For reasons of time management, personal dignity and avoiding legal liability, I did my best to avoid any involvement in securities offering due diligence when I was a banker. Still, I think I have some idea of what goes on on a due diligence call for a $1.5 billion bond offering by an investment-grade company with $10 billion of debt already outstanding. So let me attempt to dramatize for you the call between CVS Caremark, Barclays, BofA, BNY Mellon, J.P. Morgan and Wells Fargo on September 8, 2009:
Junior Barclays banker: Anything we should know about?
Junior CVS treasury employee: Nah, everything's fine.
Junior Barclays banker: Cool.
And off they went to push the button on the deal. But everything was not, in fact, fine. CVS was harboring a deep dark secret:
In offering documents for a $1.5 billion bond offering in 2009, CVS fraudulently omitted that it had recently lost significant Medicare Part D and contract revenues in the pharmacy benefits segment. Investors were therefore misled about the expected future financial results for that line of business. When CVS eventually revealed the full extent of the setbacks on Nov. 5, 2009, its stock price fell 20 percent in one day.
That's from Tuesday's announcement of a $20 million settlement between CVS and the Securities and Exchange Commission, for two unrelated bits of naughtiness. The first naughtiness is that bond deal: CVS sold some bonds without telling investors that it had lost a bunch of contracts. This means that CVS, "in the offer or sale of securities ... employed devices, schemes or artifices to defraud which operated as a fraud or deceit upon purchasers of CVS securities," i.e. those bonds.
This one is sort of weird, because those bond purchasers probably didn't care that much. They're not buying 30-year CVS bonds for that year's earnings, and while a few lost contracts don't fill debt investors with glee, nor do they seem to have filled them with terror. "When CVS eventually revealed the full extent of the setbacks on Nov. 5, 2009, its stock price fell 20 percent in one day," but the price of the bonds it sold in that offering fell by about 1.5 percent.1 You might, if you were CVS, argue a little about whether this disclosure failure was in fact material to the bondholders who were supposedly defrauded by it.2
But I guess you wouldn't argue that much, since you're not supposed to put out misleading disclosure anyway, because your equity investors might read it and get ideas. Here, the idea that you hadn't lost a bunch of contracts that you had in fact lost. I don't know. The moral here seems uncertain. Do better due diligence? (But the banks didn't get in trouble for missing this.) Don't issue bonds if you're trying to hide business problems from your shareholders?3 CVS lost its contracts in August 2009, but the SEC doesn't say that CVS needed to disclose that at the time. It let its shareholders believe that everything was fine throughout August, and the SEC is fine with that -- it never says that CVS had an affirmative obligation to disclose the lost contracts. The problem came in September, with this bond deal. If not for the bond deal, everything would have been fine.
Except the other thing, I mean! Because CVS's other bit of naughtiness was accounting naughtiness. While my dramatic interpretation of the due diligence call has no names attached, the accounting awkwardness is attributable to one particular guy. That guy -- whom the SEC went after individually, and who settled for a $75,000 fine and a one-year bar from public-company accounting -- is one Laird Daniels, the vice president for corporate budgeting at CVS until May 2009, when he became retail controller.
He has since moved up in the world,4 and you can see why. Most accountants come to work and ask themselves, "How can I comply with generally accepted accounting principles and accurately reflect the financial condition of my company?" This does not always endear them to their co-workers, who tend not to feel the same strong emotions about generally accepted accounting principles.
Daniels, on the other hand, seems to have viewed himself as a revenue center. At the end of 2008, CVS had bought Longs Drug Stores for $2.9 billion in cash, and by mid-2009 that acquisition seemed to be a drag on profitability. Daniels had other ideas:
On June 16, 2009, Daniels told a co-worker that he had been making “some good progress on the tangible asset side” with the valuation firm and that he was now “extremely confident (by the way that’s the most confident I get) that the final valuation will no longer be a bad guy but rather a good guy.” (Daniels routinely used the terms “good guy” and “bad guy” to indicate whether an item improved or harmed CVS’s purported profitability.)
CVS had allocated the purchase price of Longs among its assets, accounting for the deal as though it had bought all of Longs's assets at their fair value, and putting the excess in goodwill. Of the $2.9 billion, CVS had allocated $229.3 million of the price to stuff that was in the Longs stores. Shelves and carpets and whatnot, I guess. CVS was depreciating that property, which was a drag on earnings. But in mid-2009, CVS was also remodeling some of the Longs stores and throwing out some of the shelves and whatnot. Daniels had the bright idea of:
- saying that that was the plan all along,
- saying that CVS had always valued the shelves and whatnot in the remodeled stores at basically zero and planned to throw them away,
- changing the purchase price allocation to allocate only $39.6 million to the stuff in the drugstores, instead of $229.3 million,
- re-allocating the extra $189 million to goodwill, which unlike tangible property does not need to be depreciated, and
- reversing $49 million of deprecation that CVS had previously taken.
In general you are not supposed to change your mind about your purchase price allocation: You get time to figure out what it was, but it's supposed to be based on your plans and information at the time you did the deal, to avoid exactly this sort of thing. But Daniels managed to convince everyone that this was the intention all along, so they scrapped the original "draft" allocation that valued the stuff in the stores, replaced it with the new one that basically didn't, and so reduced CVS's depreciation cost. That increased third-quarter earnings per share by 2.4 cents, and I guess if you made CVS $49 million in one quarter you'd get a promotion, too.5 That's a good quarter.
The SEC is pretty mad at Daniels for this; they think that his accounting was wrong, and that he was not completely honest about who developed what intentions when with respect to CVS's remodeling.6 But there is another possible reading, in which Daniels, after being promoted to a role in which he had to sign off on the purchase price allocation, went back and corrected previous errors in a way that, sure, made some extra accounting income. I don't know that that reading is right, exactly, but I hope it is; I don't want to live in a world where "profit-generating accountants" and "honest accountants" are totally disjoint sets.
The SEC's complaint just sort of trails off: There's a lot about how Daniels's e-mail trail was not impeccable, but not a lot about economic substance. One question you might ask is, of that $2.9 billion that CVS spent buying Longs, how much did it spend to buy the shelves and carpets and whatnot in the Longs stores? Was it planning to keep all that stuff, or scrap it? Did it in fact keep it, or scrap it? Daniels is not your best guide to answering those questions, but neither is the SEC.7
I guess that's as it should be. Nothing in this story is that bad! The SEC came after CVS, and its individual employees, for technical GAAP violations and incomplete bond prospectuses. It's pretty ho-hum stuff; there's no sexy scandal or popular villain or even a bankrupt company to close the barn door behind. It's just: The SEC wants companies to get their accounting right, even if it's boring and technical. And if it doesn't bring the occasional boring case to make sure that happens, who will?
1 Inexact science because, y'know, it's bonds, but here is Bloomberg TDH on the price of those bonds, which "closed" at 99.9 on Nov. 4 and at 98.4 on Nov. 5, on unusually heavy volume. (Here's the same data for spread, which widened by about 10 basis points. Here's the stock.)
2 You might not win, I mean. I'm just saying you might argue. In fact putting a numerical standard to materiality seems colossally difficult to me. Also, presumably investment-grade bond investors have a lower threshold for market moves than do stock investors, so a 1.5 percent drop might be "material" to them but not to the stock investors. Here is an interesting paper discussing, among other things, materiality in a world of high-frequency trading. But we are getting rather away from the point.
Incidentally, page 84 of CVS's 2013 annual report discloses a shareholder class action over some of these disclosure issues, but no bondholder class action. The bondholders presumably didn't lose enough for it to be worthwhile.
3 Issue them under Rule 144A? Would that work? The SEC complaint dings CVS both under Securities Act Section 17 for selling the bonds misleadingly, and under Exchange Act Section 10(b) for making misleading disclosures in the vague vicinity of its stock. If you offered the bonds in a private offering memorandum that was not provided to shareholders, would that avoid the 10(b) problems? And would the case for fraud in the bond sale be somewhat harder to make? I don't know, discuss among yourselves.
4 He became chief accounting officer in 2010 and is now senior vice president for international operations and business development.
5 From Bloomberg I see 71 cents of Q3 2009 earnings, versus estimates of 64 cents. So Daniels was responsible for over a third of the earnings beat.
6 They also engage in some good literary analysis:
On September 25, the Manager of Property Accounting provided Daniels with a spreadsheet listing the amount spent on remodeling: (1) each Longs location whose assets had been written down to zero, and (2) each Longs location whose assets were included in the residual valuation of $50 million:
- You had asked me to compare the Long’s Personal Property acquisition assets that retained their value ($50M) following the recent revaluation, to the total spending for the Long’s Reset capital projects. This analysis is attached.
- As you can see on the “Summary” tab, the theory you outlined to me holds true for the majority of the stores. Those stores that have retained-value acquisition assets generally have lower levels of spending on their Reset projects. Conversely, those locations that have significant Reset spending generally do not have large acquisition asset activity. There are exceptions, of course, but in general the attached information supports your position.
The language in these e-mails is very revealing. Daniels asked for data about “reset spending” at the Longs stores to help the outside auditors get “comfortable” because CVS was “writing off a lot of the assets.” He did not ask for confirmation that CVS was actually scrapping all personal property in the stores being remodeled. The Manager of Property Accounting told Daniels that, “for the majority of the stores,” the data supported “the theory you outlined to me” and that “in general the attached information supports your position.” The “theory” and “position” which Daniels had outlined to him was simply that stores with lower retained asset values tended to have more “reset spending” – not that all assets in the “full remodel” stores were actually being discarded.
7 Pages 14-15 of the Daniels complaint does get a bit into the weeds of how extensive the remodeling would be:
The overview indicated that, despite the nomenclature, only 111 of the 361 “full remodel” stores (fewer than one-third) were going to receive “replacement of existing wall and gondola fixtures.”
There's a certain amount of that; you can almost picture the remodeled stores. But not quite.
To contact the writer of this article: Matt Levine at email@example.com.
To contact the editor responsible for this article: Tobin Harshaw at firstname.lastname@example.org.