fool.com: Building Costco Model (Bore Port) September 15, 1999

Boring Portfolio

Boring Portfolio
Building Costco Model
and random thoughts

By Dale Wettlaufer (TMF Ralegh)

ALEXANDRIA, VA (Sept. 15, 1999) --Today I've been working on the model for Costco Wholesale (Nasdaq: COST) with the goal of helping Bore readers who are interested in modeling a forward business plan. As you'll recall, I am re-building this model from the ground up, trying to get a handle on the individual value drivers all the way from store-level operations to the consolidated financial picture. Getting the store-level stuff down is going to be much more difficult and always depends upon a lot of guesswork and as much guidance as one can get from the company. Getting the value drivers down on the consolidated level is a bit easier if you like a particular sell-sider's work, you have a feel for the company, and you've laid down some of the groundwork necessary to understand what has happened in the past.

One very good report I've read is from Dan Binder at Brown Brothers Harriman (initiation of coverage, August 23, 1999), particularly because he goes beyond just providing a string of income statement projections with no accompanying balance sheets or cash flow statements. I'm sure I'll use his model for guidance along the way, and so far I've used it to plug in the components of net interest expense for the year, but the model I'm developing won't rely very much on any sell-side models.

The model here is going to be a demand-driven one, starting from the topline and proceeding from there. Working those assumptions through the balance sheet and cash flow will take some tweaks here and there to adjust for the changing composition of sales. Increased revenues from services such as the optical unit, photo, and gas stations may change working capital and margin dynamics, all of which have their effects on the three financial statements. So as we go along here (I'm taking off next week for a week in Colorado, so there will be a break in the action coming up, filled in ably by Whitney Tilson, a fund manager in New York), this model is sure to change as my knowledge of the strategic evolution of the business (and mundane things like its tax accounts) improves. Feel free to make any suggestions on the model or anything else on the Boring Portfolio board.

Eventually, the model will be look at things from a discounted earnings, discounted cash flow, and an economic value added perspective. The spreadsheet I posted on Monday contained some old assumptions -- I've trashed some stuff on this spreadsheet, so many of those things are gone here.

Other Random Stuff

You know, I like the work of Herb Greenberg (formerly of The San Francisco Chronicle, who had the "bizinsider" column on AOL long before a lot of journalists had even begun to understand online publishing) of TheStreet.com. I think he does excellent work at TheStreet.com and I never have a problem in saying so, though they do compete with the Fool and that will probably intensify when they roll out their message boards. However, I do have a problem with comments he has made recently on two companies I follow.

The first is American Power Conversion (Nasdaq: APCC), the West Kingston, Rhode Island, manufacturer of power management systems for PCs, servers, and data centers. The comment came during TSC's August 28th TV show. I didn't know about it until late last week and I wouldn't care about it that much, but after reading Greenberg's column today on another company (which I'll get to in a moment), I thought I'd comment.

Ken Shapiro of Condor Capital was talking about the company as one of his stock picks, and Greenberg made the following observation: "But you know what, their inventories are high. These are things that people don't want to talk about. Balance sheet issues, high inventories, high receivables. This suggests to you that the company's stuffing the channel maybe..."

The human mind works on models and takes shortcuts to make on-the-fly analyses easier. If a pitch is coming at your head, move out of the box quickly, or if the back-end of the car in front of you lifts up abruptly in heavy traffic, that tells you more about the car's rate of deceleration than just its taillights will tell you. We engage in these logical progressions all the time to smooth decision-making along. In this case, though, Herb's assertion that APC might be stuffing the channel is way off base.

APC has always run high inventories because:

1. The company has a high number of stockkeeping units (SKUs);

2. It wants inventory on hand to capture the sale orders that might come in from emergency situations such as power outages arising from natural disasters;

3. The inventory doesn't depreciate very quickly, given that the value-added doesn't come from rapidly depreciating DRAMs and CPUs;

5. The value-added comes partly from software;

6. Which consequently makes this a higher-gross margin company, almost twice that of the PC industry;

7. These high margins allow the company to make up the inventory carrying cost on a relatively small amount of incremental sales -- which you chalk up to customer acquisition costs or retention costs, anway;

8. Even given the unattractive inventory turnover of APC, the company still generates returns on capital in the 25% to 35% range. The strategy sure seems to be working.

The company runs its business with large inventories by design, and it'll tell you that. I think it's kind of bogus to just suggest it is doing something misleading, which is the direct implication when you're suggesting a company is engaging in such activities. While it's right to question the working capital dynamics of the company, it's not right to leave it there. And I'm not just talking up the position the Boring Portfolio has in this issue, I'm saying I don't think it's at all fair to engage in a sort of Joseph McCarthy-esque line of questioning: "No need to worry, Mr. Mostel, we're just asking questions, we're not casting aspertions on your character."

Hey, sorry to go nuclear and compare Greenberg to McCarthy, but you can't just look at a balance sheet and say a company is stuffing the channel, however much you dress up the questioning with "might" and "maybe."

The article that made me include this in today's column, however, was on Estee Lauder (NYSE: EL), which I've followed for a couple years at the Fool. Today Herb quotes a hedge fund guy who is short Estee Lauder, and who contends the company made its earnings numbers by taking an "unexpectedly" low provision for doubtful accounts (which is a debit that reduces earnings and is a credit to the accounts receivable contra account). The hedge fund guy compares 1999's allowance for doubtful accounts to 1998's (as a percentage of gross receivables) and proclaims the company as shady because the allowance at year-end 1999 is lower than last year: "[I]f the doubtful-accounts reserve at Estee Lauder had been in line with the year-earlier reserve -- which presumably included the same retailer settlements -- earnings would've been light by around 2 cents per share."

I think Herb goes deep much of the time, but he whiffs here in taking the hedge fund guy's numbers at face value. My numbers say this:

(in millions)

                    FY99      FY98     FY97    FY96  
Gross receivables $569.70  $541.40   $508.10  $466.80
Allowance         $36.00    $43.60    $36.40   $32.80
Net receivable   $533.70   $497.80    471.70  $434.00
Allowance as %     6.32%     8.05%     7.16%    7.03%
Provision.        $27.80    $25.40    $23.60   $22.00
Charged off,
net of recoveries $35.40    $18.20     $20.00  $20.00

As you can see, 1998 was abnormally high -- way above normal trend. Looking at net charge-off activity in 1999, the higher provision in 1998 (provisions lead net charge-offs) was taken in anticipation of higher losses in 1999. Since over 50% of Estee Lauder's net income before taxes and minority interest is generated internationally, and since the Russian debt crisis happened in the fourth quarter of last year, and since retailers default on payables not at the monent an economic crisis takes place, but many months afterward, I think we can see the progression of the 1998 provision into the 1999 net charge-offs.

The allowance as a percentage of receivables may be lower if the company feels there is a significant chance of recovering those charged-off accounts in fiscal 2000. Inside the receivables accounts, the composition of customers might be different, as well, if after the large charge-off in 1999 the company has not extended terms as generously to less dubious credits than before.

The bottom line is that it's bogus to ascribe to mal-intent the lower provision for credit losses when 1998's provision and ending credit loss balance was clearly above trend. I'm not talking my position here, either. I have one at a low cost basis. The hedge fund guy has a short position, which is disclosed. The difference is that I think his analysis is wrong, and I think Greenberg is as well.

Have a good Wednesday night.

Fools Wanted: Apply Within.