Almost twenty years ago, in March 2000, I was sitting in a newly-opened branch of Starbucks near my home in London, reading a copy of the Financial Times.1 I was on gardening leave between jobs, serving out my notice from one employer before I could officially join the next – one of the largest investment banks in the world. I was very fortunate. Five years earlier, straight out of university I had become a stock analyst at a smaller firm, and now the market for stock analysts was hot. There was a sense of euphoria in the air that we were at the dawn of a new economic age, funded by continually rising stock markets and a population that wanted to participate. Stock analysts sat at the centre of it all, advising people which shares to buy and which to sell, and being paid handsomely by banks to create demand for companies to issue more shares.
I bit into my blueberry muffin and read the headline: “The NASDAQ composite index suffered its biggest loss in more than three weeks yesterday after a Japanese stock rout caused selling”. I skipped to the next story and wiped away some crumbs. No bell rang, but that was it, that was the top. The NASDAQ peaked on 10 March 2000 and went on to lose 78% of its value over the next four and a half years.
Three extremes
Twenty years on, I wonder if we are at another inflection point in markets. A fund manager I highly respect wrote in his latest quarterly investment letter, “After close to 30 years of investing in European equities, 2019 is set to become a pivotal year as a result of what we deem to be three levels of market ‘extremes’ which we observe in equities currently.”
The three extremes he cites are:
Style exposures. The divergence in performance between so-called ‘growth’ and ‘value’ stocks over the past several years has been widely discussed. ‘Growth’ has outperformed ‘value’ for ten years now and the divergence has only increased over recent months. In Europe according to research by Verdad Advisors value stocks underperformed growth stocks by 8.6 percentage points in the first half of 2019. The extent of that underperformance is now largely unprecedented in modern times. In Europe, the relative underperformance of value versus growth has not been as sustained since the early 1980’s. In the US, according to research by O’Shaughnessy Asset Management, investors have to go back to 1926-1941 to find a comparable period of sustained relative performance.
Valuation. Although some of the relative performance can be explained by divergent earnings prospects, some stocks have also become more expensive at the same time as others have become cheaper. At the country level, based on long-term Shiller-CAPE valuations, StarCapital Research finds the US market 41% more expensive than in the past, the German market fairly valued and Southern Europe and emerging markets undervalued. Citigroup estimates that the free cash-flow of UK companies is over 7% in aggregate and back to levels seen in 2008. The same trend holds at the sector level – three quarters of European banks currently trade at a discount to tangible book value. And it also holds at the stock level, with many economically-sensitive (i.e. value) stocks yielding double-digit cash returns and buying back their own stock.
Positioning. It is perhaps a tautology to say that investors have fled those pockets of the market that have underperformed and where valuations have contracted. But the magnitude of the outflows from some of those areas is significant. So far in 2019, European equity funds have suffered outflows at a higher rate than in any of the past ten years. In addition, investors are crowding around fewer trades. According to a study by Bank of America Merrill Lynch reported in the WSJ the overlap in the top 50 stockholdings between mutual funds and hedge funds stands at near-record levels. Facebook for example is held by over 55% of the funds in the survey, at a weighting over 80% higher than it is in the broader index. Other well owned stocks include Visa, Mastercard, Microsoft, Amazon, and PayPal. Meanwhile short interest in the FAANG stocks (Facebook, Apple, Amazon, Netflix, Google) is at historic lows.
Calendarised flows from Global investors into European equity funds (US$)
This extreme positioning has left unloved sectors of the market languishing. European banks now make up only 9% of the market, down from 16% as recently as 2016. The counterweight is the technology sector. The top 5 tech companies in the US currently make up 17% of the market (S&P500) up from 12% in 2016; the previous peak in 1999 was 15%.
Technological Revolutions and Financial Capital
In order to determine whether these extremes have created an instability in the market sufficient to cause them to unwind, it is necessary to map them to underlying fundamentals of the economy. O’Shaughnessy Asset Management recently published a research report Value Is Dead, Long Live Value, comparing the current environment to the period 1926-41 when extremes of style exposure/valuation were as prolonged. With reference to economist Carlota Perez’s work the report suggests that both periods coincide with a turning point in technological revolutions. In her book Technological Revolutions and Financial Capital, Perez identifies 1929-1943 as a turning point within the ‘fourth economic revolution’ around oil, autos and mass production. And in a 2017 blog post she identifies the period since 2008 as the turning point within the fifth revolution – the ICT revolution. In Perez’s model the turning point sits between the installation phase of a technology and its deployment phase. The installation period is “the initial turbulent time of Schumpeter’s ‘creative destruction’, in which the new technologies establish themselves in a massive competitive experiment that defines the winners and the losers in products and companies.” In the deployment phase “the new technologies spread their transformative power across the whole economy, producing what have been called the Golden Ages.”
The O’Shaughnessy Asset Management research suggests that it is during these turning points that growth will outperform value the keenest – when growth is highest and the eventual winners from the industry are established. In the turning point within the fourth economic revolution, the age of oils, automobiles and mass production, growth was heavily weighted towards manufacturing, with General Motors the biggest contributor as it cemented its position as the leading auto manufacturer in the US. The parallels with the leading technology companies of today is plain.
An additional feature of Carlota Perez’s model that is relevant in any survey of the current environment is its acceptance in the inevitability of bubbles as a result of over-investment. Those industries at the cutting edge of economic revolutions are prone to financial excess. This makes the current extremes evident in markets less surprising, but perhaps more difficult to call.
Bill Janeway, economist and venture capitalist builds on the theme. In his book Doing Capitalism in the Innovation Economy, he describes a ‘three player game’ in which the state, financial speculators and entrepreneurs coexist to oversee economic growth. In his chronology, “state investment in fundamental research induces financial speculation to fund construction of transformational technological infrastructure, whose exploitation, in turn, raises living standards for everyone dependent on the productivity of the market economy”. Thus, government, and specifically the US Department of Defence sponsored many of the technologies (silicon, software, the Internet) that combined to create the digital revolution; financial markets then took over to fund their exploitation. Critically in Janeway’s formulation, neither the state, which is driven by mission, nor financial speculators, driven by selling at a higher price, are concerned about visible economic value creation. “And so at each stage, the Innovation Economy depends on sources of funding that are decoupled from concern for economic return”.
This decoupling is necessary to further technological innovation since the process through which innovation happens is inherently inefficient. That process works by trial and error (“and error and error” Janeway adds), which means that it generates a lot of waste. Tolerating that waste is critical to advancement. Thus, rather than being an adverse side-effect, excessive financial speculation and the bubbles it inflates allows projects to be funded that otherwise might not. Janeway argues that financial bubbles mobilise capital at irrationally large scale.
There’s a certain feature of some technologies that makes them especially susceptible to bubbles. Those that harness network effects to drive winner-takes-all economics may find themselves in receipt of investment dollars with no upper bound. The railroads were one example; certain internet-based platform businesses are another. Venture capitalist Bill Gurley calls this “the single most challenging and difficult business strategy question … in my career”. He goes on: “Let’s say you’re in business X and your competitor raises US$0.5bn-US$1bn and decides that they’re going to loses US$150m a quarter. Now if you go back prior to 2010 Amazon lost US$1bn one year, and they were the only one that came anywhere close… and now this is happening all over the place. And you can sit back and say ‘Oh, well I’ll be fiscally prudent, I won’t do that’. You’re dead.”
Of course, bubbles tend to burst in the end. However, unless those bubbles are financed with debt, the pop doesn’t have to be all bad in Janeway’s model. He quotes Brad DeLong, who wrote in the context of the collapse of the TMT bubble in 2000: “Public markets paid for a build-out of the network infrastructure … Investors lost their money. We now get to use all their stuff.”
Countervailing forces: Regulation, Markets and the Capital Cycle
A number of events usually conspire to cause a bubble to pop. Some catalysts are discussed below. But one of them is heightened regulation. That big tech companies – whose technologies were engendered by the state through patents and through occasional financial incubation – are now challenging state authority is a power shift that exacerbates the prospect of regulation. Antitrust laws were established in the US in the turning point of Perez’s third economic revolution, the age of steel and heavy engineering. Many tech companies today are becoming a victim of their own size. In several key cases valuations require them to establish a natural monopoly, but maintaining a monopoly requires the support of stakeholders outside of their shareholder base, and interests there may not be aligned. No wonder Peter Thiel (in his book Zero to One) says that “monopolists … are incentivised to bend the truth”. Recent Senate hearings on Facebook and political calls to break up big tech are evidence of this trend.
Just as regulators provide a counterweight to how big tech companies can get, so do markets. That is after all how markets are supposed to work. The theme is explored well in the book Capital Returns, a collection of investor letters written by Marathon Asset Management LLP. In short, if an industry looks sufficiently attractive, new sources of capital will flood in, leading to greater capacity and lower eventual returns. Market valuations provide a reflection of available prospective returns. The process can take many years (the book points out that the NASDAQ started bubbling in 1995 yet didn’t burst until 2000) and the impact of ‘network effects’ can make it more difficult to discern. But it is hard to avoid. It is perhaps no coincidence that Disney decided to enter the video streaming market after it had seen US$180bn of market cap accrue to Netflix. By contrast many industries are being evacuated, with valuations and returns looking too low for marginal players to compete. In the semiconductor industry four players currently compete in DRAM, compared with over 20 in 1995, and over 10 in 2005. The pattern is an old one. The charts below show how capital investment in British rail networks in the nineteenth century (the second economic revolution) correlated with share prices.
As the influence of regulators and competitive response are showing early signs of strangling the ‘gilded age’ bubble in tech, there are parallel signs that we are entering Perez’s deployment period within the current economic revolution.
Janeway cites an OECD working paper, The Best versus the Rest which shows that since the end of the dotcom/telecom bubble, the “best” 5% of firms in terms of productivity across the developed world have diverged sharply from the others in both manufacturing and service industries. He looks back to prior economic revolutions to hypothesise that initially only some companies are able to afford the cost of integrating new technologies. Before electricity supply was standardised and widely distributed, manufacturing firms needed to install and manage their own generators. He compares that with today, when firms that want the benefits of computing historically had to hire their own IT departments and manage computer operations and application development. Yet as processing and storage become as commoditised as electricity, laggards should catch up. In this way, economic benefits will accrue to the consumers as well as to the producers of technology. The only caveat to this trend in Janeway’s view is that data may usurp software as the source of productivity edge, keeping the best firms at the forefront. However, although big tech companies have access to data, they don’t have a monopoly on it. It is the biggest firms in each industry that begin to reap advantage.
One industry where the impact could be seen is the financial services industry. The four largest banks in the US last year spent US$27bn on IT, equivalent to 8.5% of revenue. At end 2018 they had over 140 million digital customers between them. In the first phase the incremental spend required is a cost to remain competitive. However, as duplicate costs are eliminated, the return on the spend becomes more visible. JPMorgan has a target to capture US$1bn in annual run-rate savings from its 2018 tech investments, which will contribute to a ROI of over 2x. The benefits these banks have in terms of infrastructure and data versus start-ups is huge, and their ability now to use technology to strengthen their competitive advantage is very real. BBVA, one of the leaders in bank digitalisation, has shown how digitalisation has helped bolster client engagement. The creation of Swish as a mobile payment system by the six large banks in Sweden has shown how banks can compete with payment solutions offered by new entrants (60% of Swedes used Swish in 2018, up from 10% in 2014). It is no wonder that Sam Woods, Deputy Governor of the Bank of England, said in January 2019, “it is so far notable that no small bank has successfully become a large bank”.
Other industries that show a similar trend of their largest players utilising technology to enhance their competitive position include oil and airlines. In the oil industry BP has used technology to reduce the depletion rate of its oil fields. In the airline industry, Lufthansa now does 52% of its business via direct distribution, versus 30% in 2015. (Hat tip to another respected investor for these).
If not now, when?
While the trend outlined by Perez towards deployment of technology may be inevitable the timing is less so. And clearly in markets being early can be as damaging as being wrong. There are three more technical catalysts for the inflection point:
First is the interest rate backdrop. There is little doubt that low interest rates have reduced the required returns expected from stocks, and that long duration stocks such as tech companies have been the relative winners. On the flip side, financial companies’ earnings have suffered from lower rates – European banks have seen 10+ consecutive years of negative earnings revisions. The recent policy shift towards easing has reduced the risk that higher rates pose for tech stocks relative to value stocks. However, the vulnerability remains, and the relative performance since the shift in policy exacerbates it. This past week the yield on the 30-year German government bond went negative, and currently a record volume of government debt yields less than zero (US$14.52tr, equivalent to 26.3% of sovereign bonds). It requires contorted logic to make sense of this, and perhaps this is the true bubble that drives everything else. Although it looks less imminent now that the Fed has cut rates again, any turn here will likely have an impact on the relative valuations within the market.
Second, fundamentals will ultimately assert themselves, particularly now that many tech companies are subject to the scrutiny of public markets. Take Uber. Last year it generated revenue of US$11.2bn but was unable to turn an operating profit on that. In its quest to dominate its market the company subsidises rides. Net of driver incentives the company loses US$2 on every US$10 of gross bookings. Its US$70bn market cap is difficult to justify on the basis of its proprietary technology and reflects the upside of securing a monopoly in its markets and/or growing its reach. Yet both of these are hard – growing its reach, because of its high degree of concentration and the constraint of regulation (24% of bookings from 5 cities, and 15% of bookings from airport rides), and achieving a monopoly because of low switching costs for users.
Finally, the capital structure used to fund the current cohort of new companies makes valuations especially vulnerable. In the late 1990s investment in technology was financed in large part via the equity markets (and high yield). In the current wave, much of the investment is private. Hence the rise of the so-called unicorns. According to CB Insights, the number of unicorn companies – private companies with valuations in excess of US$1bn – is 384 with a combined valuation of US$1,198bn. Unlike public equity, private equity does not offer liquidity. On a mark-to-market basis it is conceivable that the valuation of these companies could be lower than US$1,198bn.
In addition, unlike in public markets, not all shares are necessarily valued the same. Liquidation preference rights bestow greater value on one class of shares than another and inject leverage into the capital structure. Of course, leverage is also present in its traditional form. Softbank is the largest venture capital investor in the world. It has just raised a new fund, Vision Fund II, with US$108bn of capital, of which US$38bn is Softbank’s own capital. The company deploys leverage to back some of its holdings. As at end March 2019 it had ¥5.4tr of net debt backing its holdings (US$50bn) in addition to some margin loans secured on some if its listed holdings.
Conclusion
The market is at an interesting point. Unlike 20 years ago when there was a sense of euphoria in the air, negative bond yields suggest more risk aversion. If we are indeed moving through Perez’s turning point into the deployment phase of the current technological wave, then the opportunities available in investment markets will pivot massively. Timing is uncertain, but such are the extremes in valuation that many ‘value’ companies are paying investors to wait.
- At the time I thought the branch had been opened in error. There weren’t many Starbucks in the UK, and here was one in a Zone 2 suburb of London. I assumed that an instruction had come in from Seattle to open a branch in London’s West End, but they’d got confused and opened in London’s West End Lane instead. Obviously, I was wrong. There are now nearly 1000 Starbucks in the UK and ~250 in London. And the share price is up over 20x.