FOOL ON THE HILL
Questioning Dell's Returns on Capital

Dell is a fine company that generates very good returns on capital -- just not as good as one might think at first glance. When calculating return on invested capital for the PC maker, or any other business, it's important to carefully consider the inputs if the result is to be meaningful. Dell is a perfect illustration of this.

 Format for printing Reuse/Reprint

By Whitney Tilson
June 5, 2001

In last quarter's earnings release, Dell (Nasdaq: DELL) reported return on invested capital (ROIC) of 891%. Think about what that means for a second: Dell is claiming that for every dollar invested in its business, it generated nearly \$9 of returns. When even the best companies are lucky to sustain a ROIC above 20%, Dell's claim is pretty unbelievable. Let's take a closer look.

Return on invested capital isn't the scientific measure it's cracked up to be. If you asked 10 experts to calculate Dell's ROIC, I'd bet you'd get 10 different numbers (Fool member Andrew Chan laid out one good take on the subject). I don't spend a lot of time trying to calculate ROIC down to the last decimal place because I don't care much about a company's past ROIC. It's the future ROIC, which is a function of a company's sustainable competitive advantages, among other factors, that matters. That being said, it's critical to have a general idea of whether a potential investment has generated poor, decent, or excellent ROIC over time.

Calculating Dell's ROIC
To calculate Dell's, I take the past 12 months' after-tax operating profits and divide by average invested capital, which is generally all debt plus equity (including preferred stock and the like), minus excess cash. (ROIC can get a lot more complicated -- books have been written on it -- but let's keep it simple.) Over the past four quarters, Dell's after-tax operating income (excluding, as Dell does, a \$105 million charge in Q4 01) was \$1.92 billion.

```                              Q2 01  Q3 01  Q4 01  Q1 02  TOTAL
---------------------------------------------------------------
After-tax operating income*     \$515    573    412    421   1921*All numbers in millions```

For invested capital, I added debt plus equity, and then subtracted all other cash, cash equivalents, and short-term investments, net of 5% of annualized revenues (Dell uses this assumption for the amount of cash the business requires).

```                            Q1 01  Q2 01  Q3 01  Q4 01  Q1 02
-------------------------------------------------------------
Equity                      \$5623   6462   6057   5622   5505
+ Debt                        508    510    510    509    508
- Cash & equivalents         3459   4076   3976   4910   4886
- Short-term investments      269    240    638    528    381
+ 5% of annualized revenues  1456   1534   1653   1735   1606
-------------------------------------------------------------= Invested capital           3859   4190   3606   2428   2352
```

(Notes: All numbers in millions. Dell's fiscal years are one year ahead of calendar years -- in other words, Q4 01 ended 2/2/01.)

When calculating ROIC, one typically uses average invested capital over the time period in question. In this case, we can take the average of the first and last values (\$3,859 million plus \$2,353 million, divided by two, which equals \$3,105 million) or use the average of all five quarters, which is \$3,287 million. The numbers are close enough that it doesn't really matter, but to give Dell the benefit of the doubt, let's use \$3,105 million.

So, taking after-tax operating income of \$1,921 million and dividing by \$3,105 million, we get ROIC of 62%. That's very impressive, but it's hardly 891%. Even if we use the most recent (and lowest) invested capital figure of \$2,352 million in the denominator, ROIC is still "only" 82%.

I spoke with Dell's investor relations department today and asked for an explanation of how the company calculates ROIC, but they were unable to give me an answer in time to include in this column. If I get an answer, I'll report next week.

Thoughts on Dell's invested capital
Regardless of how it's calculated, Dell's ROIC has risen rapidly, despite weaker profits, due to its declining invested capital, which (as I calculated it anyway) has fallen 44% over the past three quarters. If this keeps up, Dell will have no invested capital at all, and its ROIC will be infinite. What's going on?

Each quarter, a company's net income is added to its equity, so over time, the equity of profitable companies typically rises. Of course, if profits are rising as well, return on equity can remain relatively constant. But the equation changes when a company repurchases its shares. In this case, the dollar amount spent on repurchases is deducted from equity, so if a company spends more repurchasing shares than it earns in net income, equity falls. That is exactly what's been happening with Dell.

You might ask how Dell spends more repurchasing shares than it earns in net income without using debt. Good question. Not many companies could do this, but Dell generates substantially more free cash flow than net income, due to its negative cash conversion cycle -- it collects from customers before it has to pay its suppliers for inventory.

In fact, over the past three fiscal years, Dell has earned \$5.3 billion in net income, but generated 43% more free cash flow, \$7.6 billion. (It also generated an additional \$2.4 billion from "Tax benefits of employee stock plans," but I exclude that from free cash flow, and discuss the effects options can have on ROIC further below.) Dell has then used \$5.3 billion to repurchase shares, including \$1.7 billion in the second half of last fiscal year alone. It's not surprising, then, that equity has declined and cash has risen, both of which reduce invested capital and increase ROIC.

The impact of stock options
Regardless of whether you use my ROIC numbers or Dell's, I think the figure is significantly overstated, not because I think invested capital is understated, but because Dell's liberal use of stock options to compensate its employees causes its profits to appear higher than they really are. (One could also argue that the value of these stock options should be included in the calculation of invested capital, but I'll leave that discussion for another time.)

To illustrate, let's take an extreme example: Say there are two identical companies, both with \$1 billion in equity, \$1 billion in annual sales, \$500 million in expenses (not including employee compensation costs), and a 30% tax rate. But Company A pays its employees \$300 million in compensation, while Company B pays its employees entirely in stock options. Here's what each company looks like:

```                           Company A     Company B
--------------------------------------------------
Sales*                         \$1000          1000
Non-compensation expenses        500           500
Compensation                     300             0
Operating income                 200           500
Taxes                             60           150
Net income                       140           350
Profit margin                     14%           35%
Return on equity                  14%           35%*All numbers in millions
```

With its much higher margins and returns on capital, Company B is much more attractive, isn't it? The situation is even better than it appears because Company B doesn't really pay \$150 million in taxes. When employees exercise their options, the difference between the strike price and the exercise price is taxable to the employee and tax deductible to the company -- just like any form of compensation -- so the company gets back part of the \$150 million. (This amount appears in the cash flow statement under "Tax benefits of employee stock plans.")

It gets even better for Company B over time. Let's say both companies generate free cash flow equal to their net income and then use this money to buy back shares. Since the increase in equity due to net income is exactly offset by the amount spent on share repurchases, the equity base remains constant. But don't forget the extra cash Company B gets from "Tax benefits of employee stock plans" (As noted above, I don't consider this free cash flow, even though it often appears, as is the case with Dell, under "Operating Cash Flow"). Let's say this amount is \$50 million for Company B, meaning that it can reduce its equity to \$950 million after one year -- and therefore increase its return on equity to 36.8%. Run this out over a number of years, and the impact becomes larger and larger.

Dell, of course, is neither Company A nor Company B -- it's a bit of both -- but I hope this example helps you understand why I think its profits and free cash flow are overstated.

Conclusion
When calculating returns on capital -- whether for Dell or any other company -- investors need to be aware of the inputs that determine the calculations, and make appropriate adjustments. Like so many investing formulas, the quality of the output is only as good as the input.

-- Whitney Tilson

Guest columnist Whitney Tilson is Managing Partner of Tilson Capital Partners, LLC, a New York City-based money management firm. He did not own shares of Dell at press time. Mr. Tilson appreciates your feedback at Tilson@Tilsonfunds.com. To read his previous columns for The Motley Fool and other writings, visit http://www.tilsonfunds.com/.