My Favorite Recent SEC Filings

This is not a stock-picking or a bear-raiding service. It represents a few stocks that I happen to follow or to look up for some reason or other. If you find any of this amusing, and you run across similar gems in your own process of securities analysis, why not pass them on for inclusion in this list?


I may have a position in one or more stocks listed here. Since I currently am not shorting stocks or buying puts on individual stocks, you may guess which of these stocks I'm likely to be long on.

The Good News

First off, to show that this isn't just some amateur Hall of Shame, here's a list of companies in whose reports I've found something positive.
  • American Power Conversion
  • Cabletron Systems
  • Stryker Corporation
  • The Rest of the News

    Now here are the others.
  • Adobe Systems
  • Autodesk
  • Leggett & Platt
  • Novellus Systems
  • Rubbermaid
  • The Companies

    Adobe Systems (ADBE)
    It's a pity to slam Adobe. Their PostScript® is a genuinely great invention. They got Bill Gates mad; and they lived to tell about it. And besides, I know a trivia question that John Warnock himself wouldn't be able to answer satisfactorily: When Information Systems Design installed its new Univac 1108 in its new offices at Oakland Airport Industrial Park in 1968 and got Exec II to boot after considerable effort, why did the program element that carried the ESI access word vector have the name WARNOCK? (You don't want to know the answer if you're not an 1108 veteran. Or even if you are.)

    But facts is facts. Adobe is in the graphics business, right? So here's a piece of graphics from their 1995 annual report, inside front cover:

    [39K image essential to the text]

    Look at the Net Income Per Share. (And for an embarrassing contrast, take a look at the graph that Stryker Corp. publishes. But back to Adobe.) The $0.62 for '93 is a little less than 3 times the $0.22 for '94. Somehow, it looks about 5 times as big, doesn't it? Well, that's because it is 5 times as big. Something like 1.34 versus 0.26 inches in the original printed version. (You absolutely don't need to use a ruler to see this effect, but a ruler confirms it.)

    We all know how to achieve that effect, which Darrell Huff in How to Lie with Statistics called the gee-whiz graph: you cut off the bottom of the graph so that it doesn't start at zero, and the differences look much more impressive. Yeah, but now look at the '95 figure. At $1.26, it's a little more than twice the '93 figure. Does it look like twice as much? No, it does not. Is it twice as big? No, it is not. It's a little over 1.5 times, by my measurement. Truncating the graph will not make one difference look too big and another look too small!

    But, to the 1/50 inch accuracy which is the best I can do without magnification, the $1.26 for '95 is exactly eight (8) times the size of the $0.22 for '94. The improvement in last year's numbers (the only thing in an annual report that readers really care about, right?) is exaggerated by an impressive 40%.

    Obviously, there is no way of just fudging the scale (in fact, no linear transformation of an honest graph) that can produce something that looks like this. And there's plenty more of this nonsense; look at Net Income for '93 and '94, for instance. The only way to make graphs like these is to hire a graphic artist to draw some nice, aesthetically pleasing lines, without any thought of actually representing the data--except, of course, that the presentation must overstate the company's performance.

    I kinda wonder about this. Back when I was involved with the reporting for a public company, the team from Ernst & Young went through the draft annual report, found all the numbers that appeared in it, and required justification from management for every single one. Well, not the page numbers, I guess, but everything else. Have standards deteriorated? Not bloody likely in these litigious times. Does KPMG Peat Marwick not care what it puts its name to? Or have the compliance people decided that misrepresentation doesn't matter, so long as it's just graphics? You pick one.

    (A year later: I am shocked to discover that Adobe didn't heed my complaints! The 1996 report shows the same scorn for accuracy, or even meaning. Remember the 62 that looked 5 times as big as 22? Well, in this year's report, it looks about 1.5 times as big as 22. I scorn to measure it; and besides, they've unflattened the bars by putting nice little bevels on all sides, so which part of the bar would you measure anyway??)

    Autodesk (ADSK)
    No big abuse here, barring the universally accepted fiction that the stockholders benefit from the company's use of their money to buy back its stock, but the graphics are a case for Professor Tufte. The 1996 annual report has on its inside front cover a set of pie charts, showing net revenues by geographic region for the last three years, and a simple bar graph showing net revenues for the same periods. The bar chart is completely honest, with explicit scale and no phony 3-D stuff to add noise to the message. Full points here. The pie charts, however, are of three different sizes. Since each one represents one year's revenue, their areas represent the relation of the revenues for the 3 years, right? Ha ha. The '95 chart is over twice the size of the '94 chart, and the '96 chart is 4 times, so that the growth of revenues is exaggerated by factors of 1.8 and 2.9, respectively. But you see, those pie charts are located on a roughly elliptical line, so the effect is that they're in orbit, see, and they're like coming closer and closer, see, so the 96 one looks bigger as it comes at you out of the plane of the page. Almost make you wish the rest of the chart hadn't been so honestly planar. Pessimum est corruptio optimi.

    American Power Conversion (APC)
    These folks make uninterruptible power supplies and such. What caught my eye in the 1994 annual report was their quaint idea that it's actually better to have software done by an internal team that knows something about your business and your products than to out-source it to the lowest bidder like a modern company! They even think the practice sets up a competitive barrier to entry (which they need, being market-share leaders). If you're a contrarian, you may want to look into buying the stock. If you know anything about product development, you will want to buy their products for any computer you actually care about.

    Their web site used to have a page of insulting commentary on competitors' marketing claims. Apparently they've gone on a quest for respectability or received a shocked memo from the legal department or something; the page is no longer there. A pity.

    Cabletron Systems (CS)
    From the 1996 proxy statement:
    Executive Compensation
    Name and Principal Position Year Salary($)Bonus($)Options
    S. Robert Levine (CEO)1996 52,200 0 0
    1995 52,000 0 0
    1994 52,000 0 0
    Craig R. Benson (COO) 1996 52,000 0 0
    1995 52,000 0 0
    1994 52,000 0 0
    ...(more officers here)

    And from the board's report on executive compensation:

    Mr. Levine and Mr. Benson currently own 26.4% the Company's outstanding Common Stock. As the Company's principal stockholders, Messrs. Levine and Benson have participated very directly in the increase in value of the Company's Common Stock since its initial public offering. Accordingly, the Board of Directors has determined that their cash compensation may remain low, both absolutely and in relation to their contributions to the Company's growth. Therefore, Messrs. Levine's and Benson's salaries have not increased since the Company's initial public offering. The relatively low salary levels of Messrs. Levine and Benson are not determined by comparison to other companies in the industry or to the Company's corporate performance. Rather, the Board of Directors has determined that it is in the best interest of the Company that Messrs. Levine's and Benson's primary form of compensation for their leadership of the Company be derived from sales and increases in value of their Common Stock.

    Can you believe it? The guys who (I presume) started the company and now own a quarter of it (it = $1.1 billion in mid-July 96) are presumed to make their money from the company's success and not need to compete in the CEO Salary Sweepstakes. The board even has the gall not to make some half-assed comparison with industry standards, because it would be irrelevant.

    By the way, stock option grants to the top 4 executive officers combined were a total of 10,000 shares, which was, hold your breath, 0.6% of total options granted. In a normal company, you're lucky if it's below 20%. (My subjective figure, not audited by a major accounting firm)

    Kind of nostalgic, isn't it? Some people still believe in profiting by owning the company and keeping the workers' interest by making them owners. And, note, they're getting obscenely rich by doing it.

    Leggett & Platt (LEG)
    You can read what I told Joe about this.

    Novellus Systems (NVLS)
    (This is not Novell, the network company.) Novellus makes chemical vapor deposition equipment for the semiconductor industry. Probably quite a sound business; I don't know. I do know that their financial management gets a booby prize almost comparable to the one for Rubbermaid.

    Guess what, they have a stock buyback program. And they have positive cash flow to pay for it; this isn't some idiot (or thief) that's going into debt to do the buybacks. True, the $9M in accounts payable due to the buyback takes a little of the luster off the $14M cash flow, but the cash is still positive.

    No, here's the good part. You're not gonna believe this. In 1994 they sold close to $54M of stock in a public offering; in 1995 theybought back nearly $35M. Taking advantage of a dip in the market is smart--but did you notice how in 1995 the stock market went up? Or do you not even have a satellite dish in your cave? No need for conjecture here; since the statement of shareholders' equity gives the number of shares, we can compute the average sale and purchase price: sold for $35.92 in 1994, bought for $55.74 in 1995.

    It's said that when Robert Schumann first heard the music of Richard Wagner, he wrote about it in a musical magazine, under the headline "Hats off, gentlemen, a genius!" We may echo what he said in the next issue (according to Professor Peter Schickele) when he had heard the music of PDQ Bach:

    Hats back on, gentlemen, an idiot!

    Rubbermaid (RBD)
    Small things first. As of April 16, 1996 they sent out a little first-quarter report to shareholders. Oddly, there's no corresponding filing in the Edgar database, or at least there wasn't as of May 12. Maybe I just don't understand these things. Anyway, their current ratio (current assets to current liabilities) managed to drop from 2.95 in 1995 to 1.63 in 1996. A lot of people see the slightly arbitrary value of 2.0 as the line between good operating capital and something rather less. What does management say about the major drop? Figure it out for yourself; there's no explanation, or mention, of the matter in the accompanying text. You might notice, though, that while the earnings were dropping ($54M to $42M) they increased the dividend ($.125 to $.14 per share; $20.1M to $21.6M total).

    Well, in the annual report for 1995, one quarter earlier, when the current ratio was down, but not yet down to 2, the Management Discussion and Analysis did mention the question of working capital (the difference, rather than the ratio, of current assets and liabilities). One of the items contributing to the drop in working capital was "increased notes payable due to share repurchases". Yes sir, you don't want to accumulate a lot of idle cash, so you pay it back to the stocksellers (stockholders? Who are they?) via share repurchases. And if you don't have the cash, you buy back the stock anyway, with borrowed money.

    Am I blowing it out of proportion? After all, the stock buyback was only one of three reasons they listed for decreased working capital. So let's look at the consolidated statement of cash flows. In 1994, they started serious repurchases, over $48 million worth. This turned a positive cash flow of $13M into a negative of $35M. Not exactly surplus cash.

    But this is nothing compared to 1995. They liked buying back stock so much that they increased it by $85,000,000 to $134,000,000. Where did they get all that cash? Well, their operations generated $18M of new cash, compared to 1994. But investments soaked up $28M more cash than in 1994, leaving them already $10M short of the previous year's cash flow. So, how did they increase the buybacks when they had less cash? That's the little glitch in working capital they mentioned: $89,000,000 in new short-term borrowing! The net, with some other changes that pretty much wash out, is $6M worse cash flow than in 1994, with huge new short-term debt to boot.

    It's real nice of them to put $134,190,000 in the pockets of the stocksellers, but the stockholders might wonder how much longer this pissing away of the value of their company will go on. Take comfort, it's not much longer. At the average cash flow for the past 3 years, they'll have zero (0) cash left in just over two years. At the 1995 rate, they can't make it to this time next year without using their last dollar of cash. This is rather like a population in exponential growth: you know that a very rapid stop is coming up, and you know pretty much when it will be; you just don't know whether the limit will come by means of war or famine or pestilence, or their fiscal equivalents. But you do know whether you intend to be holding the stock when the stop comes.

    What on earth can have been going through management's minds? If you've got a non-libelous explanation, let me know. I don't have one.

    Gee, that's a weak punch line. Here's the real thing, from the annual report: "In December 1995 the Board of Directors increased the authorization for stock repurchase over the next four years by 20 million shares in addition to those already acquired at that time."

    September 5, 1996: I'm as tired of Rubbermaid as you are. Nonetheless, a Web page that isn't kept up to date is an abomination. So here's more news on RBD.

    Stryker Corp. (STRY)
    All we have here is a simple performance graph, which I have displayed mainly to show up Adobe.

    [Here we have a 97K Graphic, which is not essential to the purposes of security analysis.]

    The ability to draw a meaningful chart isn't anything unusual, or shouldn't be. But this is a nice, clean one, and how about the shape of that earnings curve!


    Edward R. Tufte: The Visual Display of Quantitative Information, Graphics Press, Cheshire CT, 1983, in case you want to know how graphics should be presented.

    Date last modified: March 30, 2001
    Dan Drake's Home Page
    Mail to

    Copyright (C) 1997 Daniel Drake. A royalty-free license to reproduce this document in whole or in part is hereby granted provided (i) all additions, omissions, and other changes are clearly marked; (ii) the work is not reproduced as, or as part of, a work for which payment is charged; (iii) this notice is reproduced without change. Quotations for critical or polemical purposes, with proper attribution, are permitted in any case, being obviously fair use.