Regular readers of the Fool (with memories long enough to remember "Amazon.bomb") should not have been surprised to read the article "Burning Up" in the March 20, 2000 issue of Barron's. The main thrust of the piece was that many "Internet firms" are rapidly running through their available cash and may have trouble finding further funding.

Presumably, inspiration for the article came primarily from Peapod Inc. (Nasdaq: PPOD), who had announced a week earlier that it had lost $120 million in financing and its CEO had resigned. Starting from this train wreck, the article moved to some other easy targets, including drkoop.com (Nasdaq: KOOP) and CDNow (Nasdaq: CDNW) -- a litany of woes that has become de rigueur among the fashionably bearish.

Now, if Barron's had left it at that and asserted simply that some Internet companies did not have viable business models, Fools would have applauded. A discriminating look at 'Net stocks! Perhaps Barron's has turned a corner and started to distinguish the wheat from the chaff among the Wired and Unprofitable!

But no. What's true for Peapod, according to Barron's, must be true in general for "Internet firms" with negative earnings. The article quickly moved from Peapod to Barron's favorite 'Net target, Rule Breaker Amazon.com (Nasdaq: AMZN). Pegasus Research International, Barron's claimed, had determined that 51 companies did not have the cash to make it to the end of this year, and that Amazon would run out of cash in 10 months.

Others have discussed sufficiently the enormous flaws in Pegasus' methodology, which appears to have consisted solely of dividing the companies' cash on hand as of December 31 by operating losses from the fourth calendar quarter of 1999. Barron's follow-up issue featured four letters from officers of some of the named companies that detailed the unfathomably simplistic nature of the "research." I won't beat that dead winged horse.

What caught my eye in the article was this comment: "Indeed, E*Trade (Nasdaq: EGRP), Amazon.com and Ameritrade (Nasdaq: AMTD) have had to offer unusually generous interest rates and conversion terms to sell their latest offerings of convertibles."

I have to admit that I had never looked carefully at any of Amazon's bond offerings, much less the most recent one, which occurred on February 16, 2000. I decided to take this opportunity to check them out. Amazon's 10-K (which helpfully contains details of the February offering even though it didn't occur during fiscal 1999) just appeared a couple of weeks ago, so fresh data is at our fingertips.

Time for� Fun With Bonds!

Amazon's latest offering was E690 million (that's euros, folks) worth of 6.875% convertible/callable bonds, due 2010. If you understand that, you're one up on most of us. I'll try to describe it.

(Note that the euro traded at about $1.01 on February 16, so, for simplicity's sake, I'm going to use dollars. Keep in mind, however, that currency fluctuations will affect every amount mentioned.)

On its most basic level, the purchaser of a $1000 bond gets $68.75 once a year until 2010, when the initial $1000 is returned with that year's interest.

The "convertible" part means:

  • The purchaser can convert the $1000 principal into Amazon stock at a predetermined price.
  • The initial conversion price in this case is about $105 per share.
  • The price may be reset on February 16 of 2001 and 2002 to the average trading price of the previous 20 days, but not lower than $85.
  • Amazon traded at $77 on the day the bonds were issued, so the initial conversion price represents a 36% premium and the lowest possible conversion price comes at a 10% premium.

The "callable" part means:

  • Amazon can elect to redeem some or all of the bonds at any time.
  • If the bonds are called, Amazon will pay the holders their $1000, plus at least $206 minus previously paid interest (called a "make-whole" provision), giving the holders the option to convert the cash to stock.
  • There's one catch: Before 2003, Amazon can only call the bonds if its stock has traded at an average of $168 for 20 of the previous 30 days.

On the face of it, those seem like pretty sweet terms to me. The interest rate is low (10-year U.S. Treasury bonds paid 6.5% in February), and Amazon may be able to pay the debt off with stock rather than cash. It's very much like using options to compensate employees, which companies have found to be an attractive proposition for various reasons discussed here. If the stock price starts to become too high to make that option attractive, Amazon can simply call the bond for the 6.875% plus the principal in cash or stock.

Perhaps Barron's meant that this bond offering is generous compared to Amazon's previous bonds. Let's take a look at all of Amazon's offerings: 

Date  Size(MM)  Call   Rate  Premium  Rating
5/98     $326    yes  10.0%      n/a      B3
2/99     1250    yes  4.750      27%    CAA3
2/00      690    yes  6.875       36    CAA3

All of Amazon's bonds are callable (and the company has already called about half of the May 1998 non-convertible offering). Compared with the other convertible/callable bond, the new offering has the same Moody's rating, a higher interest rate, and a higher initial conversion price.

About half of the higher interest rate can be attributed to the rise in interest rates in general. You may remember that the Fed lowered rates several times in the fall of '98, after the Asian crisis struck. Those rates still held in February 1999, making Amazon's biggest offering particularly well-timed. The Federal Funds interest rate was almost 100 basis points higher this February than it was then. The 10-year U.S. Treasury bond and 30-year mortgage rates are over 150 basis points higher in 2000 than they were in 1999.

The remaining 100 or so basis points can be assigned to the higher conversion premium. The 36% premium for this year's offering is the highest ever for a premium make-whole bond (which, by the way, was introduced for the first time by Amazon in February 1999). It's true that the premium could dip as low as 10%, providing bond purchasers with something of a hedge, but the presumptive rate stands quite high.

It's too tedious a task to discover whether Barron's may have meant that Amazon's terms were generous compared to other corporate offerings of Amazon's grade, since the time and terms of other issues vary so much. From a quick glance at the numbers, it doesn't seem exceptional. At least I am satisfied that this year's offering was not so much different from Amazon's previous offerings.

Seems to me that the Barron's comment about Amazon's bond is just another example of overgeneralization in a woefully generalized article.

--Brian Lund, TMF Tardior at work and play