The Warranty Reserve Fund:
Given the availability of six months of data detailing changes in the warranty reserves of more than 600 U.S. manufacturers, what trends are becoming visible so far?
When examining the warranty tables that are now part of most manufacturers' Form 10-Q and Form 10-K financial statements, it becomes clear that while most companies are managing their warranty accruals and warranty reserve funds rationally, some are more or less making it up as they go along.
Because the reserve fund and the accruals are subject to estimates of future liabilities, one suspects that a chief financial officer can use these items to manage profitability. In good times, they can put a little extra away, and in bad times, they can take a little extra out, thus flattening and evening out the quarterly figures for net income.
One suspects that there are other factors at work here, but proving it with just the numbers is difficult. After all, how can a prediction be said to be wrong until the future becomes the present? But in at least 11 cases, warranty accruals (the amount added to cover new product warranties issued) were exactly the same as warranty claims (the amount spent in payments or in kind to cover claims) during the second quarter. If the accruals are truly estimated as a percent of current product sales and if the claims are more or less a function of past sales, the chances of that happening are about the same as a flipped nickel landing on its edge. But in four of those cases, the companies did the same thing in the first quarter.
A New Era
This is all new. Last year, warranty disclosures were not required, but many companies listed the reserve fund among their accrued liabilities. Some even detailed the ups or downs in the cash flow statement. Many included some boilerplate phrase in their financial statements outlining how they made their estimates and how they were periodically adjusted. Now thanks to Interpretation No. 45 from the Financial Accounting Standards Board, we are two quarters into the era of mandated disclosures regarding guarantees made, including warranties issued, and it has become possible to see not only where manufacturers stand, but in what direction they are moving.
Over the past month, Warranty Week has compiled a list of all U.S.-based manufacturers who are making the required warranty disclosures. In the past four columns, we have focused primarily on warranty claims, calculating them as a percentage of product sales and looking for industry averages by grouping companies together. This time, the focus is on the warranty reserve, calculated not as an absolute number but as a ratio: between the amount of claims paid during the second quarter and the reserve fund balance at the end of the quarter.
The list started with 735 companies for whom quarterly warranty data was gathered from their 10-Q and 10-K statements. A total of 80 companies had to be removed from this warranty reserve analysis because they did not provide full accounting of both their warranty claims and accruals. Most provided only the beginning and ending balance in their warranty reserve fund, or the net change in the reserve fund balance. Large and medium-sized companies in this group included Applera Corp.; Emerson Electric Co.; Ethan Allen Interiors Inc.; General Electric Co.; Harman International Industries Inc.; K&F Industries Inc.; Lufkin Industries Inc.; Mentor Corp.; Pinnacle Systems Inc.; Sauer-Danfoss Inc.; and Select Comfort Corp.
An additional 15 companies were excluded from the list for their inability to file financial statements at all for the second quarter by the close of business on Sept. 30. These companies include Advance Stores Co. Inc.; Asante Technologies Inc.; Astrocom Corp.; Baldwin Technology Co. Inc.; Calprop Corp.; DT Industries Inc.; Energy Conversion Devices Inc.; Glas-Aire Industries Group Ltd.; II-VI Inc.; Integrated Security Systems Inc.; Lancer Corp.; Read-Rite Corp.; SmartDisk Corp.; SureBeam Corp.; and Vina Technologies Inc. Each had filed a financial statement covering the first quarter, but because this is an analysis of the change between quarters, they had to be excluded.
New Arrivals
Five companies, including AMF Bowling Worldwide Inc.; Dynatronics Corp; Interwave Communications International Ltd.; Trio-Tech International; and Urologix Inc.; did in fact file their Form 10-K statements in late September, and thus their data is now included in the total. As mentioned in a previous column, among larger companies, Cisco Systems, Sun Microsystems, and Western Digital also filed their annual reports with the SEC late in September. All the online columns first published in the Sept. 8 to 29 time period have now been revised to include their warranty accounting data, replacing the respective placeholder estimates with actual numbers. All those columns in their revised form remain online at http://www.warrantyweek.com/newsarchive/.
Finally, 26 companies were knocked off the list because they reported exactly $0 in warranty claims. Lucky for them, but the computation of the claims-to-reserves ratio doesn't work if the reserves ratio is divided by zero. Companies in this group included Agere Systems Inc.; American Building Control Inc.; American Technology Corp.; Cabot Microelectronics Corp.; Computer Access Technology Corp.; Energy Conversion Devices Inc.; Evergreen Solar Inc.; GolfGear International Inc.; HEI Inc.; Hubbell Inc.; II-VI Inc.; Illumina Inc.; Interwave Communications International Ltd.; Key Components LLC; Lantronix Inc.; Loral Space & Communications Ltd.; Microsoft Corp.; Nanometrics Inc.; ON Semiconductor Corp.; Symphonix Devices Inc.; Taser International Inc.; Tektronix Inc.; Turnstone Systems Inc.; Vyyo Inc.; White Electronic Designs Corp.; and Zarlink Semiconductor Inc.
That leaves 614 companies for whom complete FASB FIN 45-compliant warranty tables were published, and for whom claims totals were greater than zero dollars. This is the list that will be used for an analysis of the claims-to-reserves ratio.
Industry Averages
The overall ratio of the warranty reserves to warranty claims for 614 companies was 17.1 warranty-months as of June 30, meaning that on average, these companies had sufficient reserves on hand as of June 30 to pay the next 17.1 months of warranty claims if their claims rates continued at second quarter rates. This represented a slight decrease from the 17.3 warranty-months seen for these same companies as of March 31. But remember that this is a weighted average, so the effect of a slight drop in the ratio of a major warranty provider such as Ford, HP, IBM, or United Technologies is magnified tremendously.
A better measure is the median, which reflects the middle point of the estimates of 614 different chief financial officers as to the liability they face in future claims, and the amount they have set aside to pay those claims. This time around, the median was 14.3 warranty-months, as opposed to the 14.2 warranty-months found in the first quarter, as detailed in the July 7 issue of Warranty Week. What this means is that 307 CFOs felt the need for more than 14.3 warranty-months of coverage in their reserve funds, while 307 CFOs felt the need for less. The CFO of Printronix Inc., meanwhile, had the median outlook: $1.413 million in the reserve fund and $99,000 per month in claims.
When we last looked at the warranty reserve balances and found 14.6 warranty-months to be the median, one of the most frequent comments from CFOs was that higher or lower amounts of reserves were typical and customary in their respective industries. Indeed, we continue to find that in certain industries, higher reserves are kept for whatever reason -- be it fear of a product recall, very long warranty periods, or a combination of factors. In the automotive industry, for instance, the claims-to-reserve ratio was 20.9 warranty-months in the first quarter, and 21.0 warranty-months in the second quarter.
Likewise, in certain industries, one-year or even 90-day warranties are customary, and the items covered are either so inexpensive or so "disposable" in nature that claims are unlikely. For instance, even if a $10 watch has a one-year warranty, how many consumers are likely to exercise their rights if it breaks? High levels of reserves therefore are inappropriate for such a manufacturer. In the software industry, most publishers don't even maintain a warranty reserve fund, because they choose to immediately expense the small amount of defective product returns, shipping out a new CD in the mail.
Here is a list of what we've found to be the weighted averages in specific industries. Keep in mind that the median for all manufacturers is 14.3 warranty-months, and the weighted average is 17.1 warranty-months. Also keep in mind that very large companies such as General Motors, Caterpillar, and United Technologies are included in more than one industry in the chart below, so the columns will not add to equal the total.
The Warranty Reserve Fund
Industry Averages, Claims vs. Reserves
Second Quarter, 2003
| Warranty | Monthly | Claims to | |
| Reserves | Warranty | Reserves | |
| Industry | June 30 | Claims | Ratio |
| Aerospace | $13.2 b | $525 m | 25 |
| HVAC/R | $2.7 b | $119 m | 23 |
| Automotive | $18.3 b | $873 m | 21 |
| Elec. Power Equip. | $2.8 b | $133 m | 21 |
| Telecom Equip. | $1.5 b | $83 m | 18 |
| Farm Equip. | $1.3 b | $80 m | 17 |
| Data Storage | $0.7 b | $47 m | 15 |
| Mining Equip. | $0.2 b | $18 m | 13 |
| Medical Equip. | $0.5 b | $41 m | 13 |
| Semiconductor | $0.8 b | $64 m | 12 |
| Computer | $4.5 b | $416 m | 11 |
| Weighted Avg. | $32.7 b | $1.85 b | 17.7 |
| Median | 14.3 |
Source: SEC Form 10-Q & Form 10-K
What this chart shows is that there's no single benchmark for all manufacturing companies. There are definite differences between industries. There are even noticeable differences within specific industries, as the last four columns of Warranty Week have detailed.. Each CFO in each industry is using the best information they have to assess the adequacy of their warranty reserves and to make sure enough funds are accrued each quarter to pay for new warranties issued. If they discover that the amount expensed in past quarters was too high or too low, they can make a change of estimate that adds or removes funds from the reserve. If their company experiences a manufacturing crisis and their reserves are insufficient to cover the sudden spike in claims, they will suffer the consequences. But if their company experiences a noticeable boost in product quality, they will enjoy the added profits that result from a downward change in estimate.
Across all industries, there were 285 decreases, 306 increases, and 23 companies for whom the claims-to-reserves ratio remained exactly the same. As for the latter group, this lack of change was entirely due to their newness to the list. During the second quarter, they filed first-time warranty data covering the first six months of the year. Because that data was then divided into two equal parts and added into both the first and second quarter lists, of course it shows no change in the ratio. It was the same data.
Rather than comparing one company against another, let's compare one company against itself. Let's assume that the great majority of companies have tried to right-size their warranty reserves given their industry and their circumstances. Let's assume they periodically assess the adequacy of their reserves and accruals, as they all claim to do in the warranty sections of their financial statements. That would mean that those above the median for all manufacturers or for their specific industry would try to move lower. Or maybe they would remain the same, comfortable in their estimates. They certainly wouldn't move higher, would they?
The surprising result is that a handful of companies that were noticeably above or below the median in the first quarter are even further away from it in the second quarter. In most of these instances, the shifts seem to be beyond the control of the people deciding how much to add or subtract from the reserve fund. But in a tiny percentage of cases, the companies do not seem to be behaving rationally. Either they had above average reserve balances and made them larger, or they had below average reserves and made them smaller.
The Big Get Bigger
Of the 306 increases, 175 were above the median last quarter, and went even higher this quarter. This means that either they 1) made an upwards change in estimate that added funds to the reserve, 2) experienced a fall in the second quarter claims rate, or 3) accrued more funds than was needed to maintain a constant claims-to-reserves ratio.
Seventeen of the above-the-median companies made upwards changes in estimates, while 22 made downwards changes in estimate, and the rest made no changes. The vast majority of the 175 saw claims rates fall, while only 27 saw rates rise. Thirty-six accrued more in the second quarter than they did in the first quarter.
Five companies made both upwards changes in estimate and accrued more than in previous quarters, despite a fall in their claims rates. Wireless chip maker Endwave Corp. ended June with $5.8 million in its warranty reserve, despite having registered claims of $49,000 in the first quarter and only $6,000 in the second quarter. Its claims rate, which was at 0.6% in the first quarter, fell to 0.07% in the second. Its claims-to-reserve ratio jumped from 347 warranty-months to 2899 warranty-months -- from very high to incredibly high.
The Small Get Smaller
Of the 285 decreases, 174 were below the median last quarter, and went even lower this quarter. This means that either they 1) took money out of the reserve through a change in estimate, 2) experienced a rise in the second quarter claims rate, or 3) accrued less funds than was needed to maintain a constant claims-to-reserves ratio.
Nine of the 174 made downwards changes in estimate, while 13 made upwards changes in estimate, and 152 made no change in estimate. The vast majority of these companies saw claims rates rise, while only 37 saw claims rates fall. Thirty-three accrued less in the most recent quarter than they did in the first quarter, while the rest accrued more.
Two companies, Navistar International and Credence Systems Corp., did all three. They took money out of their reserves through a change in estimate, their claims rates rose, and they accrued less than they did in the past. In addition, Lam Research accrued less and made a downward change in estimate, but its claims rate remained constant at 2.4% of product sales. Most notably, Ceradyne Inc., for the second quarter in a row, made no warranty accruals at all despite having claims in both quarters. Through a $148,000 change in estimate, the security company effectively emptied its reserve fund as of June 30.
Guessing the Cause
While it's dangerous to make generalizations with such a diverse group of manufacturers, and especially when more than one factor can change the claims-to-reserve ratio, some trends are becoming clear, even after just two quarterly reports. First, the changes in ratios are primarily related to changes in the percent of product sales paid for warranty claims, rather than to any overt manipulations by the companies in their warranty reserve fund balance. In other words, the CFO's actions aren't the cause; he or she has yet to react to the cause. The primary cause is a change in the claims rate, which is itself affected by both changing product sales volumes and rises or falls in actual warranty claims. With a few more quarters of data, it will be easier to tell if these changes are seasonal, atypical, or part of a trend.
However, in a handful of cases, companies that already were carrying more than is typical in their reserves decided to do something overt that resulted in them carrying even more, or companies that were carrying less than is typical decided to carry less. These are the companies that moved even further from the median. Companies in the first group are deferring recognition of even more revenue that should be freed up and declared as net income. Companies in the second group are deciding to live even closer to the edge than before.
The warranty reserve fund, claims and accruals, and warranty metrics in general will be the theme of the second monthly Warranty Week Webinar, scheduled for 2 PM on Tuesday, Oct. 14. Registration is free, and participation is open to all warranty professionals. Warranty Week editor Eric Arnum will discuss the trends visible in the first and second quarter reports, while Hewlett-Packard warranty program manager Tom Washburn will detail what is and isn't typically included in these quarterly numbers.
Thanks to the disclosure requirements of FASB Interpretation No. 45, all of this information is now in the public domain. One used to have to take companies at their word that reserves were adequate and that estimates were adjusted to match performance. Now, it's possible to fact-check these boilerplate statements, and to find the companies that not only are outside the norm, but are moving further away from the center.
| Back to Part Four | Go to Part Six |
