Back in January, I explored the popular short seller thesis that Tesla is unfairly cooking its books by not accruing enough money for its warranty work. Digging into the problem, I found the site WarrantyWeek.com, an incredible resource for those interested in the world of warranties and how they work.
As I continued my Tesla research, I kept returning to the warranty question. How was this being done? Were my original conclusions from my first warranty article correct? Was I missing something?
I got to look into it further, and it turns out I was missing something and the shorts are sort of right — something odd is going on with Tesla’s warranty accounting. It’s just not what the shorts think it is.
I used WarrantyWeek.com as the main source for my first warranty article. It’s a fascinating web site if you’ve ever wondered about warranty management systems, claims management systems, online warranty processing services, extended warranty services, product warranty program administration, and online registration.
Editor Eric Arnum reached out to me after that piece to share that he was glad I found his work useful and if I ever wanted to ask any further questions, to reach out. This connection was better than I had hoped for – direct access to one of the top experts on warranties and warranty accounting in the country! I made note to keep thinking about Tesla warranty questions Arnum may be able to shed new light onto.
In early June, a trip to a Tesla Service Center for some warranty work on my own car made me question how the experience was different than my Nissan Leaf recall experience just a few months earlier. Suddenly, I had questions, and an expert to ask.
Does Tesla Have An Advantage?
The ways Tesla and Nissan have set up their service departments are completely different. Tesla fully owns its service centers, while Nissan’s service is performed by its dealership network. Could this scenario lead to an advantage? Perhaps service technicians are guaranteed their pay by Tesla, so Tesla doesn’t need to bill labor to their warranty department. I figured Nissan, on the other hand, would need to pay their technicians a prevailing rate to perform the warranty work, as they worked for the dealerships.
Additionally, OEM auto parts are expensive. My Model 3 needed a control arm replacement. Checking for a similar part for my 2013 Nissan Leaf, an OEM control arm would cost $210.65, while a third party replacement would only cost $49.95 … including shipping! Maybe Nissan’s parts department was charging the full price of replacement parts to their warranties, while Tesla was only charging the manufacturing cost of those parts, potentially saving hundreds on each repair.
I felt like I was onto something. Perhaps the bears’ accusation that Tesla hadn’t accrued nearly enough for its warranty could be explained by a structural advantage that Tesla enjoyed due to how the divisions were set up. I emailed Arnum, expecting I’d feel more confident than before that Tesla was being reasonable with its accounting.
Alas, just how I recently used roller coaster capacity to demonstrate how being off seemingly small amounts can drastically affect the outcome, Arnum replied by stating there was no way to be sure about these facts, and reminded me of the danger of basing assumptions off of assumptions. Simply put, it shouldn’t be done.
The only real way to confirm this, he assured me, was to examine Tesla’s warranty accounting by comparing it to other companies.
What Is Warranty Accounting?
Before I go on, I think I need to explain more about warranty accounting. After the Enron bankruptcy, legislators wanted companies to be more clear in their accounting methods, and warranty accounting was one of the categories companies were now expected to disclose. This disclosure was meant to ensure that companies couldn’t use warranty accounting to hide fraud. However, there are still a number of difficulties with examining it. One of the biggest problems is companies choose for themselves what counts as warranty work. For instance, depending on the company, recalls may or may not count against the warranty account.
This lack of clarity creates a system where the shorts are correct — warranty accounting is a location that companies can stash away profits or mitigate losses by playing with their assumptions for the account.
Allow me to draw a comparison to a personal account to illustrate how this works: You have a bill to pay. You can choose to pay for that bill with a credit card or cash withdrawal from your bank account. Either way, you’re going to pay the same amount, it just changes what your monthly statement looks like.
Of course, there are limits to how much money you can shift. Unless you’re planning to carry a credit card balance, you can’t charge more to your credit card than you have in your bank account.
This trick wouldn’t work for a company, but there are accounts that allow companies to do the same sort of thing legally. Warranty accounting is one of them.
How Warranty Accounting Works
To explain it a bit further, with warranties, when you sell a product, a company is supposed to put away enough money to service that product over the lifetime of its warranty. Since it’s impossible to know the exact cost of service a given product will need throughout the lifetime of the product, the company may choose to make adjustments over time to the amount they have to better reflect the amount of service costs the product may need.
Put away $500 for a product only to discover you’re trending toward only paying out an average of $250 in warranty claims? You can claim the other $250 as profit. Discover you were wrong and there is a severe problem you need to fix and you now need $600 per product? Take some of your revenue and put it into the warranty account so it has money in the future.
I am sure conversations about how much to add or subtract from warranty accounting (and other semi-flexible accounts like it) happen in some companies to not just better adjust their warranties, but also to ensure the company is closer to matching Wall Street’s quarterly projections.
This movement based on impossible-to-know future estimates opens warranty accounting to questioning, like what we’ve seen so often with Tesla short sellers. Supposedly expert stock analysts point at the account and accuse Tesla of purposefully not holding enough money in its warranty account in order to pad its profits and limit its losses. If true, this would mean at some point in the future, Tesla will be forced to dedicate millions in revenue to its warranty accounting to make up for the shortfall.
Now that we have a basic understanding of how warranty accounting works, as well as what we can and can’t know about it, let’s take a closer look at what we can know about Tesla’s warranty accounting.
Does Tesla Save Enough For Warranty Repair? An Update
In my original article, I drew a number of conclusions in the “Does Tesla Save Enough For Warranty Repair” section, and I’m going to go through them in brief again to evaluate my claims.
I claimed Tesla owns the lowest warranty claims rate and the highest accrual rate of any US-based automaker. And it does. I stated that Tesla may be holding onto this much extra warranty due to being concerned about how much money it may need to spend if battery replacements become commonplace, but as the July 18, 2019, article from Warranty Week noted, with more than a decade of data and warranty claims not increasing, “how long do we have to wait for events that never happen?”
Next, I looked at claims capacity — or how many months would the money in reserve last if Tesla stopped adding to it. At the time, Tesla averaged 38 months of claims capacity over the past 65 quarters, while GM averaged 24 and Ford 12. While all of this is true, not being a warranty expert, I missed context that Arnum provided me with.
First, let’s update Tesla’s average claims capacity. Arnum calculated this as such: “Since Tesla spent $81m on claims in the first quarter (three months), the [average] monthly cost is $27m, and $1.13b [Tesla’s current warranty reserve as of Q1 2020] / $27m = 41.85 months in reserve.”
Second, Tesla’s $1.13 billion reserve should be put into context. Tesla is one of only ten American companies with more than $1 billion in its warranty reserve. The others are GM ($7.398b), Ford ($5.789b), Apple ($3.923b), Cummins, Inc ($2.332b), General Electric ($2.171b), Deere & Co. ($1.792b), Boeing ($1.549b), Raytheon Technologies ($1.536b), and Caterpillar ($1.533).
The companies with higher warranty reserves generally have so much due to how many products they need to service, not because they are trying to extend their monthly reserves. For instance, GM currently has 25 months of claims capacity, while Ford has 16. The average months of claims capacity that an American company held in 2019 was 20.
When sorting by claims capacity, Tesla has the third largest reserve in the nation, only behind Cummins (at 51 months) and Boeing (at 57 months). However, as Arnum pointed out to me (emphasis mine):
“Note that the only ones with higher months in reserves than Tesla are Cummins and Boeing, and both of them have massive warranty crises on their hands right now — Cummins with the diesel engine emissions and Boeing with the 737 MAX. In other words, Tesla is positioned like its peers who are in crisis, even though Tesla is not in a crisis.“
Tesla’s first Model S sedans from 2012 are just starting to have their warranties expire, so I asked Arnum if there was a way to determine how many vehicles had come off warranty. Unfortunately, it’s one of the mysteries that companies don’t need to disclose — we don’t know how many cars came off warranty due to exceeding the warranty’s mileage, due to accidents where the car is totalled, or due to sales where the warranty did not transfer. Tesla probably has a better grasp on this due to the connectivity of its vehicles, but this contributes to the lack of clarity surrounding warranty repairs.
Allow me to quote Arnum directly on what we can determine:
“From the opposite direction, however, you can figure out quite exactly how many new vehicles were sold and how much was accrued, so you can calculate the exact accrual amount per new vehicle sold. But on that measure, Tesla is literally off the scale at $1510 per unit while Ford is $591 and GM is $356. And it used to be worse: Tesla accrued $2200 in 2018 and $2400 in 2017.”
Here’s where my original conclusion was off — I stated, “If Tesla dropped its warranty reserves to the claims capacity ratio that GM or Ford averaged, it would be able to claim over $200 million in additional revenue back.”
I tend to be conservative with my estimates, and in this case I was massively conservative, to the point I feel like my calculation was flat wrong. Tesla could halve its warranty reserves, which would still put the company above the industry average of 20 months of accruals.
Halving the warranty reserves to bring them more in line with the industry average would immediately allow Tesla to recognize $565 million in revenue.
The Warranty Accounting Mystery
There’s a mystery here, as Arnum succinctly describes:
“… if Tesla knows a way to keep claims costs artificially low, why keep accruals unnecessarily high? And why postpone the official break-even day by keeping expense reports artificially high? If it cut accruals, that would boost net income and drain some of the excess idle money from the warranty reserve. Investors would be pleased.”
Additionally, there is a second mystery to me — for all the analysts who have asked questions, and for as often as the bears have brought this up, why hasn’t this been questioned further until now?
There are over half a billion dollars worth of revenue just sitting on Tesla’s balance sheet waiting to be claimed. Why is it there? And why hasn’t a “professional” analyst questioned it on an earnings call? There are clearly better uses for that money than just having it sit there.
In case it wasn’t obvious, the bears who have been screaming that Tesla is cooking the warranty books haven’t researched it, as that argument has absolutely no real-world basis. Instead, the question investors should be asking is why Tesla is potentially sitting on millions of dollars worth of revenue.
To answer that, Arnum may have the answer and some insight, so I’ll give him the last word (emphasis mine):
“One of these days, Tesla is going to realize it has hundreds of millions of idle excess funds in its warranty reserve, and when it converts some of it into net income, the shorts are gonna scream: Managing Earnings! Manipulating Accruals! Or maybe the fear of provoking that outcry is preventing them from doing it? Either way, the bottom line is that Tesla’s accruals are far in excess of its claims, and its claims are quite low while its accruals are quite high, compared to other automakers.“
A huge thanks to Eric Arnum of WarrantyWeek.com. His expertise was critical to this article. If you have any interest in the world of warranties, you should bookmark this site.