadn't even begun trading yet. 

From bad to worse ...

And that was the good news. As it turned out, Buffett's interest did more than just inflate Snowflake's IPO price. It also attracted the attention of thousands (if not millions) of investors looking to buy the stock once it began trading.

By the time IPO Day rolled around, bids were coming in fast and furious in pre-market trading. And when trading opened, by the time outside investors were finally able to buy the stock, Snowflake cost ... $245!

That's right. Three times the advertised price.

Even so, investors were still interested in owning the stock. Snowflake started its first trading day at a valuation of nearly 150 times trailing sales -- let alone earnings -- and yet the skyrocketing price seemed only to excite the momentum traders. Over the course of this first day of trading, Snowflake stock rose as high as $319, before turning tail and heading back down, eventually closing the day just under $254 a share.

... to worst

And so it turned out that, despite all the hype, from first trade to last, shares of Snowflake managed to rise less than 4%. True, investors like Buffett who managed to "get in on the IPO" before trading began made out like bandits. But everyone else, I suspect, ended up rather underwhelmed -- and the trading news remained grim for the next two days.

On Thursday, Snowflake stock fell 10%. And while it regained some of those losses Friday, the stock still closed the week at just $240 a share -- $5 below the best price most investors were able to get it for on IPO Day.

What investors can learn from the Snowflake IPO

So what is the takeaway for investors?

At a minimum, I see three key lessons for investors to keep in mind as they contemplate upcoming IPOs in popular stocks such as Robinhood, Palantir, and Airbnb later this year:

  1. IPOs are risky, and the advertised price may not be the price you ultimately pay.

Part of the reason Snowflake's IPO was so popular was because of the relatively low $80 share price at which its IPO was first announced. Had Snowflake actually IPOed at that price, the entire company would have been valued at just $22 billion and change, and could have been an arguable bargain relative to pricier cloud computing stocks such as Okta ($25 billion) or Twilio ($32 billion) or Atlassian ($42 billion).

As it turned out, by the time most investors got a chance to invest in this IPO, Snowflake cost three times as much, and was clearly overvalued.

  1. The more you pay, the less you get.

Everyone knows the old investing saw: "Buy low, sell high." But as Snowflake stock got more and more expensive before trading even began, the chance to "buy low" slipped away. By the time trading began at $245, this train hadn't just left the station -- it was already in the next state.

After paying such a high price to "get in" on the IPO, there weren't a lot of gains left to be had. That's why, by the time trading ended on Wednesday, most investors were sitting on a profit of less than 4%.

  1. An overpriced IPO needs a plentiful supply of "greater fools" to resell shares to. When that supply runs dry, look out below.

Despite the fact that Snowflake stock began trading at ridiculously expensive levels, the stock still went up -- at first. But it wasn't rising because Snowflake was becoming more valuable. (The stock was less than one day old! It hadn't had any time to get more valuable!) Instead, Snowflake's stock price ran up because traders, who had no intention of owning Snowflake for the long term, tried to make a quick buck selling the stock back and forth to each other.

Once that game ended, though, the game was up for Snowflake, too -- and the stock fell.

This, in a nutshell, is why most Snowflake investors ended last week with nothing but losses.