The makers of major appliances for the kitchen, laundry and bathroom and the makers of heating, ventilation, and air conditioning systems may seem similar, but they're very different in terms of their warranty metrics.

We found 39 U.S.-based companies that manufacture appliances or components for those units, and we found a like number of companies making HVAC systems or parts for them. And while with several metrics the split between them was close to 50/50, we found that the appliance makers have higher average claims and accrual rates, the HVAC companies have higher warranty reserve balances, and the HVAC units they sell generally have longer warranties than the major appliances do.

In past years, when we make these comparisons, we've been worried what to do about large conglomerates such as United Technologies Corp., whose Carrier division is a top HVAC company but whose other divisions are primarily aerospace and building products companies. If we include them in the HVAC cohort, their sheer size will skew the totals and averages of the others. But if we exclude them, the size of the market will look much smaller than it really is.

Partial HVAC Companies

In last year's appliance and HVAC warranty report, we set United Technologies off by itself. In previous years, we just blended them into the totals and averages. But it was always a concern, especially as some of their top competitors were acquired by parent companies from outside the industry.

So this year, we came up with a novel approach: If United Technologies is the "ringer" in the HVAC category -- introducing a large amount of non-HVAC expenses into the HVAC charts -- why not find ourselves a "ringer" in the major appliance category to counterbalance it? And the company that fits the bill is General Electric Co., a top appliance brand but also a major producer (and a major warranty provider) of turbines, jet engines, nuclear power plants, medical equipment, and other decidedly non-appliance-like products.

We always knew there was some aerospace and elevator warranty expense included on the HVAC side, because we included UTC. Now, with the addition of GE, there is some aerospace, power plant and medical equipment warranty expense included on the appliance side. Hopefully, each side is comparably inflated in size, so that one balances the other.

So in Figure 1, we have claims totals that do not agree with those of past years. Instead, they begin a new series of measurements of all companies engaged in either HVAC or major appliance production, including the parents of both Carrier and GE Appliances.

Now, with the addition of GE's heft (it's been a top 10 warranty provider every year we've measured) there's much more balance between the HVAC and appliance categories. Last year, the 39 HVAC companies paid out $1.65 billion in warranty claims, while the 39 major appliance companies paid out about $1.32 billion. That's a split of about 56/44, which is exactly where the ratio has been since 2008. Before that, it was closer to 50/50, though in 2005 the appliance makers actually paid out slightly more than their HVAC cousins.

In 2013, both the HVAC and the appliance categories saw claims rise by about four percent from 2012 levels, after both saw large contractions happen in 2011-2012. The peak year for appliance claims as well as for combined HVAC/appliance claims was 2008, though the HVAC manufacturers paid out a bit more in 2007. Either way, it's clear that they're both still far below their pre-recession peaks.

Accrual Trends

The peaks were a little different with the second warranty metric: accruals made. These are the amounts set aside at the time a product is sold, to finance its predicted future warranty costs. So it's much more closely tied to sales, and therefore it should react more quickly to the kind of sales declines we saw in 2008 and 2009.

However, in Figure 2 we can see that while the peak for the appliance makers came in 2008, the peak for the HVAC companies was in 2006. Also, while neither category is quite back up to their peak pre-recession levels, they're not that far below them either. In fact, it would be very hard to spot a recession at all in the data of either Figure 1 or 2, compared to the data we've seen for hard-hit industry sectors such as passenger car and new home sales.

This is a bit of a surprise, because both HVAC and appliance sales are closely tied to the sale of new homes. However, there's also a hefty replacement market among existing homes, and obviously some people continued to remodel and renovate their kitchens and heating/cooling systems in spite of the economic downturn. In fact, maybe they did so because of the downturn -- they couldn't afford to move, so they bought a new washer/dryer set as a consolation. It's the same effect we saw in the auto market, where people held onto their clunkers a little longer, which drove up demand for parts and repair services.

Accruals were up 8% for the appliance makers in 2013, and were up 5% for the HVAC companies. And as with claims, both categories saw accruals fall by significant amounts in 2011-2012. Those declines were fueled primarily by massively reduced accruals made by GE and UTC in 2012.

However, some of the others also cut their accruals by large amounts. For instance, Ingersoll-Rand reduced its accruals from $210 million in 2011 to $150 million in 2012. They increased last year, but only to $154 million.

Meanwhile, a relative newcomer to the list -- Green Mountain Coffee Roasters Inc. -- saw its accruals fall from $43 million in 2011 to $33 million in 2012 and only $20 million in 2013. Likewise, the percentage of sales that were consumed by warranty expenses peaked in late 2011 and early 2012. So whatever manufacturing problems the company was having with its Keurig coffee machines must have been resolved at that time.

Expense Rates Based on Sales Data

The overall trend in claims and accruals taken as a percentage of sales is hard to read. That may in fact be the price we pay for including GE and UTC -- their claims and accruals overwhelm the others. But neither company exhibits any seasonal pattern in their warranty expense data. So why is one so clearly evident in Figures 3 and 4?

In Figure 3, we've taken the claims and accrual payments of the 39 appliance makers tracked in the bright green bands of Figures 1 and 2, and divided them by the amount of product sales the companies reported over the same periods. And for some reason, it's clear that the companies' claims and accrual rates are usually highest in the second and third quarters of the year, and lowest in the fourth and first -- except in the years 2003-2005.

But despite the seasonality since 2005, there is another trend to consider: Since 2009, both the peaks and valleys of the curves have been lower from one year to the next, suggesting that there's some warranty cost reduction under way. It's not a dramatic decline, but the last five years have seen average claims and accrual rates decline from 1.7% to 1.8% in 2009 to 1.0% to 1.1% now.

Therefore, the major appliance manufacturers tend to see somewhat higher claims and accrual rates than the HVAC manufacturers. However, what the HVAC manufacturers also do is provide product warranties that last much longer. We're not charting another metric that we calculate, which is based on the size of the warranty reserve fund, divided by the average amount of claims paid per month. But if we did, it would show that the HVAC manufacturers keep a much higher level of reserves on hand than do their appliance-making peers.

For instance, at the end of 2013, the HVAC makers had reserves of $2.8 billion, and that was about 28 times as much as they were paying out each month in claims. Or to put it another way, their reserves would have lasted for more than two years, even if they stopped adding in new funds.

Meanwhile, at the end of 2013 the appliance makers kept reserves of just under $2.2 billion, which was equivalent to 17 times as much as they were paying per month in claims. So their reserves wouldn't have lasted as long, which is an indicator that their warranties have shorter durations.

It's just another way of measuring the size of a warranty reserve fund. Below in Figure 5 we're measuring it in dollars. But it also could be measured in terms of its capacity to pay claims. And that, in turn, is related to accruals, because the amount set aside when a product is sold is supposed to last until its warranty expires.

Larger Reserve Funds in HVAC?

In Figure 6, another pattern seen in Figures 1 and 2 is reversed. With those warranty metrics, the appliance manufacturers almost always paid more claims and made more accruals than the HVAC companies did in any given year. In Figure 6, however, there are only a couple of years where the reserve ratio is close to 50/50, while there are years where it's almost 60/40 in the HVAC companies' favor. So the HVAC companies must prefer to have larger balances than the appliance makers.

It's just another way of measuring the difference between the two groups. What we're saying is that two groups pay roughly the same amount of claims per year and make roughly the same amount of accruals per year. But in Figures 3 and 4 we showed how the appliance makers spend a higher percentage of sales on those warranty expenses.

Now we're saying that the HVAC makers keep more reserves on hand than the appliance makers. So to be precise, what we're saying is that while they keep more reserves, they spend it over longer periods of time. So the spending rate looks lower, but the duration of the spending is longer.

Let's take a specific example. Whirlpool spent $288 million on claims and made $292 million in accruals last year, which was roughly two percent of its product sales revenue. At the end of 2013, its reserves stood at $191 million, equivalent to about 11 months of claims payments at the time.

On the HVAC side, Lennox International (a smaller player, but one of the few "pure plays" left in the industry) spent $21 million on claims and made $30 million in accruals last year. Its reserves at the end of the year stood at $82 million, equal to 41 times what it was paying in claims at that moment.

This suggests that Whirlpool feels safe with about a year's worth of claims payments in storage, while Lennox thinks it needs three or four years' worth. And that would suggest that either Lennox has longer warranties to manage, or it is storing excess funds needlessly. We hope it's because of the longer warranties.

Accrual "Market Share"

Finally, in Figure 6, we took the annual accrual totals from Figure 2 and set them all to equal 100%. Then we figured the warranty accrual "market shares" of some of the top appliance and HVAC companies. But we did it in such a way that it would highlight some of the past decade's merger activity in the appliance and HVAC industry sectors.

The top seven companies are each represented in the chart. And then we also included four companies that are no longer reporting their warranty expenses. Maytag merged with Whirlpool in 2006 (the dark and light green bands). York merged with Johnson Controls in 2005 (the dark and light blue bands). Trane merged with Ingersoll-Rand in 2008 (the dark and light pink bands). And Goodman Global was acquired by a private equity fund in 2007, which later sold it to a Japanese company. So it hasn't reported its warranty expenses for more than three years now.

As can be seen in the chart, the warranty expenses of the merged companies were combined in a way that their "market shares" remained about the same as when they were separate. Over the years, Whirlpool's and Ingersoll-Rand's have actually narrowed a bit, while Johnson Control's widened.

Whirlpool's warranty expenses have risen more slowly than sales, which means the company's claims and accrual rates are below where they were pre-merger. That's partially because of rising product quality, and partly because of shortening product warranties. Whirlpool's product sales were up only about two percent last year, so that was probably not a factor.

Meanwhile, since their respective HVAC acquisitions, Johnson Controls' expense rates have risen, while Ingersoll-Rand's have remained about the same. However, keep in mind that Ingersoll-Rand already had a warranty-intensive product line in place when Trane came along, so the newcomer's warranty expense levels blended well with those of the other divisions (that made golf carts, refrigerated trucks, assorted pumps, compressors and tools, etc.).

Raising Warranty Expense Rates

In contrast, before buying York, Johnson Controls was making passenger car interiors, seats, brakes, and other auto components that didn't throw off a whole lot of warranty expense. So the arrival of York and its typically high-level and long-term HVAC warranty expenses was at least initially a bit of a shock to the system. The combined companies' expense rates rose to an uncharacteristically high level in 2009 and 2010, but they're back down below one percent now. They're just not back down to the 0.2% to 0.4% levels they were at before the York acquisition.

And then finally, Jarden didn't appear on the scene until after it acquired Coleman, Sunbeam Products, and the Holmes Group in 2005. In 2013 it was the seventh-largest warranty provider in the combined appliance-HVAC sector, with claims payments of $151 million and accruals of $153 million. Its expense rates were typical for a small appliance maker: around three percent of product sales.