Boring Portfolio
Further Arguments for
American Power Conversion
Plus, the power of the flow ratio

By Whitney Tilson

NEW YORK, NY (September 22, 1999) -- After learning about American Power Conversion <% if gsSubBrand = "aolsnapshot" then Response.Write("(Nasdaq: APCC)") else Response.Write("(Nasdaq: APCC)") end if %> from Dale's columns in the Boring Portfolio earlier this year and doing my own research on the company, I purchased the stock myself. I think it's an outstanding company, with a dominant and defensible position in a rapidly growing, increasingly important industry. It has high margins (13%+), plenty of cash ($272 million, equal to 7% of its nearly $4 billion market cap), and almost no debt ($1 million in short-term debt). It also generates consistent, very high (30%+) returns on invested capital.

Furthermore, for a stock that has done so phenomenally well, it is remarkably unknown (APCC has compounded at 49% annually for the past 10 years, a rate exceeded by only 12 companies of the nearly 3,000 for which there is 10-year data, according to Value Line). I think APCC's anonymity explains its very reasonable price (less than 23x trailing earnings as of yesterday's close) relative to the quality of the business and its bright future prospects.

Two events in the past few weeks have focused my attention on APCC (I am assuming that readers of this column are already familiar with APCC; if not, read Dale's columns and my buy report). First, I attended APCC's presentation at the annual Salomon Smith Barney Technology Conference (mostly for securities analysts) on September 8th and I'd like to share some highlights. Second, TheStreet.com's Herb Greenberg, during TheStreet.com's TV show, suggested that APCC is "stuffing the channel" (e.g, knowingly shipping excess product to artificially improve its financial statements). I think that's a false accusation and will explain why. I would also like to examine APCC's flow ratio over time, and show how it has been a useful tool in predicting what happens to the stock.

APCC's presentation was given by Don Muir, CFO, and Aaron Davis, Vice President, Small Systems Group. There were no major revelations, but here are some of the highlights:

• Growth in the first half of 1999 has been excellent, with revenues up 23.6% and EPS up 25.7%.
• A few years ago, APCC's customers were primarily concerned with hardware and data protection, but with the explosion of the Internet, the primary concern today is 99.999% availability (witness the billions lopped off eBay's market cap when its website crashed in June). According to APCC, power failures are the leading cause of server failures (31%), and downtime can cost some businesses millions of dollars per hour. The explosive growth of the Internet (APCC referred to it as "a great big cash register in the sky") is fueling APCC's growth, yet unlike most Internet companies, APCC is highly profitable and has a much lower level of risk associated with its business model and products.
• APCC is growing rapidly overseas, where power quality is poor and there is tremendous growth in the use of computers, servers, and other sensitive electronic equipment that needs to be protected from outages, surges and other power problems. In the second quarter, revenues grew 8% in the Americas, 46% in Europe/Middle East/Africa (EMEA), and 46% in Asia. EMEA now accounts for 29% of sales (vs. 27% for the full year 1998) and Asia for 13% of sales (vs. 11% in 1998). APCC recently opened new plants in Bangalore, India and Suzhou, China, and has already become the most preferred brand in China. APCC also has plants in Ireland, Denmark, Switzerland, and the Philippines.
• APCC's products for servers, data centers and facilities (Smart-UPS, Matrix-UPS, Symmetra Power Array, Silcon products, software, and accessories) are becoming a larger and larger share of APCC's revenues, accounting for 69% of sales in the first half of 1999 vs. 63% in 1998. The balance is products for desktop PCs (SurgeArrest, Back-UPS families).
• In a 1998 study by InformationWeek, APCC products are preferred by enterprise professionals 10 to 1 over the #2 brand. Davis said, "We have an intensely loyal customer base."

During the Q&A, Muir and Davis covered three topics:

• When asked how APCC can compete with competitors that are part of corporations that are much larger than APCC, they said they are much more focused and, with more than $1 billion in sales and high profit margins, APCC has ample resources to compete effectively. They also told the story of Hewlett Packard's attempt to penetrate the market in 1995. HP's products were inferior and customers and distributors remained loyal to APCC, so HP never achieved more than 1% market share and exited the market in 1997.
• When pressed about the inventory issue (which I'll discuss further below; Dale has also written about it), they said APCC tries to keep inventory as low as possible, but it needs to be able to offer rapid delivery to customers around the world or it risks losing them. They also mentioned that, in contrast to the PC business, there is a low risk of product obsolescence and price declines. Finally, they said that at the high end, they are moving toward a build-to-order model.
• When pressed by analysts to increase their earnings guidance for the year, they refused, saying they had met or exceeded consensus earnings estimates for 14 consecutive quarters, so they would prefer to keep guidance conservative (as Muir said, "We're focused on running the business and beating our competitors [not on these games]." Gotta love it!). Frustrated by this answer, the analysts then asked why APCC wasn't buying back shares to boost the share price. Muir answered that share repurchases precluded acquisitions via the pooling method, but conceded that this method is about to be outlawed so it's no longer a good reason. Sure enough, two days later APCC announced a 10 million share repurchase program (representing 5.1% of the shares outstanding), which contributed to the stock's 13% rise in the two days after this meeting.

Now, I'd like to turn my attention to Herb Greenberg's comments about APCC on August 28th. Here's what he said (click here for the full transcript):

Dale published a scathing rebuttal last week. I agree with Dale and want to add to his comments. Greenberg, Dale, and I all agree that APCC's inventory level is quite high, especially after it spiked in Q2 '98, and it was definitely a concern when I first bought the stock (as I noted in my buy report). However, after hearing the company's explanations, looking at the trends and analyzing the numbers -- especially the last quarter's results -- I'm sanguine about the issue for now (though I will keep a close eye on it). Here is a graph showing APCC's days of inventory since Q4 '94 (the earliest period for which I could find data):

Days of Inventory

You can see that APCC has always run relatively high levels of inventory -- it's part of their business model (which has been highly successful over time; during this period, APCC's stock compounded at 20% annually). For three quarters, from Q3 '97 through Q1 '98, days in inventory dropped significantly, but that appears to have been an aberration. It spiked in Q2 '98 due to the acquisition of Silcon, but note that it did not increase from there, and, most importantly, look what it did in the most recent quarter: it dropped from 136 days to 113. This is "stuffing the channel"? Greenberg might have had an argument a year ago, but he certainly doesn't now.

What I find more interesting than days in inventory is an analysis of APCC's flow ratio. I never cease to be amazed at the power of this incredibly simple but powerful tool for analyzing balance sheets (click here for more on the flow ratio; I also recommend Tom Gardner's amazingly prescient analysis last year of Nine West, which fell 43% in the four days after his column). Here is a graph showing APCC's flow ratio since Q4 '94 (remember, lower is better):

APCC's Flow Ratio

You can see that APCC tends to have a high flow ratio -- I generally like less than 1.25, whereas APCC's has never been below 2.0 and has on occasion soared above 4.0. However, given the inventory levels that are part of APCC's business model, a high flow ratio can be expected. It's the changes in the flow ratio that matter. Here is another graph showing both the change in the flow ratio (compared to the previous quarter) and also the change in APCC's share price (relative to the S&P) in the quarter following APCC's earnings release:

Changes in APCC's Flow Ratio Have Forecast Stock Price Movements
(click the image to see a larger version)

*Q2 '99 share price data is through 9/17/99.

While obviously many factors play a role in how a stock performs relative to the market in the period between earnings announcements, it's remarkable how well the flow ratio -- in particular, large movements up or down -- predicted what happened to APCC. For example, look at Q2 and Q3 '95, when the flow ratio worsens sharply (note that on the graph, the flow ratio scale is inverted so that an increase (worsening) of the flow ratio is represented by a downward bar, matching a downward share price). The stock declined 38% and 33% in these quarters.

The same thing happens in reverse in Q2 and Q3 '96, when the flow improves sharply and the stock soars 71% and 60% in these quarters. It turns out that in the four quarters (of the 18 in this data set) when the flow ratio worsened by 0.3 or more, APCC's stock fell an average of 29% (+18%, -41%, -44%, -49%). This method predicted all three of the worst quarters for APCC's stock (the next worst quarter was only -19%). Conversely, in the six quarters when the flow ratio improved by 0.3 or more, APCC's stock rose by an average of 17% (-19%, -14%, -1%, +30%, +49%, +59%; the latter two were APCC's two best quarters).

As an example of why I pay much more attention to the flow ratio than days in inventory, let's look at APCC since Q1 '98. In Q2 '98, the days inventory soared, but the accounts payable rose as well, so the flow ratio didn't change and the stock price continued to rise. It was only two quarters later, when the flow ratio spiked up in the Q4 '98 report, that the stock began to fall sharply (it declined 41% -- 49% relative to the S&P -- from the time the Q4 earnings were announced on 2/4/99 to the release of the Q1 '99 report on 4/22/99). The Q1 '99 report showed that the flow ratio stabilized, and the stock has rebounded 37% since then.

I am very bullish on APCC's future, especially relative to its current valuation, but as you can imagine, the first number I calculate when the company issues an earnings report is the flow ratio.

-- Whitney Tilson

Whitney Tilson appreciates your feedback at [email protected].