Revenue recognition and inventory management are key predictors of poor earnings quality, but the balance sheet and income statement offer more playgrounds for management manipulation. This chapter combines these important balance sheet and income statement concerns organized by the balance sheet’s categories of assets, liabilities, and owners’ (shareholders’) equity. It’s important to know this most basic of all relationships: assets = liabilities + shareholder equity. In these three categories you’ll find a number of line items to analyze to better assess, whether revenues are overstated, expenses understated, or cash flow unsustainable. (We covered the all-important inventory line in the last chapter.)
Proceeding from line to line, our guide is always that laws, regulations, and pronouncements from the Financial Accounting Standards Board (FASB) and others are rarely black and white. They leave management with discretion. Given the importance of choices among accounting treatments, look first for any accounting changes. Management can hide behind them to manipulate earnings quality. Even legitimate, required, and defensible changes may make it difficult or impossible to compare the current period with another in order to see company trends. Changes in accounting principles, estimates, and reporting entities raise serious concerns.
Accounting treatment changes almost always change comparisons of apples to apples to comparisons of apples to kiwis, obscuring reality and earnings quality. Three types of changes are key: changes in accounting principles, estimates, and reporting entities.
Nondiscretionary accounting principle changes are those issued by a standard-setting body such as the FASB. The company restates prior years to conform to the change—the dreaded “restatement” that, no matter how benign, frightens investors—or shows the cumulative effect of the change in the current-period financials. For the latter, investors have to consult the notes to find the effect for the year of the change.
Frequent changes in principle include changes in inventory method (last-in-first-out, LIFO, or first-in-first-out, FIFO), the method to value pension assets, type of depreciation, and choice of percentage-of-completion or completed contract method. On the surface, these may seem unobjectionable, but a company cannot change a discretionary practice willy-nilly. The change must be to a practice that’s in some way better than the prior treatment. Of course, this is not always the case—“better” is hardly a black and white standard. If a company wants to change principles to achieve a certain result, it will do so. Investors must be on guard that when a company chooses a new method, it may not be for best practices, but rather to obscure poor results by making comparisons difficult.
Financial statements are chock full of places where management must estimate, and those estimates are opportunities for manipulation. For example, if management lengthens the depreciation and amortization (D&A) period, it reduces operating expenses, boosting EPS. In 2005, VeriFone Holdings reduced its amortization expense 43 percent from 1.4 percent to 0.8 percent, artificially boosting EPS. Lengthening a depreciation period boosts EPS in the short term and hurts it in the long term, because the depreciation, although smaller on a yearly basis, lasts longer. So after the accounting change, a company will still be taking depreciation expenses in years when it wouldn’t have been under the previous depreciation term. The resulting change in EPS signals that management may be doing other things as well to meet nearer-term Wall Street expectations.
The place to start is to track D&A as a percentage of fixed assets. If the percentage declines, the company may have extended the D&A period. Then check further. Management estimates warranty obligations, collections on accounts receivable, returns, use of tax-loss carryforwards, the useful life of a car, truck, or airplane, and much more. Investors must exhume the changes in estimates from the footnotes.
In mergers , the reporting entity changes. This offers management a way to consign all sorts of expenses and charges to the past, making the newly formed company’s first reports look better. Acquisitions are tougher, because the reporting entity doesn’t change, but comparisons are cloudy or meaningless.
It’s not impossible to get to earnings quality roots with mergers and acquisitions, but often it’s better to just avoid the stock.
Throughout this chapter you will see these three areas of accounting changes crop up in different ways.
We start with two key earnings quality indicators from the balance sheet accounts receivable and allowance for doubtful accounts, and then move to goodwill, intangibles, and other assets. Each allows you to spy aggressive accounting and deteriorating fundamentals.
Reported accounts receivable have two components. You will see “accounts receivable, net of doubtful accounts,” offering another chance to fiddle with the accounts receivable number. “Doubtful” isn’t exactly scientific. This allowance offers its own set of possible concerns.
First, it’s a good thing that a company reports accounts receivable “net of allowance for doubtful accounts,” because even the very best business is going to have deadbeat or failing customers now and then. But there is extraordinary room for judgment in how likely a customer is to pay, so this is another opportunity to uncover earnings manipulation. Manipulating this allowance affects receivables, and it carries across the financial statements. This can be where management finds that penny or two the company needs to please the Street at earnings time and keep its stock price out of trouble.
Reducing doubtful accounts may be an unsustainable source of income. No ongoing business has zero doubtful accounts. Reducing the allowance can then be an earnings quality red flag.
Finding where management manipulates the allowance is complicated. Not all companies disclose the reserve on a quarterly basis, making it difficult to track. Some firms disclose reserves annually, but then it can be too late for investors to identify a trend. But where the quarterly information is available, compare the quarterly allowance for doubtful accounts as a percent of gross accounts receivable. Do this by adding the allowance back to accounts receivable and calculating the percentage relationship between the two amounts:
If the percentage of allowance for doubtful accounts drops sharply relative to gross accounts receivable, it may indicate an artificial boost to revenues and, therefore, earnings. The company likely did not record enough bad debt expense to maintain previous reserve levels. As a result, the lower expenses bring higher operating margins and higher net income.
What we want is that the allowance tracks gross receivables, rising and falling in tandem and showing that customer-credit quality is consistent as revenue grows. Declines (using the allowance as a cookie jar) and increases (customer-credit quality is worsening) alike are suspect, pointing to poor earnings quality and likely a deteriorating business.
In September 2004, John advised clients that the allowance for doubtful accounts at Silicon Laboratories and Synaptics was a probable source of EPS manipulation:
During the June quarter, Silicon Laboratories failed to increase its allowance for doubtful accounts to keep pace with the year-ago quarter. Given that DSO jumped in the quarter due to back-end-loaded revenue in the period, we believe the company’s receivables carry more risk, not less . Had the company maintained the allowance percentage comparable to receivables growth, reported EPS would have been $0.01 lower.
Table 4.1 shows that despite an over 100 percent year-over-year increase in receivables from $33 million to $69 million in Q2 2004, the allowance decreased from 2.8 percent to 1.9 percent of receivables. It is highly doubtful that a doubling of receivables—even if for good reasons, such as revenue gains—are of such a substantially high quality as to merit about a 33 percent decrease in the allowance. Business just doesn’t work that way.
Table 4.1 Silicon Laboratories’ Questionable Allowance for Doubtful Accounts: June Quarter 2003 through June Quarter 2004*
When the company contorts but can eke out only another penny of EPS, there is no clearer signal of manipulation done to please the Street.
John turned to Synaptics:
Similarly, in 2004, Synaptics also obtained an earnings boost by reducing its level of allowance for doubtful accounts in the quarter relative to gross accounts receivables. We think the large ramp in receivables increased the risk of collection and the company should have increased the allowance level. In the September quarter, the allowance was just 0.6 percent of gross receivables, which is down from 1 percent a year ago—a 40 percent drop. We also maintain that the absolute level appears to be low as well. Table 4.2 shows accounts receivables have more than doubled in the past six quarters, while the allowance hasn’t budged.
Table 4.2 Synaptics’s Questionable Allowance for Doubtful Accounts: June Quarter 2003–September Quarter 2004*
John’s June 2005 report alerted clients that Maxim Integrated Products increased accounts receivable while reducing its allowance for doubtful accounts:
Despite the increase in accounts receivable, the company reduced its allowance for doubtful accounts and may have achieved a boost to reported earnings. Table 4.3 shows that the allowance as a percentage of gross receivables fell in 2004 to 2.4 percent from 3.9 percent in the year-ago period.
Gross accounts receivable jumped 53 percent while the allowance as a percentage dropped 38 percent—from 3.9 percent to 2.4 percent. Because a certain portion of receivables represented extended payment terms to international customers—business that is growing for Maxim—we might expect to see the allowance increase rather than decrease . We note that the company reports its allowance for doubtful accounts on an annual basis. Assuming no changes in recent quarters, reported earnings per share would have been significantly less in 2004 than in 2003.
Table 4.3 Maxim Integrated Products Questionable Allowance for Doubtful Accounts: June Quarters 2002, 2003, and 2004*
Businesses just don’t see their customers’ payment quality increase faster than the rate of new receivables. The decreased allowance for doubtful accounts was a bright red waving flag. Trouble came soon. In Maxim’s 10-K filing for fiscal year 2005, ending in June, the 2004 allowance appeared as $13.4 million, not $4.9 million; 2005’s was $14 million, while accounts receivable fell from $197 million to $192 million. Maxim borrowed income from the future. Chapter 3’s Figure 3.1 shows the disaster for the stock.
Analyzing receivables and doubtful account reserves reveals whether the company is playing games with revenues on the income statement and is potentially papering over business deterioration. Further down the balance sheet’s asset side, we find more possible warnings in goodwill and intangibles.
Management can manipulate goodwill and intangibles to inflate assets. If goodwill and intangibles are large as a percentage of total assets, it could mean that assets are overvalued due to acquisitions. If goodwill changes without an acquisition, it means the company adjusted a purchase price allocation. A popular momentum stock through 2011 illustrates this.
Customer-relations management software provider salesforce.com acquired three public companies and three private companies in 2010, for $403 million. It allocated the purchases almost entirely to acquired developed technology and goodwill and intangibles, at about 25 percent and 75 percent, respectively. There is no way to know whether the company’s valuations for these are reasonable. Moreover, they accounted for 17 percent of FY 2011 total assets, and with the $277 million purchase of San Francisco Mission Bay real estate for a world headquarters, comprise the entire increase in total assets from FY 2010 to FY 2011. The latter is bad enough: one hardly thinks of San Francisco real estate in calendar 2010 as value-priced. This suggests a management capital decision based on something other than value pricing expected to bring good returns on investment. (Worse, after buying the building, reports are that costs zoomed out of control, management canceled the plan for the campus, and leased significant other space in the city.1) But previous decisions on allocating to goodwill, intangibles, and capitalized software gave management enormous latitude to manipulate asset values.
If salesforce.com were growing rapidly, perhaps none of this would matter. But its growth rates were only good—not great—and EBITDA, EBIT, net income, and free cash flow have fallen dramatically in the quarters prior to this writing. This puts the company in a bind. Because Silicon Valley “tech” companies pay heavily in stock options (“everyone else is doing it, so we have to”), this puts pressure on management to do what it can to keep the picture bright and the stock price aloft (“who really notices the option footnotes and potential dilution anyway?”). The ending is rarely a happy one.
If a company has increased goodwill when it hasn’t acquired a company, it may have created an unsustainable boost on the income statement.
When a company makes an acquisition, it is allowed up to 12 months to adjust the purchase price for the acquired company’s carrying value of assets and liabilities. A write-down in assets, or conversely a write-up in liabilities, reduces the equity of the transaction and goodwill. These purchase price adjustments do not have to flow through the income statement as expenses if done before the end of one year following the acquisition.
As a result of the purchase price reallocation within a year, expenses are kept off the income statement, artificially boosting earnings in the period of adjustment.
This treatment can also affect future periods, and investors’ inattention to these adjustments allows management a vast opportunity for manipulation. If a company reallocates the purchase price by writing down the value of an asset and increasing the value of goodwill, it builds in a future gain. If the company sells that asset later for more than its carrying value, it recognizes a gain on the income statement.
The benefit does not end with built-in gains. If the asset written down is a depreciable asset, then the related depreciation expense will also be reduced, thereby increasing earnings. Keep in mind that, under accounting rules, goodwill is no longer amortized, so operating expenses are immediately reduced. Goodwill is subject to impairment testing, but the process of valuing a company is so gray and based so much on expectations, that a company can delay recognizing an impairment charge for years.
Goodrich Corp. provided a clear example of manipulating goodwill in 2004, when John’s former firm Behind the Numbers notified clients of concerns at the company:
According to Goodrich’s June 2003 quarter 10-Q, [a purchase price] adjustment lifted goodwill by $43.4 million. But it was easy to miss on the balance sheet, because the sale of Goodrich’s Avionics business reduced goodwill by $46.3 million. According to the filing, “of the $43.4 million increase to goodwill as a result of business combinations completed or finalized, $42.4 million related to revisions to the purchase price allocation for the Aeronautical Systems acquisition and $1.0 million related to the adjustment of the purchase price of an acquisition due to an earn-out agreement.” The company did state in its 10-K that the purchase price was subject to potential upward or downward revision based on change to net assets between October 1, 2002 and May 31, 2003 as well as on the funded status of employee benefit plans. However, this was a fairly sizeable adjustment. To put the size of this increase in goodwill in perspective, the company booked $500 million in goodwill at the time of the acquisition, almost a 10 percent increase in the goodwill balance.
In addition, the company recorded a non-cash increase to the restructuring reserve of $18.4 million in the first quarter, which, again, did not flow through the income statement. This was to adjust upwards the expected cost of integrating the acquisition.
A material increase in goodwill from a purchase price adjustment requires scrutiny. It means that either assets have been written down, or acquisition-related liabilities have been increased. In the case of the former, it leads to the possibility that assets will be sold later and artificially high gains could be booked. Normal operating expenses could be booked into the acquisition liability and never hit the income statement. In either case, profits can be artificially inflated.
Goodrich’s goodwill manipulation shows one more way management can inflate profits unsustainably. Moving from goodwill and intangibles down the balance sheet to other current assets, capitalization offers management another chance to manipulate the numbers as well.
Another way companies can make the current picture brighter is through capitalizing assets. When “other current assets” divided by revenue increases, it may indicate that the company is treating an expense as an asset and overstating earnings. Track this relative to revenue—other current assets divided by revenue—and track as you do DSO. When this number is rising, it may indicate the company is increasing the practice of capitalizing expenses—treating expenses as assets that allegedly will be repaid out of future income. This treats an expense as an asset, overstating earnings now against an uncertain future when assets may or may not yield compensating earnings.
Mulford and Comiskey identify four analytical tools to detect aggressive capitalization2:
1. A review of the company’s capitalization policies
2. A careful consideration of what the capitalized costs represent
3. A check to determine whether the company has been aggressive in its capitalization policies in the past
4. A check for costs capitalized in stealth
A simple way to cover these four points is to divide other current assets by revenue. Identifying where this increases then points the investor to SEC filings for further examination.
Mulford and Comiskey note that the most conservative treatment is to immediately expense acquired in-process research and development, patents and licenses, direct-response marketing, and other research and development.3 If they yield future income, great, but until then, they shouldn’t be assets that show a stronger balance sheet and greater shareholder equity than warranted.
Software companies may capitalize software development costs once technological feasibility is reached, but this is so subject to management judgment that Mulford and Comiskey note, “In fact, one could reasonably argue that managements can raise or lower amounts capitalized by choice, raising or lowering earnings in the process.”4
Salesforce.com again exhibits a practice emblematic of many software companies.5 Table 4.4 shows the company’s capitalized software costs on the balance sheet for the past three years.
Table 4.4 Salesforce.com’s Capitalized Software: 2009–2011
The bulk of the 2011 acquired developed technology came from three acquisitions. While GAAP allows the treatment, it gives the company great latitude. The decision to boost the balance sheet by FY 2011 acquisitions through allocation to acquired developed technology, goodwill, and intangibles allows management significant discretion to manage the asset side of the balance sheet. It also raises questions that seem more pertinent when salesforce.com’s revenue is growing with consistent gross margins, but EBITDA and EBIT margins, EPS, and levered and unlevered free cash flow are declining dramatically from FY 2010 highs.
Mere questionable capitalization by itself is not the determining factor in earnings quality analysis—it’s one of many. America Online, for example, may have capitalized member acquisition expenses in 1995 and 1996 in order to show a profit. But after that and until it paid a $3.5 million fine in 2000,6 the company grew at a red-hot rate, overwhelming any earnings quality effect from that practice. That lasted until the Time Warner merger and the bottom fell out of the online advertising market and the stock. Salesforce.com is not growing anywhere near red-hot. Currently, it’s merely warm, and in recent quarters, the company is pulling out all the stops—including multiyear invoicing—to put off the day when revenue growth declines and the stock price goes with it. At this writing and based on these and other earnings quality issues, Tom has shorted via puts profitably.
Therefore the frequency of capitalization, the items capitalized, and the relative size of those items tells the investor whether capitalization is a serious concern or not.
Now we leave assets to turn to the other side of the balance sheet, liabilities, which presents its own set of potential earnings quality concerns.
Deferred revenue represents an easy opportunity for management to manipulate earnings quality. This is revenue a company receives before it delivers the product or service. Until it delivers, that revenue is at risk—it’s not yet an asset.
Days in deferred revenue are calculated as:
Track DDR relative to DSO (DSO minus DDR) for 8 to 12 quarters. If it is declining, the company is generating less deferred revenue relative to the terms it is offering customers. This may generate more upfront revenue, but it only borrows from the future.
John started shorting Medidata Solutions in September 2011, when adoption of a new accounting standard showed significant, but unsustainable, revenue growth. The accounting change formed an estimated 70 percent of Medidata’s total $10 million year-over-year revenue growth. Without the change, the growth rate would have been 7.5 percent. The huge increase was not new revenue—only revenue pulled forward—so deferred revenue trends deteriorated after the one-time boost. DDR fell sharply year-over-year (see Table 4.5).
Table 4.5 Medidata’s declining DDR: June Quarter 2010–June Quarter 2011*
Declining deferred revenue suggests that Medidata probably booked more revenue up front due to the change.7 This is a strong sign to avoid, sell, or short a stock.
So far, this chapter has shown the items most predictive of earnings quality problems. The next concerns are less predictive, but when they appear together with others, the investor should investigate further.
When the company credits deferred revenue but debits accounts receivables, balance sheet analysis becomes more opaque. The company is not receiving cash for its deferred revenue, which should make investors skeptical.
Reserves are often lumped into the “accrued liabilities,” “other current liabilities,” or “other liabilities” accounts on a company’s balance sheet, and it is difficult to determine what these accounts contain. It’s typically fruitless to contact the company for information. Therefore, the investor must infer what is driving the change in these accounts on a quarterly basis. This is another balance sheet issue that is also very important to the income statement.
As with all of our important metrics, you should track accrued liabilities (we’ll stick with “accrued liabilities” to cover those “other liabilities,” and “other current liabilities”; the detection of warning signs is the same) relative to revenue for 8 to 12 quarters. If the company takes serial charges, it may indicate that it is building up reserves. If these liabilities relative to revenue drop sharply, it could indicate the company reversed a previously established reserve. This would then be a credit to the income statement at a 100 percent margin, boosting profits and overstating EPS.
Companies maintain reserves for many reasons, usually against a contingent liability. For example, will they have to pay out on warranties or protect against customer non-payment? The key question, though, is whether the company is using reserves to smooth income—often called cookie jar reserves, because you can fill the jar and take from it as you need—resulting in questionable earnings quality. Some warning signs for reserves are balance sheet concerns, but they also affect the income statement.
Both the balance sheet and the income statement can be used to confirm suspicious behavior. For example, if accrued liabilities typically amount to 10 percent of quarterly sales, and in the current quarter they represent 5 percent of sales while operating profit margin jumps 400 basis points for no apparent reason, a reserve may have been reversed and net income artificially bolstered.
Reducing reserves for warranties boosts EPS. Management can manipulate this reserve at will. Returning to Synaptics, John’s 2004 report revealed an earnings boost from a reduction in accrued warranty liability:
In recent quarters, we believe Synaptics obtained an earnings boost through the reduction of its accrued warranty liability. This account has been reduced both in absolute dollars and relative to revenue. The cost of the warranty is estimated at the time of revenue recognition, and the obligation is affected by failure rates, material usage, and service delivery costs incurred by the company in order to correct the problem.
As Table 4.6 shows, while revenue was $38.1 million in September 2004, the accrued warranty was just $596,000, or 1.6 percent of revenue. Conversely, in the year-ago period, revenue was $29.6 million, while the reserve was nearly $1 million, or 3.3 percent of revenue. We also note that the reserve in both absolute terms and on a percentage basis has fallen in each of the past five quarters. Further detail provided in the quarterly 10-Q report shows that the cost of warranty claims and settlements nearly doubled versus the year-ago period to $535,000 from $288,000 in September 2003.
Table 4.6 Synaptics Rrevenue versus Accrued Warranty: September Quarter 2003—September Quarter 2004*
Were investors to believe that revenues steadily increased sequentially for quarters, while the company’s potential warrant liability declined each quarter? This suggests that management may be reducing the warranty reserve to boost EPS. Had the company maintained a 3.3 percent warranty/revenue ratio, earnings per share in the September quarter would have been $0.015 lower.
A penny here, a penny there, and you’re likely seeing desperation and manipulation. But it’s far more than a penny at issue where pension liabilities overwhelm many long-lived companies. Real fear can drive management to actions that can make you laugh out loud, and then avoid trouble or profit.
A huge issue for companies with traditional defined-benefit pension plans is the assumption for returns. Many have used a number like 8 percent annualized during a decade with flat to no returns overall from the market and negative or close to negative inflation-adjusted returns from fixed-income investments. This has meant huge increases in the number of underfunded pension plans. The change of even one percentage point in expectations can change hundreds of millions of dollars. Eight percentage points in a flat period is unethical at best.
Companies like General Motors declared bankruptcy, not the least from the burden of retiree benefits. The same pressure hurt Eastman Kodak, which also went belly-up. Both were for a long time considered widow and orphan stocks, but they may have created instead of supported them. Still alive at writing, Sears Holdings also has large pension liabilities and is struggling to meet them.
No wonder management waves a wand to magically raise return assumptions in lower-return environments. Increasing expected investment return and reducing pension liability also increases tangible book value. The company puts off funding pensions at the level it should if return assumptions were sensible, in effect raiding the pension account. It just borrows from the future until it can no longer do so.
Debt is less easy to manipulate, but the ways companies use debt matters. Most investors begin with one of two viewpoints. One camp believes that debt is always and everywhere bad. The other group holds that debt-free companies should take on debt, because the deductibility of interest expense reduces the cost of capital and may lower the hurdle for management to invest that capital for greater return, boosting earnings and free cash flow. It’s not this black and white.
Of course debt represents risk, and it can indicate trouble. It is axiomatic that unsustainable debt—where the company most simply states it can’t make payments—leads to an unsustainable company. “Unsustainable” can mean, at best, acquisition at a fire-sale price by a stronger company, a recapitalization that almost always leaves common stockholders with a fraction of their former ownership share, or, of course, bankruptcy.
But there is therefore no substitute for the detailed and necessary task of identifying all debt terms, including fixed or floating interest rate, maturity, recourse or non-recourse, convertibility, lender ability to call or convert the debt to equity, company repayment restrictions or options, coverage ratios, and dependability of cash flows.
The typical analysts’ emphasis on the ratio of cash to debt is superficial and a first step only. They may even go farther than one step for companies in the business of selling nonfinancial products. But their eyelids grow heavy with the complexity of debt at financial companies that depend on leverage to increase profits and offer increased risk. Those who require securitizations to maintain their business—timeshare vendors, credit card issuers, and, at various times, mortgage issuers, for example—have with alarming frequency entered bankruptcy or been reduced to a fraction of their former values when credit spontaneously combusts. Know your company’s debt intimately. An investor will encounter at least several credit crises in an investing lifetime.
Adhere to these guidelines when considering companies with debt:
1. Avoid all companies whose business models depend on securitizations or consistent access to credit markets. Prudent risk management would eliminate almost all such financial companies. Instead, consider them strong short candidates at various times in the credit cycle.8
2. Only buy companies with large debt to assets where you can model interest coverage, maturity dates, and other terms, such as convertibility and covenants, with confidence, through at least one business cycle with a serious downside. Refinancing may be required at the worst times—uncertain and unknown in advance, of course—on arduous or no terms.
3. Carefully monitor the uses of debt to determine adequate returns on capital raised. Debt becomes a quality of management and quality of earnings problem when the benefits on earnings from debt interest rate deductions are outweighed by all the ways management can misuse the cash to destroy value and hurt earnings.
4. Debt used to fund dividends and/or stock buybacks must pass the most stringent of tests—and usually fails. If value-minded management finds its shares at a significant discount to intrinsic value and to opportunity costs, debt for repurchases might be a proper use of capital (and using lower-interest debt to repurchase stock and to retire the obligation to pay those shares’ future dividend streams can be a positive allocation and very good for cash flow). But management rarely has the financial training to allocate capital this properly and subtly. Some companies, such as General Electric, can steadily increase debt while paying dividends and buying back shares, but few can do so indefinitely. And GE’s mind-numbing financial complexity should tell investors only one thing: stay away.
Because there is such a large class of income-hungry investors who choose dividend-paying stocks as a rule, debt for dividends deserves further discussion.
A company with a history of dividends eventually attracts a long-term shareholder base because of that dependability, and a history of rising dividends only locks in the shareholders more tightly. A dividend-paying company’s management therefore will sell its first-born rather than cut—let alone eliminate—its dividend. Income-loving investors would sell en masse and destroy the stock price. Therefore, a company may build up debt to pay dividends and put off the day of dividend-cut reckoning.
So when is a dividend good or bad for a company with debt? Value investors like to get paid somehow, through dividends, buybacks at opportunistic prices, buyouts, spinoffs, and other value-creating events. But growth investors view dividends as a signal that a business has no better uses for the capital and that, somehow, growth is over. What should matter is that:
The dividends are paid out of excess capital—capital not needed to run the business.
Management cannot find other investments that offer higher returns on the excess capital.
The debt is not increased to pay or maintain a dividend—unless, for example, at a negative real interest rate, such as large credit-solid companies, including Microsoft, were able to obtain in the post-2008 credit bubble low-interest environment.
Eventually, in almost every case, debt for dividends is unsustainable. What begins as a way to maintain a shareholder base ends when the dividend, at best, remains flat or worse is cut or eliminated. Like so many accounting manipulations, debt for dividends just borrows from the future—putting off a day of reckoning when dividends are cut or eliminated and dividend-hungry shareholders will bolt.
Lost in the past decades of subdued inflation has been the truism that inflation favors debtors and disfavors creditors. Debt with fixed and manageable interest is just as positive for a company as it is for the homeowner with a 30-year, 4 percent fixed-rate mortgage when inflation rises. The flip side, of course, is that a deflationary environment makes the debt more onerous. If you know which environment we are in at any time, you can make a lot of money betting on it (good luck!). In the long term, paper currency inflates, and that’s that, but in the short term, anything can happen.
All balance sheet items play roles in shareholder equity. The value investor on the long side, like Tom, focuses, not on how much can be gained , but, rather, on how much could be lost . One metric is to evaluate hard assets, or tangible book value (TBV) per share.
John’s 2005 report to clients on Helen of Troy included a reminder of why intangible assets and goodwill aren’t protection for the investor. He noted:
The balance sheet is bloated with intangible assets. Acquisitions are often less synergistic than management’s tend to believe, thus creating less shareholder value than perceived. Furthermore, in recent years it has been apparent that there is a tendency to overpay for acquired entities. With respect to Helen of Troy, intangible assets accounted for 46.8 percent of total assets at November 2004 compared with 26.5 percent in the year-ago period. As a result, tangible book value is just $1.1 million. [emphasis added] The breakdown in intangibles includes $201.5 million in goodwill, $157.7 million in trademarks, $29.6 million in license agreements, and $17.4 million in other intangibles.
So let’s stick with the tangible. But even then, in the world of investing some tangible book value is better than others. When it’s made up of a lot of cash, that’s good—except in an inflationary environment. A formula is only as good as its inputs, so all TBV’s material components must be checked—not just cash and debt. TBV is least valuable when accounts receivable and/ or inventory are manipulated, in which case the investor should examine the trend and nature of both. They could increase for the right reasons, such as growing business and more customers, or for the wrong reasons, such as poor business trends, poor customer financial strength, or getting the marketplace wrong and yet showing higher TBV. Assuming too high returns on pension investments understates pension liabilities and overstates TBV. And so on. Therefore, TBV is extremely useful but can’t be taken for granted.
Footwear company Skechers provides a great example of TBV that is static while varying dramatically in quality, due to volatile accounts receivable and inventory. It introduced a new product line in 2009, and when customers avoided it in droves, inventory ballooned. Table 4.7 shows that the company’s tangible book value remained flat for four quarters, though its accounts receivables and inventory trended worse than revenues.
Table 4.7 Skecher’s Tangible Book Value versus Working Capital and Other Trends: December Quarter 2009–September Quarter 2011*
Skechers’ new Shape-Ups shoes were initially successful. Then demand for the shoe, which offered a rolling, convex sole, fell off. Skechers misjudged demand, inventory swelled, and DSO and DSI trended dangerously upward. The effect became most visible with year-over-year comparisons beginning in the December quarter of 2010.
From the December quarter of 2009 to the Sepember quarter of 2010, inventory percentage gains dwarfed revenue increases, until the September quarter of 2011. Accounts receivable rose faster and dropped more slowly than revenues. Gross margins slid and partially recovered recently. EPS, EBITDA, and free cash flow (OCF minus capex) collapsed. Valuation multiples went sky high. The lesson and caveat: And yet, somehow, TBV was basically flat for six quarters, because it didn’t show the worsening quality of the inventory or lengthening receivables.
While certain negative accounting indicators are more important than others, no one of them alone is sufficient. Take them as a whole. The long investor can truly profit by understanding the quality of balance sheet items and their relation to long-side metrics, such as TBV. John profitably shorted Skechers. (Tom notes that at this writing Skechers has corrected the inventory issue and has brought gains to long investors who saw it in process. Today’s short may be tomorrow’s long, when management successfully addresses accounting and business concerns.)
Quest Software provides yet another example of when not to put all your chips on TBV. In October 2005, John advised clients that the company’s acquisition strategy did not create value. TBV raised a red flag. Where Quest’s acquisiting led to increased revenues, earnings, and operating cash flow, TBV increased by a mere penny. At Skechers other issues led to questioning TBV, but at Quest came the reverse: TBV signaled trouble elsewhere. John wrote then that:
We question whether Quest management’s acquisition strategy has created value for its shareholders. Our thesis is based on three observations:
While the company has turned profitable subsequent to the tech-related downturn, those earnings have not flowed through to the balance sheet, which has become bloated with intangibles.
While acquisitions benefit revenue and net income, and thus operating cash flow, the expense is an investing cash outflow. Adjustments to reflect the cash payments result in plummeting free cash flow.
Several acquisitions were the result of related-party transactions or occurred at prices substantially higher than Quest’s own valuation.
Like most technology companies, Quest suffered a downturn in 2000–2001 but has returned to profitability over the past three fiscal years, generating diluted earnings per share of $0.82. In contrast, Table 4.8 shows tangible book value increasing from $1.309 to $1.313 per share from 2002 through 2004, while nearly all of the earnings disappeared, likely the result of pricey acquisitions, and to a certain extent, massive stock option plans as well.
Table 4.8 Quest’s Tangible Book Value versus Pricey Acquisitions and Stock Options
Thus, while the acquisitions have led to increases in revenue, earnings, and operating cash flow, they’ve created virtually no value for shareholders. We are concerned that they have led only to the enrichment of Quest’s management.
Tangible book value is a useful investing tool when it signals problems elsewhere, as at Quest. The investor must analyze the entire balance sheet to know how reliable a measure of true asset value TBV is.
Profit margins hypnotize most investors. As margins rise and fall, it becomes easier not to delve into them. But it’s essential not to take gross profit or operating margins at face value. Companies can control some but not all costs, and fluctuations can be spun all sorts of ways. Sequential changes will impress or depress the quarter-driven Street, only to have the reverse effect next quarter. The Street’s focus encourages management to play along.
Most investors watch the trend in gross profit margin and stop there, yet the components matter. Management may or may not have control over inventory input costs, but their treatment is important. Remember, it is normal for gross profit margins to vary—we’d be suspicious of any that were unchanging. We know they will move. We’re concerned with unusual changes.
“Inputs” are the components of products, from cotton for jeans to silicon for semiconductors. To detect problems in input costs requires industry and company-specific knowledge. Sometimes the company will reveal when an input’s higher cost negatively affected gross profits, sometimes not. When an input rises in price and the company can’t increase product prices commensurately, gross margins decline. But when the input’s price falls and boosts gross margins, management may, like the rooster, take credit for the sunrise, though the price benefits to gross margins and profit are unsustainable.
Teen and tween clothing retailer Aeropostale is an example of a business for which the chief input is cotton, a commodity whose price is volatile, but should be well known to management of a clothing retailer whose main products are jeans and T-shirts. In 2011, clothing retailers, including Aeropostale, faced, not only consumer headwinds, leading to markdowns, but an almost parabolic spike in cotton prices—their single largest product input (see Figure 4.1). This squeezed gross profit.
Figure 4.1 Monthly cotton prices in cents per pound: November 2006– September 2011.
Source: www.indexmundi.com.
Aeropostale reported plummeting EPS for the quarter ending July 30, 2011—$0.04 versus prior year’s $0.46. It cited product mix leading to higher inventory and markdowns, and cotton prices for the decline in gross margins. The stock collapsed.
On November 3, 2011, the company provided that rare animal, a pre-earnings “update”—not a warning, but rather an increased earnings estimate for the quarter ending three days before and for which it would report at the end of the month. And not just any old penny estimate hike, but 85 to 200 percent above prior guidance. Though the prior earnings report and conference call emphasized cotton prices—which had declined at least for the time being towards pre-spike levels—this press release did not mention them, but instead attributed the improvement to “better than expected gross margins for the quarter.” The same sales a year earlier produced EPS of $0.63—more than twice as much as the November 3 update on far better gross margins. But the Street saw only an 85–200 percent jump in estimates. Gross margins or operating margins or both would likely still lag the year-ago period. Table 4.9 shows they did. “Better than expected” still wasn’t great.
Table 4.9 Aeropostale Flat Revenues, Declining Gross Margins and EPS: July Quarter 2010–October Quarter 2011
In fairness, the company didn’t spin this in the press release, and the CEO said, “We remain cautious, given industry-wide costing pressures and the current retail environment.” This may imply that the margin boost was actually due to cotton input price improvement and that product pricing remained at risk. But why did management release an earnings anti-warning, unless it intended to promote the gross margin improvement to obscure its year-over-year and sequential failure to improve margins to normalized levels? The pre-announcement showed management blatantly attempting to manage the stock price.
It worked. The stock jumped 19 percent. Those who bought before, as Tom recommended (see the arrow in Figure 4.2), benefitted from this rise, but Aeropostale’s future earnings power was still unclear. The rise in gross profit, while welcome, would likely still lag. After all, sales still fell a small 1 percent year-over-year and same-store-sales a larger 9 percent, with EPS less than half of what it was a year ago on the same sales. There was no indication of improving trends, and margins and profits were still at risk. When the actual results came at the end of the month, diluted EPS came in at $0.30—two to three cents higher than the early November update and substantially higher than the prior estimate. The stock stayed firm.
Figure 4.2 Aeropostale: May 2011–December 9, 2011.
Source: FreeStockCharts.com, by permission.
It’s possible to argue that Aeropostale couldn’t be expected to plan for such a dramatic rise in cotton prices, but we don’t buy it. They should know what business they are in—jeans and T-shirts are made of cotton, their largest input—especially given their lengthy tenure in the business. While the 2010–early 2011 spike was especially dramatic, cotton prices have varied considerably. The company could hedge but didn’t.
The takeaway here is to know your industry and its inputs for cost of goods sold. Where you can, watch market pricing. Most commodity prices are easy to find. Note whether management fully discloses any effect the inputs have on gross margins and whether execs use subsequent improvement to trumpet growing revenues and margins. Price changes in cost of goods sold inputs are not sustainable as sources of earnings quality, but a dramatic increase in consumer demand and increased volume could make up for it. It is impossible to know Aeropostale’s earning power ahead because of the input costs, so any investor must be sure to obtain a cheap-enough valuation that offers a very large margin of safety.
The next item special to overstating EPS on the income statement is serial charges.
Some companies take numerous “one-time” charges, quarter after quarter, showing that, while a particular charge may be indeed one-time only, the company’s habit is to take one-timers all the time. This is a bright red flag with respect to earnings quality. Investors and professional analysts tend to overlook one-time charges, because, by definition, they are not expected to continue. But serial one-time charges may allow a company to establish reserves or bundle normal operating costs, making future results look better than they are.
Another example is lowering asset values of ongoing equipment via a write-off, which effectively lowers future depreciation expense for an asset still in use, and therefore increases net income. Lower-than-expected selling, general, and administrative (SG&A) expenses may indicate use of previously established reserves, indicating low earnings quality.
It’s relatively easy to spot an income statement littered with charges, but as a guideline, the long investor should run from those companies with serial restructuring charges. They are not worth the effort to figure out. Table 4.10 teaches this, showing Hewlett-Packard’s restructuring charges and merger and related restructuring charges for the 12 quarters—three years! —to the quarter ending in July 2011.
Table 4.10 Effects of Hewlett-Packard’s “One-Time” Restructuring Charges on EPS: October Quarter 2008–July Quarter 2011*
Hewlett-Packard’s recurring charges are a material percentage of EPS. The danger for this company—and all other serial restructurers and acquirers—is that they can bundle whatever normal expenses they want into the charges, and these numbers will significantly affect operating margins and EPS. In any quarter, management could adjust these numbers a little or a lot, creating unreliable results for the quarter and any comparisons. When the charges someday end, as they must, the effect will be so dramatic for this company with flat revenues that the Street will undoubtedly trumpet that things have “turned around.” But an apparent turnaround based on the end of recurring charges is unsustainable.
Because there is zero confidence in HP’s earnings power, a long investor should simply pass. It would also be a dangerous short, because the Street will see a positive catalyst of unexpected and sudden year-over-year increases in this previously negative EPS despite flat revenues.
Another way for management to manipulate earnings unsustainably is through taxes.
In the start-up phase, many businesses do not generate profits. Tax accounting allows the company to carry forward some losses and use them to reduce future taxable income. Because that’s an obvious future benefit, income reduced by the loss carryforward would reduce taxes. Net operating loss carryforwards (NOLs) are a deferred tax asset.
But the value of that asset depends on what profits and tax rates the company estimates it will have. If the company believes that it may not be able to realize fully the NOL benefits, it must record a tax valuation allowance. The tax valuation allowance is a reserve against the possibility that the company may not obtain the full value of its deferred tax asset.
Because it is impossible to know what the future holds, companies may change their assumptions about the value of the deferred tax asset and therefore change the tax valuation allowance. If companies need a few million here or there to meet Wall Street expectations, they may switch their expectations about future income and use of the NOLs to lower—reverse—the allowance, improving current EPS. This may get the company past the quarter to please the Street, but eventually the allowance has nowhere lower to go.
Footnotes on taxes will show if the company reversed the allowance. To know whether to dig deeper to find it, the investor should compare the expected tax rate indicated by management (in the earnings press release or conference call, if at all) and forecast by Wall Street with what is actually reported in the quarter. An artificially low tax rate may indicate that the tax allowance was reversed, benefitting current EPS. Investors should then review the tax notes in the financial statements to find whether the allowance was reversed and whether management had its hand in the tax-allowance cookie jar.
Today, a major tax rate issue concerns companies’ foreign versus domestic tax incentives. Because the United States has the world’s second highest corporate tax rate, after Japan,9 U.S. company managements rationally attempts to maximize earnings in lower-tax foreign jurisdictions. The problem is that this restricts companies’ ability to move capital to where it can best be used.
Financial media frequently mention the huge amounts of cash that U.S. corporations earn and maintain abroad in lower-tax nations. But this isn’t a moral, ethical, or legal issue. It is the law. Therefore, it’s not only entirely rational but is required for a company to defer U.S. taxation indefinitely by keeping the money abroad, especially when the deferral isn’t deducted from net income. Repatriation of the cash can incur taxes up to 30 percent. But the tax situation distorts rational capital allocation. It may make sense not to pay higher taxes or to defer them, but companies might need the cash for domestic acquisitions. Hewlett-Packard, for example, incurred huge debt to acquire EDS in 2009, but couldn’t repatriate the 80 percent of its earnings that provide cash flow overseas. Cash also might be needed to shore up domestic operations, pay down debt, or maintain dividends and stock repurchases. A company that cannot meet these obligations will eventually have no choice but to repatriate cash and incur the tax, giving it a higher effective rate and lowering earnings.
When Behind the Numbers issued its July 2009 report on the offshore profits tax, it identified potential government action to raise revenue from foreign profits and, regardless of that action, risks to tax rates facing seven companies with huge operations, profits, and cash abroad. At the time, Behind the Numbers noted several key points—and at writing the tax distortion persists:
The Government Is Focusing on Foreign Profits as a Source of Tax Revenue
Proposals call for intercompany transactions to be recognized as taxable events.
Limiting expense deductions for companies that utilize laws to defer taxes on foreign profits are being examined.
The [then current] Administration wants to impose sanctions on countries that do not disclose information about tax haven situations.
Rep. Charles Rangel has already floated proposals to limit tax deductions for U.S. multinationals with operations overseas.
Repeating the 2004 Tax Holiday on Repatriated Cash is Unlikely
Companies with large amounts of cash overseas, but high debt and falling operating cash flow in the U.S. could use a tax holiday to bring cash back cheaply to shore up the situation.
The 2004 act was designed to repatriate cash to the U.S. to spur capex and job growth.
In reality, little monitoring was done and many companies launched large share repurchases in 2005.
Another holiday seems unlikely, as the administration has never addressed it, the government needs cash too, and terms of a new holiday would likely be onerous.
Tax Rules and Proposals May Hurt Shareholder Value by Preventing Domestic Acquisitions
Investors would think that the [2008–2009] stock market sell-off would lead to more mergers, given lower stock values.
U.S. multinationals appear unwilling to pay the taxes to use foreign cash and are therefore passing on domestic deals that may be better bargains.
Companies are fearful that offshore cash may be taxed and appear to be favoring complex foreign deals to spend it.
In fact, many large and familiar companies with the widest share ownership are strongly affected. Shareholders of Cisco Systems, Motorola (before its spinoff), Google, Merck, Hewlett-Packard, General Electric, and Coca-Cola are unaware of most of these concerns :
Cisco Systems Shows a High Cash Balance It Cannot Access
Because 81 percent of its $33.6 billion in cash is overseas, Cisco is forced to issue debt in the United States, cash flows are falling, and Cisco is cutting capex and stock repurchases.
Motorola Is an Example of a Company That Needs to Repatriate Cash to Shore-Up U.S. Operations
The company’s domestic tax rate jumped from 26 percent to 67 percent last year, as tax provisions on foreign earnings were changed and non-deductible foreign exchange and translation adjustments were incurred. Nearly all of its debt is U.S.-based, domestic operations are posting huge losses, and cash flow has been negative for three years. Dealing with these issues is hampered because Motorola cannot access $5.4 billion of its $7.0 billion in cash without incurring taxes.
Google Is in Strong Shape, Both Domestically and Internationally, but May Still Have Issues
Google’s tax rate is rising, but there is still a huge tax savings in its earnings, due to foreign tax differentials. The strong operations and cash hoard could make it a poster-child target for new tax law changes. Google lists tax treatment of foreign cash as two of its risk factors.
Merck Also In Strong Shape, but Some Risks Are Still Worth Noting
Merck carries no provision for taxes on cash that may need to be repatriated. But it may need the cash—taxes or not. Vioxx litigation shows that Merck is at risk of large contingent cash needs. Driving shareholder value via repurchases may be tough without some foreign cash.
Hewlett-Packard Shows that a Domestic Acquisition Can Be Tricky
Hewlett-Packard had about 80 percent of its earnings overseas when it bought Electronic Data Systems. The huge debt for that deal will need to be paid with domestic cash flow, but cash flow is down so far in 2009. The company has already had to reduce share repurchases.
General Electric—Here’s Another Risk for GE
Income was boosted in 2008 by increasing the prior-year foreign earnings classified as indefinitely reinvested overseas. GE continues to raise significant capital and has considerable exposure to commercial real estate. More weakness could force repatriation of cash and large earnings hit.
A tax on high-fructose corn syrup sodas is being proposed in the Senate. That could hurt case volume, and Coke already has seen huge drops in volume domestically. Coca-Cola Enterprises has enormous debt to deal with, and weak volumes will hurt more. We still expect that situation to force more margin to Coca-Cola Enterprises and less to Coca-Cola. Declining U.S. cash flow could make it tough to maintain the dividend and stock repurchases.
In December 2010, Behind the Numbers observed that tax rates were already rising for some companies whose sales increased in higher tax foreign countries, such as Japan, that tax rates rose in the United States, and that companies incurred higher taxes in some foreign jurisdictions where tax credits expired. This means that investors—absent another repatriation holiday similar to that of 2004—must look behind reported tax rates to the potential rate, possibly originating from closing the gap between foreign and domestic taxation and/or the necessity to repatriate foreign cash and incur the taxes.
In the years since the 2008 stock market crash, investors have been fond of saying that, with all the cash on their balance sheets, large-cap stalwarts can pay huge dividends and make massive stock buybacks, giving investors safety in volatile times. But as we’ve seen, these billions are often abroad, unavailable for these salubrious purposes. The cash may prove cold comfort.
Now we can pull together this chapter’s interrelated balance sheet and income statement concerns all in one place. It rarely gets this good for earnings quality junkies.
John’s March 2, 2007 report to clients about Internet content delivery company Akamai Technologies, showed practices covered by most of the earnings quality warnings identified in this chapter. Here, focus on how unbilled accounts receivable help the company beat estimates, reserves boost EPS, and margins may be overestimated. These signal that, when such serious concerns appear together in one company, it’s definitely time to avoid, sell, or short the stock.
Surge in Unbilled A/R Aids in Beating Estimates
Given the high expectations embedded in Akamai’s stock price, management is increasingly under pressure to exceed consensus estimates. In the September quarter, the company relied heavily on unbilled accounts receivable to meet, and exceed, consensus revenue estimates. This was the first instance of suspect revenue quality.
For Akamai, unbilled receivables are generally billed within a month after revenue recognition. This form of revenue can be highly subjective, as can the assumptions for profit margin earned on the revenue. As a result, a sharp increase in unbilled accounts receivable could indicate more aggressive revenue recognition.
The Street’s consensus revenue estimate for the September quarter was $108.8 million compared with reported revenue of $111.5 million. Had the unbilled receivables remained at the same proportion of revenue as in the year-ago period, reported revenue would have been $6.4 million less, resulting in a $2.7 million miss—likely to hurt the stock—instead of a nearly $3 million positive surprise.
In Table 4.11, Akamai’s unbilled accounts receivable surged in the September quarter to 16.6 percent of revenue from 10.8 percent in the year-ago period and 12.0 percent sequentially. Previously, the company had exceeded expectations without regard to its unbilled revenue, so our concern is heightened. The information for December is not yet available, but we note that subsequent to the September quarter, the company made new acquisitions, which will further obscure its revenue growth rate. In the December quarter, the company beat revenue expectations by less than $3 million, with $800,000 coming from acquisitions.
Table 4.11 Akamai’s Unbilled Receivables as a Percentage of Rrevenues: September Quarter 2005–September Quarter 2006*
The combination of recognizing a greater proportion of unbilled revenue and subsequently making new acquisitions could indicate a slowdown more marked than what’s evident in the reported financials.
Reduction in Reserve Levels Boosts EPS
In addition to the increase in unbilled accounts receivable, the company has reduced reserve levels, which have boosted EPS in the short term and accounted for a large portion of the EPS surprise. Table 4.12 shows that the reserve levels reported for the September quarter were the lowest of the last five quarters at 10.0 percent and declined by 360 bps year-over-year.
Table 4.12 Akamai’s Receivables Up, Reserves Ratio Down: September Quarter 2005—September Quarter 2006*
In the September quarter, the reduction in reserve levels added $0.01 to EPS in a quarter where the company beat expectations by $0.02. The reduction also added $0.01 in June, accounting for half of the earnings surprise. For the December period, information on the reserves is not yet available; however we note the company beat EPS expectations by only $0.01. We’re pounding the table: A penny here or there is a red flag that management is manipulating to boost earnings.
While December data are unknown at this time, the first sign of weakness occurred in the September quarter, with unbilled receivables. The combination of the potential for more aggressive revenue and expense management raises our concerns with respect to Akamai’s earnings quality ahead.
Wall Street Too Bullish on Margins
Akamai’s gross profit margin has eroded by 200–300 basis points in recent periods, and the slide is expected to continue. Table 4.13 illustrates that gross profit margin in December was 77.8 percent, a reduction of 300 basis points from the year-ago period.
Table 4.13 Akamai Gross Margins
In each of the last four quarters, there has been a downward trend in gross profit margin. Analysts further expect that the trend will continue with margins falling 120 basis points in 2007 and approximately another 100 basis points in 2008. We believe margins could fall further.
The reduction in gross profits is attributable to renewed depreciation from increased infrastructure spending. The company had been generating revenue and profit from fully depreciated infrastructure, which boosted the previous gross profit margin relative to the current period.
While gross profit margin has trended down, operating margins have expanded (see Table 4.14). For the December period, the operating profit margin expanded 420 basis points year-over-year to 34.4 percent, a trend continuing over the past several quarters. Conveniently, Wall Street is estimating that operating profit margins will expand at a rate that offsets the decline in gross profit. Operating profit is expected to expand nearly 200 basis points in 2007 and another 120 basis points in 2008. The EBITDA margin is expected to be 41 percent.
Table 4.14 Akamai Operating Margins
There are several reasons we believed that these estimates may be too aggressive:
The increase in video content will result in more price concessions to price-sensitive customers.
Competition will drive down pricing.
New distribution methods threaten Akamai’s competitive advantage.
The monetization of content is not Akamai’s core competency.
Akamai’s infrastructure expenses could be larger than anticipated.
Margins Will Further Be Pressured by Increases in Capex and Lower than Anticipated Operating Leverage
Akamai’s network operates at 20–30 percent capacity so as to ensure it fulfills performance guarantees to customers. The increase in video consumption will further strain the networks, in our view. In the past, management has managed cash flow by underinvesting in capital expenditures, and we believe reinvestment could accelerate faster than earnings models assume. Therefore, the company will generate less revenue off of a fully depreciated base of servers (depreciated over three years), which further pressures gross profit margins.
Table 4.15 next shows capital expenditures and capitalized software in relation to total revenue over the past five years. Capital expenditures have increased to 16 percent of revenue from around 10 percent in prior years.
Table 4.15 Akamai’s Rising Capital Expenditures and Capitalized Software Costs as Percentage of Revenue: 2002–2006*
If video consumption continues to grow at triple digits rates as some expect, then Akamai’s infrastructure will have to keep pace in order to fulfill service guarantees. However, given the development of new technologies that will compete with Akamai, either the company will have to make more acquisitions or reinvest more aggressively in infrastructure than it has, either of which will increase depreciation and weigh on gross profit margins.
As the market shifts toward video, which opens the door to new technologies and pricing models, we feel that the operating leverage embedded into Akamai’s financial expectations is aggressive. First, increased competition from companies such as LimeLight, with lower pricing points, could slow revenue growth more than expected. The pricing further pressures gross margins. If new delivery methods gain traction as some suggest, this will require greater investment in R&D and other operating expenses.
Table 4.16 shows that the company has invested approximately 6–8 percent in R&D over the past three years and had cut it significantly from prior periods. If the advent of video downloads is a “massive disruption” as many suggest, then the company risks being woefully underinvested in new technology. But even a small increase in R&D spending would blow a hole in the operating leverage assumptions current in Wall Street’s models.
Table 4.16 Akamai Declining to flat R&D versus Revenue: 2002-2006*
This also does not consider increased expenses on the sales and marketing side. Reductions in headcount led to a decrease in spending on S&M relative to revenue in prior periods; however it has begun to accelerate, as shown in Table 4.17.
Table 4.17 Akamai Increasing Sales and Marketing to Revenue: 2002–2006*
These proved more than enough predictors of trouble. At the second following quarterly earnings release, the stock dropped 30 percent in a few days while the best market was positive. Figure 4.3 shows Akamai’s shares losing 78 percent from the issuance of John’s report—the black arrow—to their bottom in the following year.
Figure 4.3 Akamai: Mid-2004–October 16, 2009.
Source: FreeStockCharts.com, by permission.
Our goal here and throughout the book is to find the key items—the best predictors for discerning earnings quality—and view them together. Revenue recognition, which affects the income statement’s top line and all the financial statements, may be the king of the earnings quality mountain, but other lines on the income statement may raise concerns too. Similarly, inventory management is the key balance sheet earnings quality concern, but the investor must closely analyze the other crucial balance sheet lines: accounts receivables, doubtful accounts, goodwill and intangibles, deferred revenues, pension obligations, and debt.
