This is the case reaction for how Amazon survived the dot-com bust. I'm sensemaking this case in the context of surviving disruptive change, though there are other ways to approach it. You could analyse it as an investor, or as an example of what it looks like when a technology darling has to prove, in the wake of the bursting technology bubble, that it is a legitimate business. Amazon certainly does that here. Bezos takes the company to profitability over the four years following the dot-com bust.
There are four points I want to make when sensemaking this case.
The first point is to notice how Amazon raised a convertible debt offering at just the right time — right before the dot-com bubble burst. This is partly luck, as I'll discuss in the second point, because predicting when a bubble will burst is neither easy nor really possible. That said, they were very savvy financial operators to have issued a convertible bond when they did.
A Common Cog member who was sensemaking this case kindly sent me their notes, and I quote: "It is advisable and common to raise convertible bonds where the conversion price is a premium to stock price when your stock price has run beyond intrinsic value. Effectively, it makes it cheap debt, or cheaper debt compared to normal debt without the equity option." This turns out to be true, and well known in finance circles. Let me break down what it means.
Amazon clearly had a competent finance team at the time. The legendary Joy Covey, who was the first CFO, had already left by then — she had burned out — with Warren Jensen taking over during the events of this case. Covey sadly died young, but she was a very well-respected CFO. Along with Bezos, she was responsible for writing the legendary 1997 annual shareholder letter that Bezos subsequently attached to every single annual report, talking about day one. If you read that 1997 shareholder letter, you would realise that Covey and Bezos were deeply respectful of their investors.
There are many stories out there, if you go looking, about investor analysts recounting how Covey would call them up in complete anger after they had written an incredibly bullish, glowing report on Amazon. She would tell them, "Hey, I'm trying to run a business here, and my employees are getting too excited over the value of their stock because they're reading news reports citing your analysis. Can you lay off?" Covey, Bezos, and other members of the leadership team clearly knew that their stock price did not reflect underlying reality.
So why, if you believe your stock is well above intrinsic value, should you issue lots of convertible bonds with the conversion price set at a premium? The short answer is this. If you raise money through straightforward debt — say you issue a bunch of bonds — investors will demand a certain interest rate to compensate them for the risk. But if you issue a convertible bond when your stock is really hot, you can effectively say, "Can you give me a lower interest rate in exchange for the option to participate in the equity upside if my stock gets bid up further?" You are exploiting the market's optimism, and at the time, because of the dot-com mania, everybody was excited about Amazon stock. Amazon's stock performance had also been very strong from 1997 onwards.
You are basically saying to investors, "You want the equity upside? Give me a discount on the interest rate." The reason you want to set your conversion price higher than your current stock price is that you don't want your bondholders to convert immediately. If the conversion price is lower than, or the same as, your current stock price, that is just unnecessary dilution — they convert straight to equity the moment they receive the bonds, and you have diluted your shareholders for no good reason. That is not great capital allocation.
So you want the conversion price set higher. As an executive, ideally you want it as high as possible, while investors want it as low as possible, close to the current stock price. There is a give and take on the interest rate too: investors want it higher, you want it lower. The equilibrium between these two tensions is what ultimately determines whether you get the conversion price and interest rate you are after.
Conceptually, though, this instrument makes a great deal of sense, especially when your stock is hot, because you are securing much cheaper debt by exploiting the market's excitement over your perceived equity value. Amazon closed this offering, and the timing turned out to be fortunate — which brings me to my second point.
The second point is that the dot-com bubble bursting was not predictable. Any serious student of markets will tell you that you cannot predict when a bubble bursts, because there is enormous volatility both on the way up and on the way down from the peak — and the peak itself can only be identified retrospectively.
Consider the experience of an Amazon executive. You IPO in 1997, and almost immediately there is the Asian financial crisis. It is a huge deal in Asia, but thankfully not in Western markets — though the market is still frightened. The S&P 500 falls 6.8% in a single day, with the total decline in response to the crisis reaching around 13%, before the market recovers and zips right back up in the grip of the dot-com mania.
Then, a year later in August 1998, Russia defaults on its domestic debt. That alone might not have been catastrophic, but one of the world's most sophisticated and famous hedge funds, Long-Term Capital Management, blows up at the same time. The Russian sovereign debt crisis devastates LTCM's convergence trades, and because LTCM had so many cross-holdings, the New York Fed and various other financial market players feared the contagion would spread to other firms and banks. The New York Fed ends up running a coordinated bailout. Between August and October 1998, the S&P 500 falls by around 30%, and the Nasdaq falls by even more — around 50%, I believe — with multiple trading halts during this period.
Now imagine you are an Amazon executive living through this. Your stock has been going up. There was a blip in 1997, a frightening few weeks, and then it kept climbing. Then over the course of three months in 1998, the market gets smashed by 30%. There was no way to know in advance that the New York Fed would save LTCM. In fact, Marc Andreessen — the wunderkind running Netscape at the time — said years later that everybody thought the dot-com bubble, which most people sensed they were in, would finally burst during those months in late 1998.
But what actually happens? After October, the market recovers 50% in just three months, and then over the following year roars vertically by 256% to the peak of the dot-com bubble at the end of 1999. So there is simply no way to predict this. Having lived through that experience, who is to say that the decline from the peak in late 1999 into 2000 would not have corrected once again and sent the market ripping up another 200 or 500 percent? This is precisely what makes bubbles so difficult to navigate.
The most famous story from that era is probably Stan Druckenmiller. He managed to sell out of the dot-com bubble, but could not stomach the final vertical rip of the market and bought right back in at the worst possible time, losing a great deal of money — so much that he had to step away from trading entirely for a period. He reportedly put himself on a beach somewhere with no access to any trading screen.
So when Amazon closed that convertible debt offering, let us be clear: the timing was luck. They closed it just a couple of months before the peak of the bubble, but nobody could have predicted when it would burst.
The third point is to notice how different the reality of Amazon's business was from the slaughter happening in the public markets. There was a lot of blocking and tackling in response to Ravi Suria's analysis, to make sure their negative cash conversion cycle did not unwind. But the fact is that after the dot-com bubble burst, Bezos and his team did a round of layoffs, slowed expansion, and then set about demonstrating to Wall Street that they could be profitable. Amazon shifted to profitability over the course of four years, simply to show that they could. In 2001, they posted their first quarterly GAAP net profit. In financial year 2002 — with results released in early 2003 — they showed the first full-year positive GAAP operating income. Then in financial year 2003, announced in January 2004, they delivered the first full-year positive GAAP net profit. After that, of course, Bezos took Amazon back into the red in 2004 with the launch of Amazon Prime and a range of other initiatives covered elsewhere in the case library.
The point is that Amazon had a working business. As the saying goes, they put up or shut up — and they put up, showing the world they were not just another dot-com darling with nothing behind them.
The fourth point is really two related observations. First, Bezos was unwavering in his vision to build an everything store, which makes it striking that during these difficult years he had to generate a huge chunk of revenue by white-labelling e-commerce solutions for Toys "R" Us, Target, and I believe Borders. That must have been painful, but he needed to generate the cash. Second, notice how almost the entire leadership team churned during this period. It was simply so psychologically difficult, stressful, and painful to live through the dot-com bust and then spend four years grinding out proof, milestone by milestone, that the business actually had legs.
So, that is my sensemaking for this case. I would be very interested to hear what else you notice. My focus has been on the fact that Amazon shored up its balance sheet at exactly the right moment, and that it just so happened to have a working business underneath all the hype — which proved decisive once the bubble burst and external funding dried up for years.
This is the case reaction for how Amazon survived the dot-com bust. And I am sensemaking this case in the context of surviving disruptive change. There are other ways to sensemake this case. You can approach this case as an investor, um, as an example of what it looks like when a technology darling, uh, has to prove in the wake of the technology bubble that it's in bursting, that it is a legitimate business. Um, and Amazon certainly does that here. Bezos takes the company to profitability over the subsequent four years after the dot-com bubble burst. I think there are a couple of points. I think there's like four points that I wanted to make, uh, when I'm sensemaking this case with regards to this context. Um, the first one is notice how Amazon raised a convertible debt offering at just the right time, right before the dot-com b- bubble burst. Uh, this is partly luck, as we'll talk about in our second point. It's not easy and even possible to predict that a bubble is gonna burst. Um, uh, but, uh, the-- they were very savvy financial operators to be able to issue a convertible bond. Um, now, a Common Cog member who was sensemaking this case and kindly sent their, uh, sensemaking to me, uh, they said, and I quote, "It is advisable and common to raise convertible bonds where the conversion price is a premium to stock price when your stock price has run beyond intrinsic value. Effectively, it makes it cheap debt or cheaper debt compared to normal debt without the equity option." And this turns out to be true. Uh, it turns out to be well known actually in finance circles, and I'll break down what this means in a bit. Um, but if certainly at that time, Amazon had a competent, uh, finance team. Uh, I think the very legendary Joy Covey, who was the first CFO, had already left by then. She had burned out. Um, but, uh, oh, she, she, she had left during this case because Warren Jensen takes over during the events of this case. Uh, sadly, Covey died young, but Covey was a very well-respected CFO. She, along with Bezos, was responsible for writing the legendary 1997 annual shareholder letter that in subsequent years, Bezos included attached to every single one of his annual reports, um, uh, uh, talking about day one. Uh, and if you read that '97 financial, uh, sh-- uh, that ninety, ninety-seven shareholder letter, you would realize that Covey and Bezos certainly was very respectful of their investors. Covey, uh, herself, there are many stories out there if you go looking for them about various investor analysts, uh, talking about how Covey would call them up in complete anger when they had written an incredibly bullish, glowing report on Amazon and would tell them, "Hey, you know, I'm trying to run a business here, and my employees are getting too excited, uh, over the, the value of their stock because they're reading the news reports that are citing your analysis. Can you lay off?" Um, so very clearly, Joy, uh, Covey, as well as Bezos and other members of the leadership team, knew that their stock price did not reflect underlying reality. So why is it that when you believe that your stock is way above intrinsic value, you should, uh, issue lots of convertible bonds where the converting-- the, the conversion price is at a premium, right? Um, and, and the short answer to that is that, uh, if you are raising money through debt, um, uh, you just say you issue a whole bunch of bonds, uh, investors will demand a certain interest rate, right, to, uh, uh, compensate them for the risk. Um, if you issue a convertible bond and your stock is really hot, right, uh, you can sort of say that, "Hey, can you give me a lower interest rate for the option to take part in the equity upside if my stock gets bid up further?" Um, you're effectively trying to exploit the market's optimism, which at the time, because of the dot-com, uh, bubble or the dot-com, uh, hype, the mania, uh, everybody was excited about Amazon stock, right? Amazon stock performance was also really good from '97 on- onwards. Um, now, uh, you can-- You're basically saying to investors, "Hey, you want the equity upside, give me a discount on interest rate." And the reason why you want to set your conversion price higher than your current stock price is because you don't want your bondholders, holders to convert immediately. If the conversion price is lower than your current stock price or the same as your current stock price, then that's just unnecessary dilution. When you issue it, they get it, they immediately, you know, convert it to equity if they want to, and you've just diluted your shareholders. That's not great capital allocation. Um, so you wanna set the conversion price higher. As an executive, ideally, you want to set it as high as possible, but investors want, uh, the conversion price to be set as low as possible, uh, close to your current stock price. So there's a bit of a give and take here, both in terms of the interest rate. Investors want the higher, higher interest rate, you want a lower interest rate. Uh, and also the conversion price. Investors want a low conversion price, you want a high conversion price, right? And so the, the equilibrium point, the give and take between this two, these two things, um, is what eventually determines, like, whether or not you, you get the price you want and the conversion price you want and the interest rate you want. Anyway, all of this was being worked out, but, uh, conceptually as an instrument, this makes a ton of sense, especially if your stock is really hot because you're getting a, you're getting much cheaper debt at a lower interest rate when you go and exploit, uh, the market's excitement over your equity value or the perceived, uh, possibility of, uh, of your equity value going up. Okay, so Amazon closes this, and the reason why I said that they are lucky, this is my second point, right? Um, the dot-com bubble bursting was not predictable. Uh, any serious student of markets would say that you can't predict when a bubble bursts. And the reason for that is because there's huge amount of volatility both on the way up and on the way down from the peak of a bubble, which again, you can only know retrospectively. Uh, what does this mean? Well, consider, uh, an Amazon executive's, uh, experience. You IPO in 1997, right? And in 1997, there's the Asian financial crisis. Uh, it is a huge deal in Asia, uh, but it's thankfully not a huge deal in Western markets. But still, the market is scared. It's scared for a couple of days. Uh, in, in 1997, the S&P 500 fell 6.8% in one day, um, and the total decline, uh, in response to the AFC was, uh, 13% before the market recovered and then zipped right back up in the midst, uh, in the grips of the dot-com mani-mania. And then one year later, right, in August 1998, Russia defaults on its domestic debt, and, uh, that, that wouldn't be, uh, bad news in itself. However, uh, there is a, uh, the world's most sophisticated, one of its most famous hedge funds, Long-Term Capital Management, blows up during this time, right? It, it-- the Russian sovereign debt crisis, uh, devastates, uh, LTCM's convergence trades. And the reason this is significant is because LTCM had so many cross-holdings that people was, uh, the, the New York Fed and various other, uh, financial market players were all scared that the contagion would spread to other, uh, firms, to other banks, such that the New York Fed actually ran a coordinated bailout. So there was so much fear that LTCM would take down the entire stock market, uh, the banking system. Between August 1998 and October 1998, the S&P 500 falls by around 30%, right? And, um, you imagine that you are an executive in Amazon, your stock is going up. Uh, there's a blip in 1997, you're a bit scared for a while, and then it continues going up. In 1998, over the course of three months, you see your stock market getting smashed by around 30%. So the entire S&P 500 falls by 30%. Nasdaq even more, actually. I think Nasdaq was like 50%. Um, and, and there were multiple, uh, what do you call it, trading halts during this period. Um, now after, uh, the New York Fed, um, um, saves LTCM, and there's no way to know this, right, before it happens. Uh, in fact, I think years later, te-- more than a decade later, Marc Andreessen, uh, who was, uh, at the time the wunderkind in charge of Netscape, said that everybody thought that the dot-com bubble, which, you know, most people knew that they were in a bubble, they thought it would finally burst in, uh, those months in 1998. But then what happens? What happens is that after October, right, uh, the market recovers 50% in the subsequent three months after October 1998, and over the next year z- roars vertically by 256%, uh, to the p-peak of the dot-com bubble, uh, slightly more than a year later at the end of 1999. So there's no way you can predict, right? If you have gone through this experience, uh, who is to say that the decline of at the, at the top of the bubble, um, in, uh, the end of 1999, the start of 2001, uh, who, who is to say that that decline wouldn't have then corrected once again and the market would rip up another, you know, 500, 200, 500%, right? Uh, this is why bubbles are so, um, difficult to deal with. The most famous story, I think, from that time was Stan Druckenmiller. He managed to sell out of the dot-com bubble, but he couldn't take the vertical, uh, ripping of the stock market at the final phase, and so he bought right back in at the worst possible time, and he lost a lot of money, so much so that he had to take a break from trading completely. Like, I think he went off-- He said that he put himself out to pasture for a couple of months, uh, put himself on a beach somewhere with no access to any trading screen. So when Amazon closed the convertible debt, let's be clear, this is luck. They closed it at just the right time, just a couple of months before the peak of the bubble, but there was no way anyone could have predicted, um, that it would have burst when it did. Um, now the third point I want to make is that notice that the reality of Amazon's business is very different from the slaughter of the public markets that was, uh, happening at the time, right? Um, they, they were a lot of blocking and tackling in response to Ravi Suria, uh, to make sure that their negative cash conversion cycle didn't unwind. But the fact is, after the dot-com bubble burst, right, Bezos and team, um, they had to do a round of layoffs, they had to slow down on expansion, and then they had to demonstrate to Wall Street that they were profitable, that, that they could be profitable. And so Amazon shifted to profitability, uh, just to show that they could over the course of four years. So in 2001, they showed their first quarterly GAAP net profit. Then in 2002, financial year 2002, right, which means that they show this, uh, when the results are released in 2003, right, they show the first full year positive GAAP operating income in the financial year 2002. And then finally, uh, in financial year 2003, they demonstrated first full year positive GAAP net profit, uh, announced in January of 2004. Now, of course, after th- after that in 2004, Bezos takes Amazon back into the red again, uh, with the launch of Amazon Prime and a whole bunch of other initiatives which are covered in other cases in the case library. But, um, notice that Amazon had a working business and they could, uh, uh, you know, as they, as they, as you webmates put up or shut up, right? So they, they, they did put up to sort of, uh, show everyone that, "Hey, we have a working business here. We are not just one of those dot-com darlings that have nothing behind them." Um, and, uh, the final point, like this is like sort of two related points. Uh, Bezos was very strong in the vision. He knew he wanted to build an everything store, and so the fact that he had to, uh, switch and ha- get a huge chunk of his revenue during these difficult times from white labeling e-commerce solutions from Toys "R" Us, uh, Borders, I think it was Borders, right? Um, well, certainly Toys "R" Us and Target. Um, that must have been, uh, painful to, to do that, but he needed to generate the cash. Um, and, and related to that, notice how the entire leadership churned during this period because it was just so psychologically difficult and painful and stressful, um, to, uh, go through the dot-com bust and to then demonstrate and transform the business enough to demonstrate to Wall Street over the course of four years, uh, that, "Hey, this business actually has legs." So yeah, um, anyway, that's my sense making for this case. I would be very interested to see and hear, uh, what other things you notice. But I was focusing on the fact that Amazon shored up its balance sheet right at the right time, um, and it just so happened to have a working business, uh, around the same time that, uh, the dot-com bubble burst and it couldn't really tap any, uh, external funding, uh, for its business for a number of years.