accounting for warranty provisions

[luvb2b]

i need help understanding warranty provisions.

please review the following items from 10q's for q1, q2 2017.

notice q1 2017 provision for warranty = 66,822
notice q2 2017 provision for warranty = 52,797
q1 + q2 = 66,822 + 52,797 = 119,619

notice the total provision for six months ended june 30 2017 = 122,371

why is there a difference?

if you check 2016 figures you will find there is no issue, q1+q2 2016 provisions = provisions for 6 months ended june 2016.

thanks in advance for any insights.

 

[luvb2b]

paging for assistance @brian45011 @surfside @neroden @Reality and anyone else who understands accounting a lot better than me. please see question above.

 

[brian45011]

paging for assistance @brian45011 @surfside @neroden @Reality and anyone else who understands accounting a lot better than me. please see question above.

I think it nets out against "Net Changes In Liabilities for Pre-Existing Warranties, etc"

($3,510) + $8,915 = $5,405 (6 months shows just $2,653) Difference is $2,752 same as the discrepancy you asked about. Accrued "Warranty at End of Period" is accurate. No sure what was going on but that was about the time the transition from Wheeler back to Ahuja was happening

 

[cpa]

My guesses are (a) that someone did not bother to check the numbers (yes, it happens to all of us at one time or another), or (b) Tesla had an immaterial change in its accounting policy that they applied retroactively in quarter two (in other words, they closed the books for quarter one, and they made a post-closing entry in quarter two retroactive to March 31.) Probably should have made a comment in the verbage in the Q2 2017 10-Q like, "certain costs from quarter 1 have been adjusted for retroactive changes in estimates, so some amounts may not directly correspond."

I suspect that Tesla has a type of formula that they use to accrue estimated warranty costs. There is probably a different formula for S and for X, and now for the 3. The formula will in all likelihood change periodically once actual results are determined. This estimate might have been too low (especially considering all the problems that were incurred with Model X.)

Another contributing factor (and this is purely a guess) could be that there was a currency exchange hiccup for all the warranty costs that are reserved for all foreign sales, and this might have been corrected retroactively as well.

 

[Reality]

Likely an immaterial adjustment of prior periods wtihin the Q's, this happens sometimes and as long as its not material it is ignored. Also could be an expense reclass although that would be strange imo.


Its weird they wouldnt just pickup the difference in the current period though.

The 6th month number is correct, and the 3 months ended in Q1 2017 is slightly understated by the difference you are seeing

 

[Reality]

Actually it's a reclass


Perform the exact same exercise on the "Net CHanges in liability for pre-existing warranty......"

and you get the other side of the reclass.

 

[Reality]

I think it nets out against "Net Changes In Liabilities for Pre-Existing Warranties, etc"

($3,510) + $8,915 = $5,405 (6 months shows just $2,653) Difference is $2,752 same as the discrepancy you asked about. Accrued "Warranty at End of Period" is accurate. No sure what was going on but that was about the time the transition from Wheeler back to Ahuja was happening

lol, ya should have read thread, sorry

i agree with this, non issue albeit a bit sloppy accounting

 

[luvb2b]

thanks for your replies @brian45011 and @Reality. i can see how the numbers work now.

any idea on why this retroactive adjustment to provisions takes place? i always thought provisions were a fixed periodic accrual based on product deliveries and that provisions for warranty couldn't be retroactively changed.

it feels unusual considering it didn't happen at all in 2016. i get the cfo changed, but would something like this happen due to a cfo change? should i go back and check 2014 to see if ahuja did the same thing back then?

separately has it gotten this bad that i have to ask bears for help with the accounting?

 

[Reality]

i dont know, but i have seen it in other places.

I wouldnt consider it significant or anything, or even sketchy, more just an error likely caused by something in the consolidation rollup and the Q2 numbers were already audited. So instead of changing the Q2 numbers to make it roll, they just let it slide because it's immaterial.


You will see this sometimes, it's not common but it's not rare either. Quarterly reviews are very light compared to the K audit.

 

[luvb2b]

thanks. i don't recall seeing things like this but am willing to move on.

i dont know, but i have seen it in other places.

I wouldnt consider it significant or anything, or even sketchy, more just an error likely caused by something in the consolidation rollup and the Q2 numbers were already audited. So instead of changing the Q2 numbers to make it roll, they just let it slide because it's immaterial.


You will see this sometimes, it's not common but it's not rare either. Quarterly reviews are very light compared to the K audit.

 

[cpa]

First off, interim statements are not audited. They are typically reviewed by the external auditors, but they are not subject to the same testing and procedures that the year-end audit requires.

The accrued warranty reserve is an estimate, just like the allowance for doubtful accounts, accumulated depreciation/depreciation expense, accrued absences, and too many more to list. These are just fall-outs from the fundamental accounting principle of matching expenses with revenues.

As Footnote 1 so clearly states, the financial statements are prepared using estimates; actual results may vary from estimated amounts. A change in estimate is not a change in accounting method, so it does not require any disclosure to the financial statements or any retroactive restatement as would a change in depreciation method or a change in inventory valuation from FIFO to LIFO.

It is entirely possible that Tesla changed its method of estimating warranty reserves in quarter two retroactive to January 1. It is further possible that they calculated the actual costs of warranty work for 2017 and compared the actual results to the 2017 expense, and determined that they lowballed the reserve. But this information was not completed until after the first quarter financial statements were presented on the 10Q.

Auditors ask for this information as part of their routine testing and evaluations. If the client does not provide it, the the auditors must do these calculations for their workpapers. Auditors must evaluate the reasonableness of estimates based upon historical evidence and projected assumptions. If the auditors deem that the client has not accrued enough in the liability, the auditors will propose an adjustment if the amount is material. If the amount is immaterial, the auditor has the option of dropping it or adding it to their cumulative list of unadjusted differences to be weighed as a whole before the auditor signs off on the statements.