MARKET WATCH: What’s Happening? Earnings reports aren’t what they used to be: New research shows that perfect foresight of quarterly earnings growth no longer generates as much alpha today as it did twenty years ago. Why? Perhaps the market has lost its faith in earnings guidance. Maybe managers are using more accounting tricks to dress up (or dress over) their company’s earnings. Or it could be that GAAP income statements don’t properly account for the growing investment by today’s market-leading firms in intangible assets.
Our Take: This last trend, the economy-wide shift toward intangible investment, is an important phenomenon in its own right. Optimists interpret this shift as evidence that years of heavy investment in hi-tech innovation are paying off (or will soon pay off) in a higher economic growth rate. Pessimists say a great deal of spending on intangibles is wasted on jealously guarding intellectual property (IP) rather than creating it. The pessimistic case squares nicely with the emergence of a “moated,” rent-driven U.S. economy.
Are corporate earnings losing their market-moving potential? That’s the question posed by researchers Feng Gu and Baruch Lev in a new report published in Financial Analysts Journal (FAJ), which finds that even perfect prediction of earnings no longer generates a hefty above-market return.
This striking conclusion has obvious relevance for equity asset managers, who rely heavily on earnings-based performance models. Even more impactful is the companion discovery that the earnings-return relationship may be weakening because of an economy-wide shift toward intangible assets.
The broader implications of this shift are not yet clear—we will run through the possibilities later on in this piece. But let’s first investigate the authors’ analysis of earnings and its impact on markets.
THE IMPORTANCE OF EARNINGS
Quarterly earnings reports, to a Wall Street analyst, are the beginning of all wisdom about a company’s profitability. Within these reports, various measures of earnings (net, pretax, per-share, etc.) take center stage. More than any other figure, earnings growth (or a lack thereof) is used by investors as shorthand for whether a given stock is a buy or sell.
The logic is simple: In essence, buying company stock is a bet on alpha. Through some combination of value appreciation and dividend payout, investors want to get more return than they could have gotten by investing in a benchmark. And of course, that return depends upon the company’s ability to grow its earnings. Thus, as earnings rise (or surprise on the upside), alpha should also rise. Conversely, as earnings fall (or surprise on the downside), alpha should also fall. This relationship is the reason why analysts spend so much time and energy building predictive earnings models and seeking earnings guidance from companies’ investor relations (IR) departments.
A RELATIONSHIP IN TROUBLE
But new research shows that the correlation between earnings and return is weakening. In the FAJ report, researchers Gu and Lev examine historical corporate earnings beats and misses alongside stock market returns going back to the late 1980s. They consider a hypothetical “perfect” investor who, at the beginning of a given quarter, held a long position in every company that would subsequently beat its quarterly earnings estimate and held a short position in every company that would subsequently miss its quarterly earnings estimate.
The conclusion: In that given quarter, such an investor would have generated a rather sizable 6% premium over the rest of the market in 1989-91. But the same investor would have generated just 2% more than the market in 2013-15.
What if we wanted to look not only at beats and misses, but also at absolute earnings growth and decline? The researchers accounted for that as well by constructing a second hypothetical scenario in which an investor, at the beginning of a quarter, held a long position in every company that would subsequently post quarterly earnings growth and held a short position in every company that would subsequently post quarterly earnings decline.
Here again we see evidence of a weakening relationship: In that given quarter, such an investor would have generated a 4% premium over the rest of the market in 1989-91. But the same investor would have generated just 2% more than the market in 2013-15.
It’s noteworthy that the premium fell further in the “beats and misses” scenario than in the “absolute earnings” scenario. Why? Perhaps because investors have caught up to the fact that beats and misses are increasingly gamed by company management, making them more of a corporate construct than an indicator of value. (More on this later.)
WHAT THIS MEANS FOR ASSET MANAGERS
The weakening relationship between earnings and return clearly affects equity asset management as a profession. After all, if even perfect prediction of earnings leaders and laggards no longer generates much of a premium, a fallible human manager who can only occasionally see the future correctly may more often drift down to (or below) the benchmark—making it harder to justify the premium that investors are paying for his or her services.
What can be done? Clearly, asset managers should be on the lookout for alternative valuation models that look at more than just earnings. The FAJ report urges managers to adopt a framework that hinges on “strategic assets,” or assets that confer a competitive advantage on their holder.
Gu and Lev define strategic assets as having three attributes. The first is that they generate net benefits that are a prequel to earnings (e.g., a growing customer base). These are investments a company can make to maximize its earnings within the confines of a competitive business landscape. Dell is a good example of a company whose strategic assets generated net benefits that yielded higher earnings: The firm’s main strategic asset throughout the 1990s was its build-to-order business model, which won the company higher earnings and market share at the dawn of the direct-to-consumer PC revolution. But as imitators cropped up in the late ‘90s and early ‘00s, Dell lost its strategic advantage—and before long, its robust earnings growth.
The second and third attributes of strategic assets are somewhat alike because they both confer quasi-monopoly power on their holders. The second is that they are rare (e.g., wireless telephone spectrum); the third is that they are difficult to imitate (e.g., patent-protected IP). If a strategic asset is rare and difficult to imitate, it gives the company holding that asset a lasting competitive advantage. These are the “business moats” that Warren Buffett loves to talk about; after all, even the most spectacularly productive chateau loses its value if its domain can be easily invaded and occupied by outsiders.
Another common thread tying together these types of strategic assets is that they increasingly involve IP of some sort. IP, to the extent that it is impactful and long-lasting, is like economic rent—companies who don’t own it must pay to use it (think of licensing fees that films must pay when they want to use a certain piece of copyrighted music).
Strategic assets are beneficial not just to companies, but also to asset managers. How so? The presence (or lack thereof) of strategic assets is a better indicator of a company’s future performance than its earnings. Moreover, a company’s strategic assets must necessarily erode before its earnings ever show signs of trouble (because earnings wouldn’t decline while a company still had its competitive advantage intact)—meaning savvy asset managers could potentially predict an earnings decline.
The bottom line: Asset managers who can incorporate strategic assets into their valuations will surely provide more value added than analysts relying solely on earnings reports.
DRIVERS
Let’s now zoom out and focus on the broader picture. What could possibly explain why earnings growth no longer guarantees above-market return?
The “earnings game” rewards smooth talkers, not outperformers. One reason that earnings are losing their predictive power could be that the numbers presented in quarterly earnings reports are no longer truly indicative of a company’s performance. Indeed, an earnings beat increasingly reflects a proactive IR department rather than an outperforming firm.
It’s no accident that, according to a 2016 analysis by The Wall Street Journal, around 75% of S&P 500 firms meet or exceed earnings forecasts. This ratio remains constant quarter after quarter, in good economic times and bad. How is this possible? Because companies have become proficient at ensuring that analysts aren’t overly bullish in their projections. Nearly 2,000 times from Q1 2013 to Q1 2016, according to the Journal, companies would have fallen short of the average earnings forecast if analysts hadn’t changed their figures in the 40 days ahead of the quarterly earnings release.
Negative guidance positions the company to communicate to investors (a) that we are realistic about today and (b) that we are optimistic about the future. Not only did we post an earnings beat, but we expect a huge growth figure in the next quarter. There may have been a time when this tactic was novel and inspired confidence in shareholders. But as the formula has been copied and rolled out marketwide, most analysts have clearly learned the game. In other words, the relationship between earnings and the stock market may have weakened simply because cynical investors have learned to disregard an increasingly manipulated earnings season.
The impetus for companies to talk down analysts has never been stronger than it is today. Increasingly, firms are pressured to turn out positive quarterly performance from shareholders who have grown unwilling to stick with underperformers. By the end of 2016, the average holding period for a NYSE-traded stock was just 8.3 months, down from eight years in the 1950s.
This what-have-you-done-for-me-lately business environment is causing company executives to abandon long-term strategic planning in favor of “making earnings.” This is amplified, according to “short-termism” critics, by stock-option incentives. A widely cited 2005 white paper based on interviews with more than 400 financial executives found that a significant majority would trade long-term value creation for short-term earnings growth. Fully 78% of respondents said they would sacrifice at least some of their firm’s value to avoid a bumpy earnings path, with 26% saying they would make a moderate or large sacrifice. What’s more, 80% said they would cut discretionary spending, including R&D, in order to avoid a quarterly earnings miss. Companies such as Amazon, which boasts about its indifference to near-term earnings reports, are a rarity in today’s market.
To be sure, the earnings game can only be part of the explanation for the weakening earnings-return relationship. After all, as Gu and Lev point out, informal earnings guidance “is neither prevalent nor a new phenomenon. Yet…until recently, meeting or beating the consensus was a winning strategy.” Moreover, earnings manipulation would only affect the “beats and misses” scenario—it would have no impact on the return generated by correct prediction of earnings growth or decline.
Accounting is in the eye of the beholder. A more complete explanation points to the rising prevalence of subjective accounting methods. Today’s corporate income statements contain an unprecedented number of line items that rely on company discretion, which could lead to overstated earnings.
Consider, for instance, “fair value” accounting—which determines the value of an asset based on a company-generated definition of present value, rather than on the objective price paid for that asset. While post-GFC accounting rules attempted to standardize the formula for deriving fair value, this accounting method remains highly subjective and controversial. The rise of intangible assets like patents and IP with no real-world sticker price adds a further degree of subjectivity into fair value accounting. Other accounting tricks that companies use to distort their earnings include taking a “big bath” (overstating the share of losses attributed to one-time charges in a given quarter) and establishing “cookie-jar” reserves (cash used to pad earnings in future down quarters).
The numbers confirm that companies are indeed getting more creative with their accounting practices. Data from consultancy Audit Analytics reveals that SEC-registered firms made 820 combined changes to their accounting methods in 2016, the highest since the firm began measuring such changes in 2000. This figure is 49% higher than the 2000-16 average of 550. Not only are firms making more accounting changes overall, but a greater share of these changes inflates earnings. In 2016, the ratio of accounting changes that positively affected income versus those that negatively affected income was 1.6 to 1—far above the 2000-16 ratio of 1.3 to 1.
Companies could be using favorable accounting to turn a negative earnings growth figure into a positive one, a process that wouldn’t improve the company’s return-generating potential. Such “false positives” could be bringing down the average return of companies that post quarterly earnings growth.
The rise of the intangible economy. Yet there may be a better explanation of the weakening earnings-return relationship—one which takes into account the huge rise we’ve seen in strategic assets that generate net benefits, are rare, and are difficult to imitate. And that’s the fundamental shift toward an intangible economy.
Until very recently, most of the market’s top firms were rich in “tangible assets”—think physical assets like factories, machinery, land, and inventory. As recently as 2001, the most capitalized U.S. firms included industrial giants like General Electric (#1 largest company by market cap in 2001) and oil barons like Exxon (#3), all flush with fixed capital.
But over the past three decades, the rise of super-scalable tech—especially in new industries like e-commerce, the sharing economy, and digital media—has produced a new crop of “asset-lite” market leaders. An early example was Microsoft, which (as of 2006) had a market cap of $250 billion, of which measurable assets constituted $70 billion (mostly financial reserves) and actual plant and equipment constituted a mere $3 billion, barely one percent of market cap. Today, Microsoft has been joined by the so-called “FAANGs”: Facebook, Amazon, Apple, Netflix, and Google. Instead of plants and factories, these companies deal largely in intangible, conceptual assets like patents, copyrights, data, “brands,” and other types of IP.
The shift toward intangible assets is on full display in high-tech fields. The average U.S. exchange-traded company in the software and services industry group spends 42.0% of its revenue on intangible investments, as measured by trailing 12-month (ttm) SG&A expenses plus ttm R&D expenses. (Because most companies don’t break out all of their intangible investments separately on their balance sheets, researchers tend to use SG&A combined with R&D as a proxy.) This is the third-highest mark of any industry group, behind only household and personal products (at 57.0%) and Big Pharma (49.2%). But it bears mentioning that CPG companies have been cutting their intangible investments at a rate unmatched by any other field—down nearly 60% since 2010. Software and services, on the other hand, has more than doubled its intangible investment over that period, the highest growth rate of any field.
The idea that the knowledge economy is expanding at the expense of the goods-making economy is not a new one. (See: “The Spread of the Pink-Collar Economy.”) But this systemic shift has powered the growth of intangible assets in a way that still is not fully understood. According to the FAJ report, U.S. investment in intangible assets rose from 9% of gross added value in 1977 to 14% in 2014. Over the same period, investment in tangible assets dropped from 15% of gross added value to 9%.
To be sure, we are getting better at including intangibles in our national accounts. In 2013, the BEA announced a comprehensive revision of its NIPA data, which included a move to reclassify R&D expenditures as investments. Prior to 2013, expenditures for private R&D were not recorded at all in the calculations for GDP—but as a result of the reclassification, these expenditures are now part of a new series called “intellectual property products,” which also includes investment in software, entertainment, and other non-physical goods. This series has even been backdated to 1929 using estimated data. We can see from these data that IP production now amounts to 4.1% of GDP, a share which has more than doubled since 1980.
How, exactly, does this shift toward intangible assets affect corporate earnings? Quite simply, modern accounting understates earnings for companies that produce a lot of intangible assets. A tangible asset is capitalized; only its depreciation, rather than the cost of the entire asset, is counted as an expense on the income statement. By spreading out the cost of the asset over its useful life, accountants are recognizing that the asset will or could be used to create future value. By contrast, under GAAP, all IP production is expensed immediately. Essentially, value-creating R&D investments are treated the same as everyday expenditures such as salaries. This has the effect of lowering earnings in any given quarter in which a company spends heavily on intangible assets. (Because the BEA by contrast does count IP production as capex, the national accounts do enter an annual IP depreciation charge against national income.)
A further accounting quirk is that not all intangible assets are treated this way—only ones that a company generates internally. The same intangible assets, if acquired from a third party, are capitalized under GAAP just like tangible assets. Thus, because internally generated IP is treated unfavorably on the income statement relative to acquired IP, companies that want to make their earnings appear larger have a clear incentive to buy innovation rather than to generate it themselves. (To be sure, there is a trade-off: Firms more concerned with the tax implications of higher earnings may rather fund innovation in-house than buy it from third parties.)
IMPLICATIONS OF THE INTANGIBLE ECONOMY
What are the attributes of companies that invest heavily in IP? To the extent that such firms are growing in their relative size and importance, how does that growth affect the behavior of the U.S. economy? For an in-depth discussion, see the recent book by Jonathan Haskel and Stian Westlake, Capitalism Without Capital: The Rise of the Intangible Economy (2018). Let us here mention a few overall features of this growth.
More sunk costs. Unlike an investment in tangible (fixed) assets, an investment in IP is typically a “sunk cost.” This means that it if does not pan out, you may not be able to sell what you’ve invested in to anybody else. There will always be someone to buy your used warehouses or truck fleet or machinery. But most likely no one will want to buy your failed brand or your half-baked software or systems design, no matter how much you spent on it. This is why GAAP is reluctant to allow you to put it on your balance sheet. This is why tech firms don’t borrow much (where’s the collateral?). And this is why equity owners in IP-heavy startups bear more risk: If things go bad, the company will have little if any liquidation value.
More scalability and synergy. Along with more downside risk, IP-heavy firms also have more upside risk. A successful IP innovation is likely to be highly scalable—meaning that its lower price or superior deliverability can quickly achieve global dominance with few extra costs. Social media platforms with “network effects” (like Facebook) are so scalable that they seem to enjoy infinitely declining returns to scale. A successful IP-based firm is also likely to experience vast synergies, as its initial great idea is likely to spawn additional great ideas as it grows. A good example of abundant synergy is Amazon, which continuously leverages its IP and business experience to enter new markets, from cloud services and home delivery to food retail and AI robotics.
More market concentration. Economists have tried to explain the recent trend toward rising market concentration (that is, a smaller number of dominant “take-all winners” in each industry) by pointing to several drivers: These include less vigorous antitrust policy; regulatory capture (as with big banks or big pharma); and signs of declining business dynamism. (See: “Where Have All the New Businesses Gone?”) Yet clearly one driver also worth highlighting is the rise of the intangible economy. High downside and upside risks mean that fewer firms will dare enter industries where IP plays a large role—but also that those who do dare enter may soon end up with high profit margins and most of the customers. Where there is easy scalability and synergy, an equilibrium with many competitors is fleeting and unstable: Sooner or later, a winner tends to appear who does indeed take all.
More struggle to protect IP ownership—and acquire spillovers. As the intangible economy grows in importance, the determination of people and firms to define their IP broadly and to exclude others from using it will naturally intensify. And so it has: Witness all the headline-grabbing legal battles over patents and trademarks, ranging from Waymo and Uber to Qualcomm and Apple. To the extent large IP owners are successful, we would expect a greater “spread” in firm profitability and ROR and perhaps a greater spread in the society-wide income and wealth distribution. To the extent large IP owners cannot avoid a “spillover” of privileged knowledge to workers and small firms in close proximity, we would expect these small fry to crowd into urban centers where the winners are located. (Sound like Millennials, anyone?)
OK, so there is broad agreement that the rise of the intangible economy is probably implicated in a wide number of structural shifts in business behavior.
But when we get to next question—whether the net impact of all of the above is a good thing or a bad thing for the U.S. economy—well, here opinions are divided rather starkly.
The optimistic case. An optimistic (and pretty much the mainstream) view holds that this recent, decades-long intangible boom represents an extraordinary wave of technological innovation—centered largely on digital IT breakthroughs—that will ultimately generate dramatic gains in living standards. Like earlier waves that gave us the steam economy or the railroad economy or the electrical economy, this new wave rivals or exceeds its predecessors. So what’s ahead? The optimists point to everything from sentient robots, driverless vehicles, and block chains to gene therapy, the quantified self, and rendered realities.
It's an alluring vision. Silicon Valley ideologists like Erik Brynjolfsson and Andrew McAfee belong to this camp. As McAfee said in 2015: “Digital technologies are doing for human brainpower what the steam engine and related technologies did for human muscle power during the Industrial Revolution.” Ultimately, this side believes, the historic investment in technology by market-leading firms like Google and Amazon will bear fruit in both a high ROR on capital and a high rate of overall economic growth. Sure, this camp acknowledges certain challenges such as growing inequality and underemployment—but they say we can bridge the transition to the future with sufficiently generous public benefit programs. (Some say Mark Zuckerberg will be running on this platform in 2020.)
There is however one big problem with the optimists’ account. Before we start legislating a universal basic income for this dawning tech-enabled cornucopia, we need to confront a very awkward fact: Thus far, there is no evidence that this recent IP boom is generating any economic growth. To the contrary, labor productivity growth is actually decelerating.
Back in 1987, economist Robert Solow famously quipped: “You can see the computer age everywhere but in the productivity statistics.” The problem thirty years later is not so much that we can’t see it—but rather that it has already come and gone. So we see it in our rear-view mirror. Productivity growth rose to a fairly impressive 3% per year during the late ‘90s and early ‘00s. There’s your boom. In the decade-plus since, growth has ground down to 1.2% per year. And during the first full year of President Trump (Q1 to Q4 of 2017), it weighed in at just under 1.1%.
Optimists have issued several responses to this inconvenient fact. Some concede that it is a problem for their account. Haskel and Westlake suggest that the flattening out of IP production as a share of GDP since 2001 (see again Chart 7), after decades of accelerating growth, may explain the recent ebbing of productivity performance. Others say, just hold on: There is always a lag between IP investment and the productivity payoff. But that lag is getting longer and longer.
Still others dismiss the productivity issue as a problem of mismeasurement. Slowing productivity growth since 2005, they say, can be chalked up to the rise of unmeasured “free goods of the Internet” which constitute services that the Bureau of Economic Analysis refuse to count. (Think of a navigation app, which cuts the amount of time users spend in traffic, potentially freeing up more time to spend being “productive” in an economic sense.)
To be sure, this could be happening to some degree. But as mentioned earlier, we are in fact getting better at incorporating the production and consumption of intangibles into our national accounts. And as we’ve written before (see: “The Great Productivity Slowdown: Fact or Fiction?”), a growing body of research undercuts the mismeasurement hypothesis, concluding that, even if it is happening, it could not possibly account for most of the productivity decline.
Besides, if technology were improving our lives in tough-to-measure ways (and, equivalently, if we are overstating inflation and therefore understating real gains in household income over time), we would expect to see evidence of this improvement in quality-of-life surveys.
We see no such evidence: A recent Pew Research survey shows that a plurality (41%) of U.S. citizens say life in their country is worse than it was 50 years ago for people like them. (Only 37% say it’s better.) Meanwhile, according to a 2016 CNN/E-Trade poll, more than half of Americans (56%) believe the next generation will be economically worse off than they are. That’s near an all-time high. The Age of Trump clearly is not an age of standard-of-living satisfaction.
The pessimistic case. Problems with the optimistic account have bred a newer, contrarian viewpoint. According to pessimists, much of what we measure as a rise in intangible investment is attributable not so much to more genuine technological or cultural innovation performed by firms, but rather to rising monopoly rents claimed on the already-owned innovation that firms are exploiting. In other words, a growing share of today’s IP activity consists of the efforts of dominant firms to create new ownership rights over existing IP and to zealously defend them using strict patent, trademark, and copyright laws.
Recall that what’s truly unique about IP is not so much how it enables owners to use knowledge, but rather how it enables owners to exclude others from using knowledge. And, by any measure, the U.S. legal system recently has gotten much better at rewarding expensive efforts by firms to expand and defend what constitutes ownable knowledge.
For starters, lawmakers keep extending the duration of the protections afforded to innovators. When the Copyright Act was originally drafted, for example, it granted 14-year protection to copyright-holders—with the option of another 14-year term if the holder was still alive at expiration. Now, the copyright term has been extended to a full century. (This lengthening duration has been nicknamed the “Mickey Mouse curve,” a nod to the conspicuous fact that the term seems to be extended every time Walt Disney’s copyright on Mickey Mouse is set to expire.)
What’s more, lawmakers have also bolstered the legal defensibility of patents, copyrights, and trademarks. Strong IP protection has become a staple of U.S. trade policy: Regulations like the TRIPS Agreement and the WIPO Copyright Treaty, which were intensely lobbied for by the United States, establish broad-brush protections for IP owners. It’s also become ever-easier for companies to secure patents: Firms like Apple can, with a little persistence, push through patents that had formerly been rejected. While official Patent and Trademark Office (PTO) data show that just half of all patent applications in any given year are granted, this doesn’t count reapplications. In 2013, University of Richmond researchers computed a revised patent “allowance rate” to account for this discrepancy—and found that nearly all patents (92%) are eventually accepted.
Finally, lawmakers have hugely broadened the definition of patentable IP. As recently as the early postwar era, U.S. policymakers took a fairly casual attitude toward patent and copyright protection. To maintain your competitive edge, your firm needed to stay at the cutting edge of innovation in-house. It was assumed that your past accomplishments would soon devolve into the public domain.
Indeed, many leading inventors of that era hardly thought about patentable IP. When the renowned physicist Sir Alexander Fleming realized that penicillin could rid the world of age-old pestilence, he allowed the U.S. and U.K. governments to patent his discovery. Similarly, Sir Tim Berners-Lee, inventor of the World Wide Web, never sought a patent for his idea. The concept that gave birth to the modern computer was never patented, and the transistor itself (sometimes hailed as the most important invention of the twentieth century) was released to the public by AT&T’s Bell Labs for nominal fees—all of which accelerated early IT innovation.
But today, corporations and individuals rush to patent anything of even middling value—from crustless sandwiches (Smucker’s) to rounded corners (Apple). There’s even the notorious case of an inventor successfully patenting a stick (as a pet toy) in 2002. Back during the American High, a copyright was for a written published work and a patent was for a particular and describable concrete device, period. Today, a firm can own exclusive use of a “user experience,” of a “business method,” or of a person’s DNA and personal information.
This strengthening of patent and copyright law coincides perfectly with the Boomer life arc. The same Boomers who once forged a mythic reputation as originators of pathbreaking, system-smashing new ideas have since learned to use the system to transform every one of those ideas—real and imagined—into valuable and protectable property. It was, after all, the late Steve Jobs who once vowed to “patent it all.” And it was David Bowie who, improbably, pioneered how to issue 10-year bonds in return for future royalty streams on his pop music.
A shift in the political economy of idea creation has reinforced this Boomer arc. Back when Boomers were kids, and the G.I. Generation ran America, most U.S. R&D (67% in 1964) was funded by the federal government. Indeed, the federal government spent so much back then that if we look at total U.S. R&D spending over the entire postwar era we see no rising trend over time. Yes, there is indeed a rise in business spending on R&D, which explains why the BEA’s “intellectual property production” measure (which draws directly from these R&D numbers) rise over time. But this rise has been matched by the decline in federal spending—to a mere 22% of all R&D by 2015.
What’s distinctive about federal research? Well, for starters, the public sector rarely if ever patents its high-tech breakthroughs (say in semiconductors or rocket engines) on behalf of taxpayers. Typically, it just lets the knowledge quickly and freely flow through to industry, which accelerates economic growth. Moreover, a much larger share of federal research (then and now) is “basic,” that is, oriented toward fundamental technological discoveries. Business, by contrast, spends mostly on “applied” research and product development that helps firms bring profitable products to market.
In fact, basic and applied research as a share of GDP has hardly grown at all from President Kennedy to President Trump. This may help explain why the strong rise in private-sector IP production since the 1980s has not generated dramatic productivity gains. Looking at national IP production, especially “basic” technology research, there may be no rise at all. (As Robert Gordon, a leading pessimist, puts it: “With a few notable exceptions, the pace of innovation since 1970 has not been as broad or as deep as that spurred by the inventions of the [previous] century.”) It may also explain why pre-Boomer generations of innovators were less focused on IP ownership. In an era when most R&D was a taxpayer-funded public activity, the libertarian ethos of private reward for private initiative seemed generally out of place.
In every modern society, patent and copyright laws are enacted to benefit society over the long haul. And, to be sure, maximizing that benefit requires balancing the need to incentivize the inventor and author (by granting them property rights) with the need to foster progress by transmitting new knowledge speedily to the public. There is no question that this balance has shifted in the direction of the private interest—so much so that it often makes more business sense to buy and protect patents (creating “patent thickets”) than to underwrite real innovation. Apple famously spends more money on lawyers (pursuing and protecting patents) than it does on R&D. And then there are the “patent trolls” who buy up patents and sit on them for the express purpose of suing violators. (See: “Patent Nonsense.”) All told, research suggests that between 40% and 90% of patents are never used by their owners or even licensed out to third parties.
The pessimistic bottom line: Much of what is counted as spending on intangible assets actually pays for patent litigation, the purchase of patents that the buyer never intends to use, and other patent-centric expenditures that can hardly be considered productive investment.
Here are Buffett’s “moats.” Companies with patent-protected IP, and the cash on hand to defend that IP in court, hold a virtually insurmountable competitive advantage over industry newcomers. This lends itself to a darker interpretation of the modern economy—as a place not of innovation, but rather of wall-building. In a moated economy, we would expect to see rising market concentration (see: “Shhh! The Markets Are Concentrating”), rising intrafirm inequality, and even rising income inequality as the employees of “superstar firms” reap the biggest earnings gains. Check. We would also expect that businesses would be hesitant to invest even in an era of historically cheap money and high returns on capital. Check.
PARTING THOUGHTS
By now you’re probably asking, well who’s right, the optimists or the pessimists?
Looking only at the data won’t give you the answer. Though we are getting better at measuring intangible investment, the available data are agnostic—they don’t provide any context. Growing intangible investment could be taken as evidence that businesses are creating huge new technological gains (the optimistic viewpoint)—or as evidence that businesses are simply better able to exert their monopoly power (the pessimistic viewpoint). But the corroborating evidence we’ve presented in this piece clearly favors the pessimists.
The implications for investors: If you believe in the optimistic case, then you will want to track the emergence of game-changing innovations, those most likely to generate a higher ROR, wider profit margins, and more rapid revenue growth for entire industries by way of scalability, synergies, and spillovers. If, however, you find the pessimistic case more persuasive, then you will want to track those particular firms which produce or acquire the most "ownable" IP and which can best use that IP to create, through pricing power, islands of privileged incumbency. Optimists will be more interested in the underlying technology. Pessimists will be more interested in the fine print and loopholes of each deal--and in any abrupt changes in the direction of public policy (patent law, antitrust, taxation) that could reshuffle the playing board overnight.
