Top 50 Warranty Providers:
While 20 saw declines in the percentage of revenue spent on warranty claims, all but 12 paid out more claims in dollars during the first half of 2005. Rising product sales made the difference.
Warranty claims rates continue to drop for a significant number of major manufacturers, as measured by rates seen at the end of June 2005 and compared to a year's before.
According to data filed with the U.S. Securities and Exchange Commission and compiled and analyzed by Warranty Week, 20 of the top 50 warranty providers saw claims rates fall during the first half of 2005. However, 30 of the top 50 actually saw claims rates rise -- some by significant amounts.
Last week we charted the claims and accrual rates of the top ten warranty providers. Next week we'll detail some of the industry-by-industry trends in terms of average claims and accrual rates and average reserve levels. This week, we'll provide in one great big table the dollar amounts of claims paid and the percentage of product revenue that represents, along with comparisons to the middle of 2004.
In the chart below, 50 companies were chosen based upon the dollar amounts they reported having been spent to satisfy warranty claims (in cash or in kind) during the first and second calendar quarters of 2005. They were then ranked by the change in their claims rates. For companies on a fiscal year schedule not corresponding to the calendar, data from the quarters that ended in 1) Jan., Feb. or March and 2) April, May or June were used.
Eight Missing Companies
For one reason or another, eight companies that could have or should have been included in the top 50 were omitted. For Black & Decker, Eaton, General Electric, Motorola, and Textron, the problem is that they have chosen to overlook the quarterly disclosure requirement of FASB Inquiry No. 45, and have instead chosen to provide the required information only once a year. For Sun Microsystems and Western Digital, the problem is that their fiscal years ended on June 30, 2005, so they aren't scheduled to file their annual reports until the middle of September. For Terex, the proximate problem is a lack of any filings for the past four quarters, which has in turn led to an SEC investigation into the company's accounting practices and its ties to United Rentals.
Were it not for these special situations, all eight of these companies probably would have been included in the top 50. We could have made estimates of each company's warranty spend and gone back later to replace it online with actual data. However, while that may be fine for annual reports where typically only a few companies are a few weeks late, in this case the real data for at least five of the eight won't be available until March 2006. Therefore, just for this comparison of the most recent half-years, we have chosen to leave them out.
Top 50 U.S.-based Warranty Providers
First Half, 2005
Improvement in Claims Rates vs. 2004
(in $ millions and percent)
|Company||6 mo. '05||Prod Sales||2004|
|Deere & Co.||$186m||1.5%||-0.7%|
|Nortel Networks Corp.||$129m||3.0%||-0.6%|
|Rockwell Collins Inc.||$29m||1.8%||-0.4%|
|Tyco International Ltd.||$62m||0.3%||-0.1%|
|Mohawk Industries Inc.||$24m||1.0%||-0.1%|
|Pulte Homes Inc.||$60m||1.0%||+0.05%|
|Cisco Systems Inc.||$187m||1.9%||+0.1%|
|Beckman Coulter Inc.||$32m||3.7%||+0.1%|
|Eastman Kodak Co.||$37m||1.4%||+0.1%|
|Apple Computer Inc.||$93m||1.7%||+0.1%|
|Thor Industries Inc.||$26m||1.8%||+0.1%|
|Ford Motor Co.||$1,965m||2.5%||+0.2%|
|General Motors Corp.||$2,366m||3.0%||+0.3%|
|Standard Motor Products||$24m||5.8%||+0.8%|
|Applied Materials Inc.||$91m||2.6%||+1.0%|
|Novellus Systems Inc.||$40m||6.1%||+1.1%|
Source: Warranty Week from SEC data
(1) Five companies -- Black & Decker, Eaton, General Electric,
Motorola, and Textron -- have chosen to detail their warranty
expenses only once a year, and are therefore excluded from this chart.
(2) Sun Microsystems and Western Digital are not scheduled to
file their annual reports until September, and so are excluded from this chart.
(3) Terex has been unable to file either a quarterly or an annual
report since the second quarter of 2004, and is therefore excluded from this chart.
The fact that these eight companies were not included here allowed eight other manufacturers to take their place: Avaya; Champion Enterprises; Masco; Mohawk Industries; NCR; Standard Motor Products; Storage Technology; and Thor Industries Inc. All reported claims amounts between $23 million and $26 million during the first half of the year. All eight of the excluded companies are likely to eventually report more than $26 million in claims for the first six months of 2005. When all the data becomes available in six months, we'll amend the online edition of this article to include their actual rankings.
Compared to last year, in absolute numbers, only 12 of the 50 manufacturers paid out less claims in dollars during the first half of 2005 than they did during the first half of 2004. They were: Agilent Technologies; ArvinMeritor; Boeing; Champion Enterprises; Honeywell International; Ingersoll-Rand; Lexmark International; Maxtor; Mohawk Industries; Nortel Networks; Seagate Technology; and Tyco International Ltd. For all but Agilent, this translated into a corresponding drop in claims rates. But because Agilent saw product sales fall a slight bit faster than claims, its claims rate went up fractionally.
All the data presented here relates to the worldwide warranty spending of U.S.-based manufacturers. As far as we know, none segment their warranty data into U.S. and international components, and none do very much to apportion their aggregate warranty data by division or product line. In addition, none provide any details about total outlays minus reimbursements. So all these figures are for companies as a whole, and all these figures represent net claims paid. And despite the inclusion of Nortel, Seagate, and Ingersoll-Rand, companies which are foreign-based and which therefore are not required to file Form 10-Q or 10-K statements with the SEC are not included.
Estimates for Service & Finance Revenue
Both the claims rate and the accrual rate are computed using the best information available for product sales, as opposed to total revenue (which might include finance, service, or consulting revenue, for instance). When a company does not segment its product and non-product revenue, or when it provides no segment data at all, Warranty Week estimates the size of the warranted product segment using data from close competitors. For instance, Dell provides no product line data but Apple does, so we applied Apple's warranted hardware segment ratio to Dell's total revenue number. For Whirlpool, Maytag, and a few others, we simply assumed that 10% of total revenue originated from non-warranted sources such as finance, service, and extended warranty premiums. All of these estimates were held constant from one year to the next.
Of the 38 manufacturers who laid out more in dollar terms to pay claims in the first half of 2005, only nine saw their claims rate fall in percentage. For the others, claims rose faster than sales, or sales fell, so the size of the percentage fraction grew. For instance, IBM paid out $400 million in claims during the first half of 2004 and $409 million during the first half of 2005. Its hardware sales fell from $14.2 billion to $12.3 billion over the same period, and so its claims rate grew from 2.8% to 3.3%. Cummins Inc. saw sales rise by 21%, but claims rose by 67%, so the claims rate rose from 2.1% to 3.0% -- a 39% increase in the percentage.
Big Changes in Claims Rates
When it's viewed on a percentage basis, the change in claims rate reveals some shocking year-to-year changes. For instance, at Boeing the claims rate fell by 0.6% to 0.6%. But that represents a 50% decline! Seagate Technology, Deere, and Ingersoll-Rand each saw their claims rates plunge by 30% or more. Tyco, Maxtor, Avaya, Lennar, and Rockwell Collins saw declines of 20% or more.
Conversely, there were some disturbing increases in claims rates, again as measured by the percentage increase rather than merely the mathematical difference in rates. Masco saw its claims rate soar by 154%, while Delphi saw its rise by 73%. There were 17 companies that saw a rise of 10% or more, while these two plus Fleetwood Enterprises, Palm, Cummins, and Applied Materials were the six who saw their claims rates rise by 30% or more.
Of course, there are numerous reasons for the massive increases and decreases. Most we would dismiss to timing issues, either in favor or against the manufacturer. For instance, companies who are primarily suppliers to OEMs could find that in one period they receive an early credit from their OEM while in another period the credit is late. Because we're measuring only six months in each case, small changes in timing can produce large changes in rates.
Likewise, for some companies there have been acquisitions or divestitures, or major changes in the product line that affect both sales volumes and warranty payments. Some companies may actually still be paying claims for products they no longer manufacture. That would drive sales volumes down and claims rates up.
Can't Compare One Company to Another
Still, these are the numbers, and just because there could be 50 detailed footnotes for 50 "special" situations doesn't mean the data shouldn't be presented at all. Here is where we insert our standard disclaimer that one company's percentages may not be directly comparable to another's. However, we believe it is statistically valid to compare one company to itself over time, even if different companies include different cost items under the warranty heading than others.
Then again, it's still astonishing that some 30 months after manufacturers began providing detailed warranty data in their Form 10-Q and 10-K statements, very few stock market analysts are actually tracking the numbers. They could probably tell you sales volume per cubic meter of factory space out to three significant digits, but they don't have the slightest clue about such a leading indicator of quality and reliability as warranty.
|Back to Part One||Go to Part Three|