The Case for Amazon, Part 2: Lehman Report Didn't Scare Away Legg Mason

When was the last time you remember a bond analyst knocking the wind out of a stock? Can't say I can, but that's just what happened last week, when Lehman analyst Ravi Suria questioned Amazon.com's (AMZN Quote) biz model. His report caused Amazon's stock to tumble, which was a buying opportunity for the Legg Mason Value Fund. (See Part 1 .)

Randy Befumo, a Legg Mason analyst, intimately involved in the company's Internet research, gives his analysis of the Lehman analysis:

The key issues raised in the Lehman report are, Will Amazon run out of money by the first quarter of 2001? And, Does Amazon's economic model work?

I think it is critical to realize that a bond analyst of credit has a different job than an equity analyst. The value of equities is 100% in the future. Creditworthiness, however, is completely tied to the present.

If I estimate a company's value based on last quarter's results and do not look into the future, I have a pretty low chance of being right. However, if I estimate its creditworthiness on last quarter's results, I have a 100% chance of being right.

Obviously, if I want to forecast changes in creditworthiness, last quarter won't help me very much at all. Being an analyst of equities is more like forecasting future creditworthiness than obsessing over last week's box scores.

Will Amazon run out of money by the first quarter of 2001? The Lehman conclusion is dependent on the fourth- and first-quarter results being perfectly indicative of the future. "We believe that for Amazon to be a successful business, it must be able to generate the cash-operating profile of a successful retailer. It is essentially this yardstick that we use to analyze the company, and ... we find it woefully lacking from an operational aspect..."

Of the four variables cited, capital expenditures and working capital [inventories and payables] are worth examining. The Lehman analyst himself notes that despite the leverage, we do not believe the company is currently faced with any default issues [given the first debt matures in 2008], indicating that the debt issue is really not worth paying attention to from a liquidity standpoint.

On capital expenditures, to refresh our memories, Amazon currently believes that it will be operating cash-flow positive through the fourth quarter and that capital expenditures will be equal to depreciation for the full year. Given that the company sets its own capital budget, this makes the claim of "rising cap ex" at least challenged if not specious. For some reason, every single analyst who looks at Amazon and comes up with a negative conclusion takes as a given that capital expenditures will continue to rise at a rapid pace despite the company's claim to the contrary. The leading indicators seem to support this, as capital spending dropped from $105 million in the fourth quarter to $37 million in the first quarter.

Amazon has almost complete control of its capital budget, and if it were necessary, could slow down its spending to whatever level is required. Is Amazon's claim of lower capital expenditures for 2000 reasonable? In 1999, Amazon increased its distribution square footage about 10 times to support what I estimate to be $8 billion to $10 billion of annualized revenue based on the productivity the company achieved prior to the rapid buildout.

Amazon should not need to expand the U.S. distribution capacity until 2001 at the earliest. Capital expenditures just don't go up over time magically ... you actually have to spend it on something. Unless you think Amazon is lying about the amount of distribution space it bought or that productivity per square foot will drop 90%, there is no real need to spend at that level in 2000 unless the company exceeds these sales levels. I believe that Amazon will be able to hit its goal of capital expenditures equaling depreciation, which from a cash standpoint will be a wash.

As for working capital, I am unclear about the Lehman analyst's level of knowledge about the company. His comment that "Amazon has limited operating expertise in logistics..." appears to be ignorant of the fact that Amazon's logistics are run by the former logistics team that managed Wal-Mart's (WMT Quote) system. In fact, Wal-Mart was so peeved that Amazon hired these two gentlemen away that it sued Amazon back in 1998.

Furthermore, the Lehman analyst appears ignorant of the benefits of a negative working capital float. Even at its worst [fourth quarter 1999], Amazon's payables of 48 days eclipsed its inventory of 18.3 days. This means that Amazon pays for stuff about 30 days after it actually manages to sell the goods. Essentially, the company is being financed not by the capital markets [as claimed] but by its own suppliers. To claim otherwise, you have to isolate the first quarter from the fourth quarter, which for a seasonally driven retailer would be a mistake.

The fourth quarter sees an inflow from payables; the first quarter sees an outflow to suppliers and a drain on cash balances. It is the combination of these two quarters that creates an operating trend, not one or the other in isolation. In fact, it is even better to look at the full trailing four quarters. The Lehman analyst actually agrees with this at the beginning of his report, as he writes, "typically in a seasonal business like a retailer, the operations draw cash into working capital in the third and fourth quarters, as inventory is piled up, and release cash as the holiday sales are paid down." Or translated, any one quarter's results will be impacted somehow by seasonality and to get a sense of the overall trends, looking at the full year is likely to be more constructive.

Beyond focusing on the results of one quarter [rather than looking at one-year time periods where all of the working capital issues are balanced out], there is also basic misunderstanding of the nature of Amazon's working capital. The Lehman analyst is particularly focused on Amazon's ability to manage working capital, as he claims that "fulfillment, high inventory turnover, and effective working capital management are all operational issues that Amazon has shown an exceedingly high degree of ineptitude in."

This seems strange given that Amazon, since inception (despite significant operating losses of $791 million), has a cumulative loss of operating cash flow loss of $61 million. Operating cash flow contains all operating and working capital cash flows. What this means is that working capital and other operating items have contributed $731 million to the company since inception. The claim that Amazon is inept is somewhat forced. Certainly, the fourth quarter and the first quarter were not great, but I think the evidence suggests that overall, since inception, the company has done an excellent job.

As for whether or not the model works, I think the Lehman report actually holds the answer to this -- an affirmative answer. The Lehman analyst notes that Amazon's cumulative free cash-flow loss since Q4 of 1997 is $718 million. Over that same period, the company has done $2.9 billion in sales. This means that the average dollar invested since 1997 [$359 million] has generated about $8 in sales.

Obviously, this is a massive cash outflow, but for a company that has raised $2.8 billion in cash since inception we can clearly see that the $718 million in outflows is not even close to equaling the cash hoard the company should have left.

The net working capital investment looking at the fourth quarter and the first quarter together was only $22 million less than the inventory the company ended up writing down. Even if the net investment in working capital over the fourth quarter and first quarter is quintupled in fourth-quarter 2000 and first-quarter 2001, it will not even come close to draining Amazon's cash balances. Consequently, I think the most inflammatory claim of the Lehman piece -- that Amazon will run out of cash in first-quarter 2001 without more financing -- is just wrong.

Coming up later ... Part 3: The Bottom Line -- Legg Mason's (and Randy's) argument for Amazon.