Roula Khalaf, Editor of the FT, selects her favourite stories in this weekly newsletter.
In part 1 we explained the basics of purchase vs pooling accounting. In part 2 we looked back at the history leading up to FASB’s 2001 decision to mandate purchase accounting and at some of the contentiousness around that decision.
Here we will leave the posturing and examine whether the decision made conceptual sense.
Cash vs stock
One way to frame the question is by thinking about the differences between cash buyouts and stock swap mergers, which have different treatments in both law and financial analysis.
When it comes to takeover law in Delaware, where most companies are incorporated, cash buyouts face a higher standard of review than do stock mergers. Shareholders in cash buyouts are forsaking any future claim to profits and upside. Boards of directors are thus obliged to seek the highest price.
In stock mergers, shareholders are swapping interest in one company for interest in another, and the legal standard of review is the more lenient “business judgment rule”. Moreover, stock-for-stock exchanges are counted as tax-free reorganizations in which no capital gains tax is owed.
At Lex, we’ve spent a lot of time thinking about how investors should evaluate the economic difference. We’ve argued that in stock deals, the most important metric is the exchange ratio that determines how much ownership each set of shareholders gets in the new company. Analyses such as premium paid, multiple paid, and synergy value are more relevant in cash buyouts.
Then should a distinction also apply in financial statements? If two companies are simply melding, why should the seemingly arbitrary “target” company need to have its books updated to market value while the nominal “acquirer” keeps its books at historical cost?
But that’s what happens under the purchase accounting method. Under pooling, at least the treatment of both companies is equitable: each gets to keep their books at historical value.
That said, the FASB’s case for eliminating the pooling treatment is also compelling. Should an acquirer pay a steep price for a target, it seems reasonable that the financial statements should capture this.
Impairment vs amortisation
As for the shift from amortisation to impairment testing, it might seem like a capitulation to lobbying from the business community. But it also appears to be the right approach. Straight-line amortisation over some period of years is arbitrary. The erosion of that premium paid to target shareholders will almost certainly have a non-linear path.
And annual impairment testing, while irritating to companies and a boon to accountants, is conceptually satisfying. If carrying value exceeds fair value, then companies take as big a hit as is necessary to reflect that reality, however painful. Roger Lowenstein, presciently writing in the Wall Street Journal in 1996, proposed such a scheme where purchase accounting would not be accompanied with compulsory amortisation.
Above, we mentioned that the impairment fallout from the dot-com bust laid bare just how atrocious M&A was in that era. More recently, Hewlett-Packard has been humiliated by the $8.8bn impairment charge it took from the Autonomy transaction not just for the size but how quickly (within a year of the deal) the computer giant took it. Under a regime of amortisation, H-P shareholders may not have learned as quickly of its new subsidiary’s troubles.
One provocative alternative to the goodwill and impairment regime: Just have the premium paid be an immediate charge to the buyer’s equity. From Professors’ Brian Carpenter and Daniel Mahoney writing in the CPA Journal in 2008 regarding the 1970 APB Opinion 16:
Even among the many accounting professionals who favored the required use of the purchase method for all business combinations, there was substantial criticism of the purchase method’s required treatment of goodwill as an asset. Many who opposed this treatment thought that goodwill should instead be recognized as reduction of owner’s equity, arguing that the acquiring company willfully sacrificed some of its equity in the hopes of (through the purchased company’s hoped for superior earnings) recouping it in the years that followed the acquisition.
This seems extreme to have to take a large charge immediately but companies could conceivably prefer the “big bath” right away instead of the steady drip of annual amortisation or having to disclose an impairment down the road.
The impact on financial statements
Another problematic consequence of pooling: how it treated historical financial statements. Because companies were thought to be combining, all historical financial statements were restated as if they always were together, making it tricky for observers to figure out the true performance of a company’s operations.
The industrial conglomerate Tyco was a serial acquirer during the 1990s, and in 1999 was laid siege by an investor who questioned its merger accounting.
According to BusinessWeek, CEO Dennis Kozlowski, of toga birthday party fame,
…scheduled a seminar on Wall Street to explain to analysts and investors the intricacies of merger accounting. …
…With the bull market already wobbling, the simple suggestion that something is amiss at Tyco sent skittish investors heading for the exits. And Tyco’s devilishly complex financial statements make it hard for anyone to decipher its results. Unless Kozlowski, 52, can convince Wall Street that his stellar operating gains are solid, Tyco stock is likely to remain depressed.
That would be quite a blow for a company whose whole strategy is built around rapid-fire acquisitions using its lofty stock as currency. Over the past three years, Tyco has spent some $30 billion on deals, $23 billion of it paid for with stock. In fiscal ’99 alone, it spent a record $20.2 billion, including $3.2 billion for medical supplies maker U.S. Surgical last October, $11.3 billion on electronic connectors maker AMP in April, and $2.9 billion for Raychem, another electronic parts company, in August.
From a later story in the New York Times:
Tyco International, whose share price has declined in recent weeks over concern about its accounting practices, is likely to refrain from all-stock acquisitions that use pooling-of-interests accounting, Tyco said today.
The shift in strategy, reported in The Wall Street Journal, was accompanied by an announcement that Tyco would buy back as many as 20 million shares. The company said the purchase of 20 million shares was the maximum allowed under rules governing its purchase of AMP Inc. earlier this year.
Of course, Kozlowski would land in jail for looting Tyco. In 2003, Tyco was forced to restate its 1998-2003 books for a variety of accounting sins including “faulty acquisition accounting”.
Apart from the odd accounting fraud, should merger accounting even matter to companies, shareholders or policymakers? We haven’t mentioned it until now but goodwill amortization, after all, won’t impact cash flows. Yet since earnings and cash flow are often conflated, companies were worried.
Before FASB made the rule changes, there was talk of earnings statements separating out the goodwill line so investors and analysts could isolate the impact of goodwill amortisation. AOL Time Warner was pointing to a new metric in 2001 according to the Chicago Tribune.
The proposed repeal of pooling-of-interest accounting includes the addition of a new line of corporate income statements that would allow companies to report earnings per share before and after goodwill amortisation. “The hope is that the investment community will focus on what’s called cash earnings” before goodwill amortisation, Somer said. If so, the quarterly ritual of companies posting net income per share in comparison with analysts’ estimates could be transformed with an emphasis on cash generation, a reform long advocated by many market observers.
In the years since, companies have been successful in promoting cash-oriented profit metrics.
The most momentous accounting change for tech companies in the past fifteen years isn’t the purchase accounting rules but rather FAS 123R (published in 2004), which mandated the expensing of stock compensation.
Stock compensation, like goodwill amortisation, is a non-cash charge. As such, tech companies have been deep innovators in “non-GAAP” and “adjusted” earnings reports excluding stock comp expenses (always with reconciliation to GAAP in font size 8 on page 74 of the press release).
Twitter, for instance, just reported a GAAP loss for 2013 of $645m, of which $600m could be attributed to non-cash stock compensation expense. And most analysts have been perfectly willing to worship at the altar of those sanitised numbers.
The academy weighs in
There is plenty of anecdotal and circumstantial evidence of bias towards deals using pooling accounting, some of which we’ve just cited.
Let’s look at two academic studies that investigated the link between accounting and dealmaking.
The first study looks at all-stock deals prior to the rule changes, when companies had a choice between purchase and pooling, exploring which method they chose and why.
The title of the report—authored by David Aboody, Ron Kasznik, and Michael Williams —nails the key point, “Purchase versus pooling in stock-for-stock acquisitions: Why do firms care?”
The implication, of course, is that firms shouldn’t care. They study 687 all-stock deals between 1991 and 1997, of which 60 per cent were accounted for as poolings.
Their results are intriguing. They find that “CEOs with earnings-based compensation plans are more likely than others to use pooling to account for acquisitions with potentially large step-ups”. This finding reflects the notion that earnings-based bonus plans are often based on mechanical formulas that are not modified to compensate managers for the “earnings” penalty associated with the purchase method.
In contrast they found “no association between stock-based compensation plans and the purchase-pooling choice, suggesting that top executives are less concerned about the implications of recording large step-ups for their market-based compensation than for earning-based compensation”.
In other words, bosses did not worry about the impact of accounting on stock price performance that would determine how much they got paid. Yay, Efficient markets!
(The Wall Street Journal recently reported about how companies are increasingly crafting pay formulas to avoid being impacted by rigorous accounting.)
A second study examines the question of whether the change in accounting rules increased or decreased transaction activity post-2001.
Caith Kushner, a corporate lawyer, published the study in the NYU Journal of Law and Business in 2006. It examined whether deal volume had been depressed by the rule changes.
Kushner looked at all deals, both cash and stock from 1990 to 2004 and tested the level of deal activity against three factors variables thought to influence deal activity: 1) the availability of pooling treatment, 2) the economic cycle, and 3) stock market value.
He simultaneously tested all three factors through multi-variable regressions. The respective dependent variables were the absolute annual dollar value of deals in a given year and the annual number of deals.
He found that the abolition of pooling did have a negative impact on the dollar value of deal-making after 2001, though it was not statistically significant. But when he looked at the number of annual transactions rather than the aggregate annual dollar value of deals, the impact of pooling went the other way. Pooling was hampering deal activity, though, again, not at a statistically significant level. Kushner hypothesized that in smaller deals, companies wanted the flexibility of purchase accounting, but the amortisation expense was too burdensome. Once the new rules didn’t call for amortisation, companies were more freely able to chase smaller deals.
So we have anecdotal evidence (AT&T-NCR price bump, Yahoo!-GeoCities merger agreement, breathless commentary of tech executives and Wall Street research houses) as well as some academic evidence that pooling treatment was positively associated with M&A activity.
Setting aside accounting esoteria, perhaps what’s most interesting for market observers is what has happened to the volume of stock-for-stock deals after the rules changed in 2001. Remember there is a level playing field between all-stock and cash deal from a bookkeeping perspective.
Stock-for-stock deals have shrunk considerably (this year’s proportion is distorted by Comcast-TWC). Some of this may reflect the abolition of pooling. But other salient factors in the drop may be what’s happened to the cost of debt and the cost of equity. Cheap debt has made cash deals more attractive.
And as P/Es have fallen off, stock currency doesn’t look as appealing.
(Excuse the different time periods, good data for S&P 500 earnings only begins in late 1998.)
To the extent that cash buyouts are preferred to stock mergers, we don’t think that is a positive development.
Economically, cash buyouts are more dangerous for buyers (though if they turn out well, they are more lucrative to existing shareholders too). Companies face two kinds of risk: business and capital structure. Capital structure risk can mitigate or exacerbate business risk. In a cash buyout, existing shareholders are paying a premium up-front for synergies and are taking on all the business risk of achieving them. The added financial risk from debt makes their margin for error even less.
In a stock swap, the business risk remains and, if things head south, shareholders will suffer from only from dilution. But there is far less financial distress risk because no incremental debt has been taken on (apart from what is assumed from the target’s books).
We argued at Lex that when Facebook bought WhatsApp for $19bn, it was risky strategically but financially defensible, since $15bn of the price came in Facebook paper. If WhatsApp flops, then shareholders will be hurt by the extra shares issued, but otherwise the company won’t face liquidity or solvency problems.
The tech M&A boom of 1999-2001 ended in misery not because the deal themselves were necessarily ill-structured—stock swaps are financially conservative—but because the underlying acquirers became dogs with bigger fleas.
The internet service provider and web portal Excite@home declared bankruptcy in 2001. But when Excite and @home merged (very clever combined name), it was an all-stock deal, though one that chose purchase accounting. But since few wanted to search the web using Excite or get their internet from @home, the company failed. From CNET in 2001:
In an attempt to provide content that would complement its high-speed connections, @Home bought the Excite Web portal for $6.7 billion in January 1999. Also as part of this strategy, the company in 1999 spent $780 million for Blue Mountain Arts, a provider of online greeting cards.
But the pipes-and-content strategy failed as online advertising revenue shriveled and investors fled high-flying Net stocks. As a result, Excite@Home went through several management shake-ups and strategy shifts, all of which failed to pull it out of a downward spiral.
But that would have likely been the outcome if they never had gotten together. The collapse of 2001-2002 was more about poor companies than about poor deals. Here is a great post-mortem of Excite@home from Wired in 2001.
There are reasons for hope that today’s startups are of higher quality. Here’s a nice chart from Bloomberg showing how the higher quality and cheaper valuation of newly listed companies that demonstrates greater prudence in today’s boom. Lex and the FT’s James Mackintosh similarly argued last week that revenue, rather than eyeballs or narrative, are more present in today’s high fliers.
And the buyer universe for these still pricey newbies is limited to a short list: Apple, Facebook, Google, Amazon, Microsoft, Oracle, Cisco and perhaps a handful of other juggernauts. With the emphasis on building “ecosystems”, they are the most active acquirers of pricey assets even if the targets reside well outside their current core strength. Say, online advertising king Google buying a thermostat maker for $3bn.
Not that these companies won’t err in deals. Rather they are so big, valuable and (especially) cash-rich that however they choose to fund a big deal, the failure won’t be catastrophic.
Just embarrassing, especially when the goodwill impairment charges are disclosed.
Comments