There is a constant battle between perception and financial necessity. Any hint of failure will be self-fulfilling, in that it will immediately shut down access to capital. To survive, the company must always be bullish about its prospects of success. As capital markets are influenced by the general level of optimism, success is also affected by the prevailing economic and monetary conditions.
In figure 11.1, the stages of development outlined above are presented schematically. What can be seen are the points at which funding is typically required. These represent potential crisis points for the company, since failure to raise the necessary funds means the game is over. Typically the level of funding rises for each successive stage, as the technology moves from feasibility demonstration to commercial deployment.
This means confidence levels also have to be raised, as increasing numbers of external investors need to be pulled in. Historically even ultimately successful companies have often been forced at some stage to offer to sell their technology in desperation at their inability to raise capital, only to scrape through to ultimate profitability. For example, Bell was forced to offer his patents to Western Union for $100,000 and was lucky that his offer was turned down. There was a point in Google’s history when it badly needed development capital and came close to surrendering control over its future development. We can only speculate how differently the shape of an industry might have developed had a different outcome occurred in these cases.
The terms investors can obtain also depend upon the stage of investment and overall market conditions. The earlier an investment, the greater the equity that is typically received, simply because there is a greater number of hurdles to be overcome on the way to profitability. In some cases, the process telescopes and profits appear quickly, reducing the number of necessary funding rounds. This mainly happens where a company succeeds in obtaining patent protection over the technology, although even this is not always sufficient to ensure success. If there are high capital costs, access to capital becomes of at least equal importance. Edison, AT&T and the computer manufacturers all found this out. In the Internet boom, exceptionally, the opposite happened. Easy capital and availability of funds meant that companies were able to raise extraordinarily large sums purely on the promise of demonstrating feasibility. That sustained them well beyond the first phase of the typical development cycle – but did not in itself guarantee their ultimate success. The importance of retaining investors’ confidence returned with a vengeance when the day of reckoning could no longer be postponed.
The economic impact
The economic impact of technological change is distinctly mixed. In the short term it is positive in the sense that it typically contributes to higher economic growth, as part of the invested capital devoted to the technology is effectively converted into consumption. This tends to bolster the feelings of wellbeing which surround strong capital markets. However, in due course the new businesses reach the point where their cash burn rate can no longer be sustained and new capital is no longer forthcoming. In aggregate, the new businesses contribute little to the stock of productive capital because of duplication, the costs of raising capital and the costs of maintaining necessary propaganda.
In the medium term, the impact is more likely to be negative than positive. The destruction of capital that follows a period of over-investment creates the need for a reduction in consumption and a rise in the savings rate to repair the stock of available capital for future funding. Survivor companies refund and retrench such that their overall contribution to economic growth is negligible, and is counterbalanced by their competitive impact on existing companies.
As a side effect, inflation tends to be lowered, although it is the reaction of the monetary authorities that determines whether the ultimate outcome is inflation or deflation. In most cases, excess capacity is created by the funding of new infrastructure, and this has to be written down or depleted before the economic impact can be finally measured. Employment patterns meanwhile have to change, which creates problems of adjustment for those parts of the labour market most adversely affected.
In the long term, however, the process is generally positive for the economy. New technology delivers more efficient production and distribution, raising growth rates and the standard of living. More than 60 years ago, Joseph Schumpeter described how a process of ‘creative destruction’ enabled the capitalist system to regenerate itself at periodic intervals. Technological change is the handmaiden of that process, and there is nothing planned or pre-determined about its effects, judging by the history of previous episodes.
The Internet and the technology cycle
Anyone who followed the saga of the Internet described in chapter 10 will recognise many of its characteristics in the technology cycle. Stage 1 grew quite slowly out of the embers of the personal computer funding boom, and was slightly unusual in that the basic research and much of the subsequent infrastructure funding was provided by the government. This set the scene for slow but accelerating incremental improvements in Internet connectivity and access. The public funding of large-scale networking created ‘end user’ demand in the academic communities. A series of technical problems had to be solved before the potential had any possibility of becoming commercial reality. But as connectivity and access spread, new private sector entrants were able to look for profits from the opportunities that the Internet provided. Many were initially drawn towards developing software, where the cost of entry was significantly lower. An extended period of feasibility and prototyping followed.
Along the way a number of individuals and groups created solutions with commercial potential. The more far-sighted venture capitalists placed small sums with a number of these embryonic companies. Success did not come overnight – and many failures occurred along the way – but eventually the various pieces of the jigsaw came together to produce an Internet that was fit for commercial exploitation on a global scale. Rapid traffic and user growth followed. Once this became apparent to the public markets, unprecedented levels of venture capital funding and an IPO boom followed, made possible by a low-interest-rate regime and the willing suspension of disbelief by investors, based upon the promise of ‘sure’ returns in the short term. As we now know, this second stage ended abruptly in early 2000 as the interest-rate regime shifted upwards and reality began to intrude.
The duration and magnitude of stage 2 was supported by the benign economic environment that surrounded it. What is unusual about the TMT/Internet bubble, compared to many historical precedents, was that for many companies it was possible to raise funds without going through the early stages in the standard technology cycle. At the peak, companies did not have to demonstrate prototypes, viability or deployment to an unquestioning outside world. They did not have to achieve profitability or a robust cash flow position. In many cases, they did not even have to forecast the achievement of such a position within a meaningful time frame. This put them in a position where the ‘cash crunches’ that periodically occur as a technology moves through the various stages of the cycle were postponed.
In the absence of traditonal cash-crunch tests it became easier for valuation metrics to migrate away from their historic parameters, which they did with a vengeance, as described in chapter 10. In normal periods investors typically require to see at least the promise of a dividend stream from the majority of their investments. The ability to return cash has always been seen as an important measure of a company’s health. Even before the Internet bubble, however, this had started to change, largely for tax reasons. Share buybacks became more popular with both investors and corporations, which increasingly looked to replace dividends as the favoured form of distribution to shareholders. Changes to corporate incentive schemes, and in particularly the widespread adoption of share options as management incentives, further encouraged this change.
Nevertheless the Internet bubble was notable for taking the trend away from dividends and cash flow as a measure of corporate performance to new extr
emes. If one monitors published recommendations on securities during periods of technological change one can often observe a shifting of valuation parameters over time. As valuations become impossible to sustain against one benchmark, analysts move on to another more ‘acceptable’ benchmark. This reflects their natural bias towards optimism, based on the empirical fact that in the long run equity markets provide the strongest returns of any asset class. The Internet bubble was a textbook example of this process at work.
11.3 – Bubble valuation anatomy: from fundamentals to concepts
What this migration in valuation techniques actually demonstrated was the ever-escalating risk of investing in companies whose prospects of generating cash flow and profits were receding into the future. The further one goes down this path, the closer in form investing becomes to gambling. Just as with gambling, the only person consistently to win from this game is the ‘house’ – in this case represented by the investment banks that provided a seemingly inexhaustible supply of new issues to satisfy investor demand and the brokers who earned fees by pushing those stocks on to their clients. Those who saw the emerging Internet bubble as a chance to make some easy money had to learn this timeless investment lesson for themselves.
The market impact of the Internet bubble
By the time the first edition of the book was published in 2001 we had clearly entered stage 3 of the technology cycle, the period when investor scepticism is rising and bringing into question the viability of many businesses that have hitherto seen few barriers to fundraising. As funds dried up, all the traditional valuation metrics that had been ignored up to that point came back with a vengeance – and returned real power to companies that had either raised funds already or had genuinine self-financing businesses. This, combined with rising capital costs and general investor unease, created the conditions under which the valuation edifice collapsed.
History strongly suggested that it would take many years for the excesses in stock markets and the levels of valuation seen in 1998–2000 to unwind, and so it proved. Most leading stock markets fell by around 50% when the bubble burst and the market did not start to revive for three years. Having briefly matched its 2000 peak in 2007, the S&P 500 index only decisively surpassed the level it reached at the height of the bubble in 2013. The Nasdaq index, where many of the new economy stocks were listed, did not reach its 2000 peak again until 2015. In real, inflation-adjusted terms, most stock markets took at least a decade to recover their losses.
11.4 – Hangovers can last a while
Source: Thomson Reuters Datastream.
It is true that you cannot attribute all this performance to the fallout from the Internet bubble. The global financial crisis of 2007–08 killed off the new bull market in equities that began in 2003 and was close to regaining the former highs when it was aborted, prompting a second severe bear market. That crisis had many roots, however, at least one of which – the easy monetary policy adopted by the Federal Reserve and other central banks – can be attributed in part to the fallout from the earlier Internet bubble.
In the bond market, long-term yields declined steadily after 2000. For some this was the consequence of a number of global deflationary forces continuing to weigh on the fixed-interest markets. One of those forces, as noted earlier, was the destructive side effect of technological advance on established business models, a factor which was both explicit and widely forecast in the case of the Internet and the coincident changes in telecommunications and social behaviour. A great many things became cheaper. A huge amount of capital was also destroyed in the decade. This effect was partially masked for a while by the extraordinary explosion in credit that preceded and helped to cause the banking crisis of 2007–08.
The shocking events of 9/11 rightly traumatised much of the world but arguably served to alter the path of economic policy. From a position of raising interest rates to moderate economic growth, the monetary taps were turned on both in the banking sector and the non-bank sector, leading to an explosion of securitised debt. The normal economic consequences would have been inflation which, in turn, would have eventually been met with monetary tightening. However, the backdrop of the efficiencies brought about by the Internet, together with a flood of new incremental supply from the reintegration of China and India into global markets, meant that there was little obvious sign of inflation. As a consequence the main inflation which was sustained was an inflation in asset prices, while debt continued to balloon to high and unsustainable levels.
The misallocation of capital to telecoms
Contrary to the experience of earlier bubbles, the Internet itself has not produced meaningful amounts of excess capacity. Because of its very different history, having been financed largely out of the US public purse for so many years, it did not inspire a string of speculative ventures raising capital and laying lines in the hope of future returns, as was seen in the railway age. In that respect, the early stages of the Internet had much more in common with the early years of the telephone, when the Bell Companies retained their monopoly and had no need to lay excess lines to squeeze out competitors and forestall new entrants. The capital raised for Internet businesses effectively gave a swathe of new entrants sufficient funding to go on paying their running costs for many years, well beyond the normal time horizon at which innovators historically faced cash flow shortages. It produced instead a wild competitive race to achieve first-mover advantage, a trend which effectively transformed capital into current consumption rather than long-term sources of growth. Many of these ventures proved to have short lifespans.
It was a different, but related, story in telecommunications, where massive amounts of capital flowed into building both fibre optic and wireless networks to support the expected growth of a ‘connected world’, a phenomenon that the Internet was rightly expected to help materialise. Much of the capital in this case was diverted into the hands of governments in the form of wireless licence auction proceeds. An auction of third generation (3g) mobile phone licences in the UK and Europe resulted in $150bn being paid to various European governments in auction proceeds, simply for the right to operate. The same amount again was raised to build the 3g infrastructure before the licences could be used and any revenue earned. Total global capital expenditure for the telecoms sector in 2000 alone was $243bn, a figure two and a half times greater than that invested five years previously. Within this figure the largest elements were Internet-related, either for broadband access, optical networking or wireless infrastructure.
You could argue that, far from empowering a ‘new’ economy, what followed was a classic Keynesian make-work project, paying people to dig useless holes in the ground. Many of the telecom companies whose shares came to the public markets in the form of IPOs did little for productivity or product innovation. It took years for the substantial communications infrastructure put in place to be fully utilised and for many companies it never produced a meaningful rate of return. That is one reason why a number of prominent ventures that commanded huge market valuations for a while, such as Global Crossing, eventually went bust.
How does this level of fundraising and capital expenditure compare with other examples in our historical survey? In today’s money the railway industry in its peak year raised roughly $250bn in authorised capital (although only a proportion was actually spent on infrastructure). The telecoms investment spend was therefore similar in magnitude. Or, looking at it another way, the $250bn invested in one year was roughly equivalent to the GDP of Russia! Was this huge investment in the development of the Internet and commuications infrastructure wisely spent? How long would it take to generate sufficient revenues to justify the investment? When would it feed through into a positive increment to economic growth? These were the questions that should have been posed at the time.
Over a decade and a half later the answers are painfully and predictably clear. For the many investors drawn to the telecoms sector by the promise of returns from the brave new interconnected world, the out
come in aggregate has been dismal. Shares in the sector fell nearly 60% from their peak, and although there has been a long and slow recovery since, they remained 40% lower in 2017 than they were then. It is true that profits have declined as the overspend washed its way through the system and competition became more intense, but the main cause was inflated expectations rather than poor profits growth. As in previous periods, the valuations that prevailed during the mania phase were simply detached from reality.
Where we are today
The process of rationalisation and refinancing that has historically been the central feature of stage 4 in the technology cycle has been ongoing for well over a decade, while the Internet age has moved into its final consolidation phase. We can see those companies that have emerged as long-term winners, including second-generation adapters of the new technology, and the many more who have either faltered, been taken over or simply failed. Companies such as Apple and Amazon have survived and profited hugely from the second order effects of the disruptive new technology, enabling them to achieve – at least for now – dominant global market positions in their specialist areas of smartphones and online retailing. They have been joined by Google and Facebook, which have given away services to harvest vital consumer information that can be monetised by helping advertisers and retailers better target their customers.
Engines That Move Markets (2nd Ed) Page 62