This is a reasonable thing to do if you think these charges are (a) reasonable and (b) non-repeating.
If the "one-off charges" are not really "one-off" then the "non-GAAP" earnings (presented net of these charges) are a fraud on the gullible.
This is the central point of the series. It would be dead-easy to fake "earnings after one-offs" by putting ordinary expenses in the restructuring budget. I could make margins almost as large as I liked by telling you to ignore costs. Take an extreme example: if I called marketing expenses one-off (and put them in a bucket which I ignored) my margin would look much higher. Telling you to ignore those expenses of course is a sort of con - a Wizard of Oz trick where you tell people to "ignore those expenses by behind the curtain".
The "non-GAAP" earnings presented by Valeant however are not audited. GAAP accounting does not ignore the one-off expenses. The question is whether you - as an inventor - should ignore them like management encourages you to do.
The purpose of this post is to assess whether one-off charges as booked by Valeant are reasonable.
To assess reasonableness I looked at a few mergers where the acquired company had public accounts prior to the merger. It is against those accounts and that business said merger charges arise.
I start with the Medicis Pharmaceuticals merger.
Here, from Medicis's last filed annual report on Form 10-K here is the business description:
Medicis Pharmaceutical Corporation ... together with our wholly owned subsidiaries, is a leading independent speciality pharmaceutical company focusing primarily on helping patients attain a healthy and youthful appearance and self-image through the development and marketing in the United States (“U.S.”) and Canada of products for the treatment of dermatological and aesthetic conditions.Also according to that form 10-K Medicis had 646 full-time employees. No employees were subject to a collective bargaining agreement, and as seems obligatory in a 10-K they believed they had good relationships with their employees. 253 employees were in sales.
Medicis was acquired by Valeant during the year following these disclosures - and the deal closed in December 2012.
The Valeant form 10-K for the year ended December 2012 gives details on the merger including the restructuring charge. Our job in this post is to assess whether those charges are reasonable.
If they are reasonable then we can accept the "non-GAAP" earnings. If they are not reasonable then the "non-GAAP EPS" is a Wizard-of-Oz style con.
Here is what the 10K says about the charges.
Medicis Acquisition-Related Cost-Rationalization and Integration Initiatives
The complementary nature of the Company and Medicis businesses has provided an opportunity to capture significant operating synergies from reductions in sales and marketing, general and administrative expenses, and research and development. In total, we have identified approximately $275 million of cost synergies on a run rate basis that we expect to achieve by the end of 2013. This amount does not include potential revenue synergies or the potential benefits of expanding the Company corporate structure to Medicis’s operations.
We have implemented cost-rationalization and integration initiatives to capture operating synergies and generate cost savings across the Company. These measures included:
We estimated that we will incur total costs in the range of up to $275 million in connection with these cost-rationalization and integration initiatives, which are expected to be substantially completed by the end of 2013. $85.6 million has been incurred as of December 31, 2012. These costs include: employee termination costs payable to approximately 750 employees of the Company and Medicis who have been or will be terminated as a result of the Medicis acquisition; IPR&D termination costs related to the transfer to other parties of product-development programs that did not align with our research and development model; costs to consolidate or close facilities and relocate employees; and contract termination and lease cancellation costs. These estimates do not include a charge of $77.3 million recognized and paid in the fourth quarter of 2012 related to the acceleration of unvested stock options, restricted stock awards, and share appreciation rights for Medicis employees that was triggered by the change in control.
See note 6 of notes to consolidated financial statements in Item 15 of this Form 10-K for detailed information summarizing the major components of costs incurred in connection with our Medicis acquisition-related initiatives through December 31, 2012.
This is a surprising number - at the last 10-K (admittedly released about nine months prior to the merger) Medicis only had 646 full time employees. The termination of 750 people looks aggressive. If the reserves for this are unreasonable the non-GAAP earnings are also unreasonable.
Providing for redundancies for 750 employees when you bought a business that only had 646 employees sounds like over-provision to me - but other people might have a different view and there were a large number of people fired. This article from the Phoenix Business Blog that 319 people were fired the day the merger closed and that they were paid two months pay in lieu of notice. Two months pay time 319 people gets nowhere near the $275 million provision in the above quote. We need to look elsewhere.
More generally we should compare the total charges disclosed or anticipated ($275 million) to the pre-acquisition balance sheet of Medicis. If for example the pre-acquisition balance sheet contained only $20 million in plant it would be unreasonable to write off $100 million. The excess write-off would create a cookie jar which could be used to fake non-GAAP earnings. Indeed that is the central allegation we are addressing.
Here is the final quarterly balance sheet for Medicis as an independent company:
Balance Sheet as of:
Cash And Equivalents
Short Term Investments
Total Cash & ST Investments
Deferred Tax Assets, Curr.
Other Current Assets
Total Current Assets
Gross Property, Plant & Equipment
Net Property, Plant & Equipment
Deferred Tax Assets, LT
Deferred Charges, LT
Other Long-Term Assets
Curr. Port. of LT Debt
Curr. Income Taxes Payable
Unearned Revenue, Current
Other Current Liabilities
Total Current Liabilities
Other Non-Current Liabilities
Additional Paid In Capital
Comprehensive Inc. and Other
Total Common Equity
The source is CapitalIQ but it checks against their last 10-Q.
Not all of these assets are subject to write-down or balance sheet adjustment on acquisition. For instance the cash and short-dated securities are almost certainly able to be converted to cash near par and hence money-good. No write-down there. It might be true that the accounts receivable are not entirely solid, but you would think they are mostly money-good, after all the customers before the merger were roughly the same people as the customers after the merger. And it doesn't make sense to write down the tax assets - after all Valeant is claiming that these are very profitable businesses after the merger - so former tax losses are probably money good.
Writing down intangibles is a wash and has no effect on Valeant's accounts. Valeant has to work out what each of the tangible assets is worth at acquisition, and using this new balance sheet and the price they acquired they deduce the goodwill to add to their own balance sheet. They could write-off some of the property plant and equipment. I guess at the same level they could provide some liabilities, eg they could have a provision for sacking staff. However the main liabilities (long term debt and the like) are not subject to much write-up either. [Looking at a balance sheet usually the debt is a solid number!]
So lets do a (im)plausibility check - lets imagine a write-off so large it is implausible (at least if the business was worth buying for $2.6 billion. Here goes:
Suppose - and this is very nasty,
*. that half of the receiveables are bad
*. and half of the inventory has to be written off,
*. and the entire net property, plant and equipment needs to be written off.
*. And further suppose they have to sack every one of the 646 full time employees and provide $200 thousand per employee (and suppose we provide this notwithstanding that the Phoenix Business Blog suggested that most those employees were paid only two months salary). That might give us the largest plausible provision.
So the answer is $252 million which even with these extreme assumptions is not as much at the $275 million stated in the Valeant form 10-K.
The $275 million number looks like a porky to me. It sure as hell looks like the only way that you can get to that number is to dump ordinary expenses into the one-off bucket. And if you do that the "non-GAAP cash EPS" that the bulls in the company tout is rubbish.
The alternative hypothesis is that Medicis really was awful - and the receivables were bad, and the inventory did have to be written off, and the property and plant was useless because they moved all the manufacturing. And they were so efficient they got to sack 750 of their 646 staff.
Stranger things have happened in my investing travels. Maybe it is all reasonable after all.
Post script. Even Valeant managed to reduce the merger-charges. The 2013 form 10-K re-estimates the provision for this merger at $250 million and gives further break-up as per this paragraph:
We estimated that we will incur total costs of less than $250 million in connection with these cost-rationalization and integration initiatives, which were substantially completed by the end of 2013. However, certain costs may still be incurred in 2014. Since the acquisition date, total costs of $181.3 million (including (i) $109.2 million of restructuring expenses, (ii) $32.2 million of acquisition-related costs, which excludes $24.2 million of acquisition-related costs recognized in the fourth quarter of 2012 related to royalties to be paid to Galderma S.A. on sales of Sculptra®, and (iii) $39.9 million of integration expenses) have been incurred through December 31, 2013. The estimated costs primarily include: employee termination costs payable to approximately 750 employees of the Company and Medicis who have been terminated as a result of the Medicis Acquisition; IPR&D termination costs related to the transfer to other parties of product-development programs that did not align with our research and development model; costs to consolidate or close facilities and relocate employees; and contract termination and lease cancellation costs. These estimates do not include a charge of $77.3 million recognized and paid in the fourth quarter of 2012 related to the acceleration of unvested stock options, restricted stock awards, and share appreciation rights for Medicis employees that was triggered by the change in control.I wonder where in the P&L the difference between 250 million and $275 million appears. Whatever, both numbers seem very large to me.
So far the buy-case for Valeant looks weak.
However this is a single acquisition, Medicis, and the main products of that company Valeant are selling to Nestle. I would like to do this more generally but as a later post shows that becomes increasingly more difficult after the Medicis merger because the one-off charges are not sufficiently broken out by acquisition.
PPS. Someone on twitter is saying that - like the jobs - many of the restructuring expenses to do with the Medicis merger could have taken place in the pre-Medicis Valeant. I will try to address this in future posts.