Top 50 Warranty Providers:
The data is in, and there are four new names on the list of companies with the largest warranty programs. American manufacturers reported more than $23.675 billion in warranty claims during 2003, a rate very close to $2 billion a month.
The year-end warranty claims and accrual reports are in, and we can now be more certain of the size of this OEM warranty industry. Based upon the reports of some 750 U.S.-based manufacturers, over $23.675 billion was spent honoring warranty claims during 2003. Claims, which jumped by almost 5% in the third quarter, grew by only 1.3% in the fourth quarter. For at least nine companies, warranty is now a million-dollar-a-day enterprise.
|In recent weeks, four major warranty providers made first-time reports of their warranty activity: Black & Decker, Eaton, Motorola and Textron. Added in retroactively, these four heavyweight companies spent almost half a billion dollars on warranty claims during 2003. With their addition, it turns out that all through the second half of 2003, American manufacturers spent more than $2 billion a month on warranty.||
There are still a handful of companies who are not scheduled to file their Form 10-K annual reports with the U.S. Securities and Exchange Commission until late March or early April. And there are a few others that have said they will be unable to file on time. But outside of Nortel Networks Corp., Standard Motor Products Inc., and Goodyear Tire & Rubber Co., they are for the most part smaller companies spending much less per quarter on warranty claims than any of the Top 50.
So let the counting begin! For those companies that have yet to report, we'll insert placeholder estimates for their claims totals until the exact information becomes available later in 2004. For the vast majority of manufacturers, however, there is now a comprehensive database available containing 12 to 15 months of exact data about their warranty reserves, claims, and accruals.
We'll begin the mining of this data with a brief look at just the 50 largest warranty providers, as measured by the amount of claims they reported during calendar 2003. Though there was some data available for some companies for some portions of 2002, it was never enough to allow for meaningful analysis of this kind. So what follows in this issue and in subsequent columns in the weeks ahead is likely to be the first time actual multi-company and multi-industry warranty data was ever sliced and diced.
Three Data Points
As an appetizer for what promises to be a lengthy meal, the chart below contains three essential data points for each of the Top 50 companies. The first is the amount of claims they paid (in cash or in kind) during the calendar months of 2003. The second is the proportion that amount represents as a percentage of product sales. The third is a measure of the capacity of the company's warranty reserve fund, expressed as the number of months it would last if accruals ceased and claims remained the same.
The figures for warranty claims paid for the entire year are in most cases arithmetical additions of the claims reported for the first, second, third and fourth calendar quarters of 2003. For companies on a fiscal year schedule ending in March, June, or September, figures were collected for the four calendar quarters of 2003, no matter what the companies called them.
However, several companies are on a fiscal schedule where their years end on off months such as Oct. 31 or Jan. 31. In those cases, the quarter ending in October -- a month in the fourth calendar quarter -- would be counted as their fourth quarter, while one with a quarter ending in January would be counted as a first-quarter report. There are exceptions -- for instance a few companies whose fiscal years ended on the weekend after New Year's. Yes, it's confusing. But in general, whatever Form 10-Q or 10-K had been (or should have been) filed by March 19, 2004 was assumed to represent the company's financial report for the fourth calendar quarter. In previous reports, we've used May 31, Aug. 31, and Nov. 30 as the cutoff dates for the first, second, and third quarters.
The figures for claims as a percentage of sales are a bit more open to interpretation. First, eight of the Top 50 companies report little, if any, revenue data that would be helpful to somebody trying to subtract out service revenue, finance revenue, or non-warranted product revenue from the overall total. In those eight cases -- American Standard, Brunswick, Dell, Gateway, Ingersoll-Rand, Maytag, Nortel, and Whirlpool -- estimates based upon the reports of competitors were constructed for service and/or finance revenue, and were subtracted from the totals.
Second, product revenue figures for the other 42 companies in many cases probably still include some non-warranted products, some service revenue, some finance revenue. Readers are invited to help Warranty Week remove these components from future calculations. However, readers would probably prefer to help with the calculations of their competitors, since any revenue removals would necessarily increase the percentages and bump the company a few notches up the list.
Capacity to Pay Claims
The measure of warranty reserve capacity is a comparison of the amount currently spent on claims per month, divided by the amount most recently left in the reserve fund. If warranty reserves were water tanks, this would be an expression of how long the water would last given the current rate of outflow, and no further inflows. As such, it is just a measure of capacity, which allows small and large companies to be compared.
Also, we should note that while the claims totals are for the entire year, both the claims rate and reserve capacity measurements are snapshots taken as close to the end of the year as possible. Therefore, they can be compared with snapshots of rates and ratios taken earlier in the year. However, several top warranty providers made first-time warranty accounting reports in March 2004. So there is no earlier data, and no way to know if these rates were increasing or not. For all Top 50 companies, claims and accrual rates were highest in the fourth quarter (2.5%) and lowest in the third quarter (2.3%). Reserve capacity was highest in the first quarter (18 warranty-months) and lowest in the third quarter (16 warranty-months).
The quarterly warranty claims totals among these Top 50 companies ranged all the way from $2 million (Dana 4Q03) to $1.2 billion (GM 4Q03). But it takes serious money to play at this table. Of the 200 different claims totals (50 x 4 quarters), only five were below $10 million. Only four (all by GM) were above $1 billion. But thirty quarterly claims totals from ten different companies were between $100 million and $1 billion per quarter. So there are perhaps a dozen companies for whom warranty is a million-dollar-a-day operation.
The annual claims figures for the Top 50 ranged from $57 million to $4.4 billion. Only three companies reported in excess of $1 billion in claims for the full year (GM, HP, and Ford), and 26 more reported between $100 million and $1 billion in claims. Claims rates for the 50 companies varied from 0.2% to 9.6%. Reserve fund capacity ranged from 3.2 warranty-months to 137 warranty-months.
Top 50 U.S.-Based Warranty Spenders
Full Calendar Year, 2003
|Claims||Percent of||Months in|
|Lexmark International Inc.||$168m||9.6%||11|
|Sun Microsystems Inc.||$373m||4.9%||8.9|
|Applied Materials Inc.||$203m||4.5%||5.4|
|General Electric Co.||$749m||4.1%||23|
|General Motors Corp.||$4,435m||2.8%||24|
|Deere & Co.||$334m||2.8%||15|
|Ford Motor Co.||$3,524m||2.5%||18|
|Rockwell Collins Inc.||$57m||2.3%||34|
|Nortel Networks (1)||$169m||2.3%||15|
|General Dynamics Corp.||$60m||2.0%||38|
|Black & Decker Corp.||$86m||1.9%||5.7|
|Cisco Systems Inc.||$314m||1.8%||8.6|
|York International Corp.||$66m||1.8%||16|
|United Technologies Corp.||$405m||1.8%||30|
|American Standard Cos.||$124m||1.5%||32|
|Agilent Technologies Inc.||$76m||1.5%||11|
|Apple Computer Inc.||$74m||1.2%||11|
|Ingersoll-Rand Co. Ltd.||$82m||0.9%||26|
|Lucent Technologies Inc.||$162m||0.9%||62|
|Pulte Homes Inc.||$74m||0.8%||9.3|
|Tyco International Ltd.||$165m||0.5%||30|
|Johnson Controls Inc.||$57m||0.2%||21|
|Subtotal Top 50||$19,539m||2.5%||17|
|Total All Mfgs.||$23,590m||1.9%||17|
Source: SEC Form 10-Q & Form 10-K
Note: (1) -- Company is unable to file its 10-K, so estimates were created using 9-month data.
Steady readers will note some new faces on the list. General Electric has decided to report annually upon its warranty accounting, even though the rules require both quarterly and annual reports. But we're in luck. This was the period in which GE made its annual report. The company reported $749 million in claims for 2003, and $751 million in accruals. It also added $131 million to its reserve fund, which was by far the year's largest upwards change of estimate.
First-time-ever warranty reports were filed in early 2004 by Black & Decker, Eaton, Motorola, and Textron. These four companies reported a cumulative $497 million in warranty claims for calendar 2003, so obviously some revisions will be necessary to past editions of these Warranty Week quarterly reports. We will begin that process next week in a quarter-by-quarter breakout of the claims and accrual totals for all manufacturers.
What's a "Normal" Rate?
Inevitably, the question arises why certain companies are at the top of the list, or what should be considered the "normal" range. That depends quite a bit on the industry in question. For instance, readers will note that many of the largest computer companies are towards the top of the list, while many of the air conditioning and aerospace companies are towards the bottom. This has to do with the type of products sold, the length of warranties they typically provide, cost of repair, number of parts, and many other factors.
It makes more sense to speak of normal ranges. For claims rates, anything between 1% and 5% is in the normal range, although there are exceptions. For instance, in the printer industry, it is common practice to sell the initial hardware at a discount, with lost profits to be made up by later sales of paper and ink. This has the effect of driving up warranty claims rates, which are based only upon the initial hardware price, and which don't include later sales of consumables such as paper or ink.
Conversely, numerous companies sell product to the U.S. military. Sometimes it's warranted but most times it isn't. A fighter jet might be covered by maintenance agreements and performance guarantees, but not by product warranties. However, a GPS handheld might be covered by warranty, especially if it is just a hardened version of a commercially-available product type. Even in the strictly civilian market, some products might be routinely sold without warranty coverage. The point is, some of the apparently low claims rates on the list above may have more to do with a lack of warranties, and therefore a lack of claims. There is no simple correlation between these percentages and any notion of product quality.
In terms of reserve fund capacity, again there is no normal. The median is 17 months, which means half the companies are above and half the companies are below that level. The typical "normal" range seems to run from 10 to 20 months, but again that depends upon the length of warranties, the type of product, and numerous other factors. The higher the number, the higher the capacity, and the more conservative the company's planning assumptions. The lower the number (and some companies outside the Top 50 go as low as half a month), the lower the capacity and the more optimistic the company's planners about the future.
Companies way above this range might have recently seen a sudden and unsustainable drop in claims, perhaps due to seasonal factors or just out of sheer luck. For instance, Dana Corp., after seven quarters of claims in a range of $18 to $20 million, suddenly reports just $2 million in claims. That is not sustainable, but it temporarily pushed the company's capacity up to 137 warranty-months by year's end. Previously in 2003, the company's reserve fund capacity fluctuated from 11 to 21 warranty-months.
Companies way below this range have either recently removed significant sums from their reserves, as did AutoZone Inc., or they have seen a sudden spike in claims, as did Applied Materials Inc. and Maxtor Corp. However, there are a few companies that have chosen to remain outside the "normal" range for long periods of time. If they have left themselves short, they will take a big hit to earnings if claims should one day spike. If they have over-accrued, one day they can treat their shareholders to an unexpected one-time profit boost.
Finally, there are the totals at the bottom of the chart. These 50 companies accounted for roughly 83% of the total claims reported by all manufacturers during 2003, although numerous importers are not reflected in that total. As detailed in the chart below, the Top 50's claims totals for the four quarters of 2003 were $4.719 billion; $4.775 billion; $5.008 billion; and $5.037 billion.
Top 50 U.S.-Based Warranty Spenders
Claims and Accrual Rates
Full Calendar Year, 2003
Source: Warranty Week
Two questions immediately pop out from this data. First, why the dip and then upturn in both claims and accruals over the course of 2003? Second, why are the claims and accrual rates so much higher for these 50 companies than they are for all manufacturers in general?
More detailed answers will hopefully come in the weeks ahead as we more closely examine the individual company data. Right now, let's just say that the fluctuations in rates have more to do with rising sales than with changes in warranty claims and accruals. However, as the chart makes clear, the warranty claims totals (in blue) expanded each quarter.
As to why the rates are generally higher for the larger companies, this deserves some further investigation. Could it be that the larger companies somehow encourage claims? Could it be that the Top 50 companies tend to be the OEMs closest to the customer while the smaller companies tend to be suppliers located a few notches further up the food chain? Or perhaps the larger companies keep better track of their warranty accounts? For now, we'll just note that the difference exists, and ask readers for the possible reasons.
|Go to Part Two|