Putting technological evolution in context:
What entrepreneurial capital does; why it succeeds and fails; and why that is relevant to energy technology Investing
I. Introduction. Venture Capital has a tendency to produce spectacular and unexpected returns and to spectacularly disappoint just when returns are expected. It remains a poorly understood science – one often described by its practitioners as “art.” Nonetheless the last 40 years of venture capital investing have produced their share of patterns and learnings. A closer look at those patterns suggests a new approach to investing in emerging energy technologies.
Technological evolution proceeds through a process that can be characterized as 1) explosive ideation, 2) competitive carnage, 3) adoptive inflection points, 4) insightful recombination, and 5) secondary expansions and network effects. These can be summarized as follows:
Explosive Ideation is the point at which a new idea, invention, product captures the imagination of a significant number of entrepreneurs, each of whom now believes they can perfect that idea and successfully take it to market, typically convincing groups of investors to back them in doing so. This tends to be the hardest of all of the stages because it involves the purest form of invention – intentionally solving a hard problem by attacking it directly.
Competitive Carnage is what happens when those entrepreneurs and investors realize perfecting that idea was more difficult, more expensive, and/or simply took longer than expected and the hoped for market adoption has not materialized as quickly as hoped. As a result, there is rampant competition for what little market acceptance there is. The result tends to be a killing off of the weaker competitors and a rapid lowering of prices, even as progress in perfecting the idea continues. Often, by the time this stage is completed only a handful of meaningful competitors remain.
Inflection Points represent the continuing phenomenon, particularly as regards significant “hardware” inventions, that once the new technology has been adopted by a couple of percentage points of the addressable population it tends to grow at exponential rates (i.e. CAGR’s of 50% plus) until it reaches approximately 70% or more of that addressable market.
Insightful Recombination occurs as this group of entrepreneurs and those from adjoining fields begin to look more closely at the range of ideas that resulted from the “explosive ideation” phase and realize that some of these concepts can be recombined both with each other and with technological advances in adjacent fields to more rapidly move the technology forward. This is often the financially most powerful of the stages because these recombinatory insights typically are of the variety that once demonstrated, are obvious to everyone, even if they were anything but obvious before one entrepreneur had that critical insight.
Secondary Expansions and Network Effects represent the overlay of new services and business models and applications software on top of the underlying hardware technology: Microsoft to the PC, iTunes to the iPhone, Netflix to the VCR, Social Networking to the Internet, etc. These both leverage the difficult work put into the underlying hardware technology and tend to add even more recombinative effects, thereby allowing for far more rapid growth trajectories than the underlying inventions they could not have existed without.
A single phase of this complete process can span 50 years or more. Therefore those in the middle of the new technological revolution often lack the experience of having been through the prior cycle and make faulty assumptions both about where they are in the process and just how fast the next process can move forward. Typically, they only remember the pace at which things progressed at the end, not the decades it took for the process to get started. It is in those misunderstandings that both investor opportunity and investor disappointment lie.
Specific to the issue of dealing with resource limitations, climate change, stranded fossil assets, solving water, food and other critical human needs, most of us were much more familiar with the pace of the IT revolution, the data communications revolution and the biotech revolution in the period between 1995 and 2010 than in the preceding 15 years, where the pace of change was much more deliberate, more capital intensive and more hardware and infrastructure than software and networking focused.
But we have now completed the first fifteen years of accelerated sustainability investments and we have again moved certain hardware and infrastructure technologies forward in tremendous ways that have, in turn, opened the door to software, business model, service and networking inventions that further accelerate the attractiveness and adoption of these technologies. Understanding just where that puts us in the timeline of the inventive process and which opportunities it leverages is key to making intelligent investments in the sector.
II. Explosive Ideation. “Creative destruction” is another term for this part of the process; first coined by economist Joseph Schumpeter in 1942. In fact, Schumpeter referred to capitalism as “the perennial gale of creative destruction;” seeing "deliberate disruption" as the key to "transformational growth" by both individuals and organizations.
Nature fosters innovation through diversification (mutations) to create new species and variants of existing species able to survive the forces that challenge the existing forms of life. Charles Darwin was one of the first to recognize and describe this natural phenomenon. Nature also represents forces that allow it to correct for overpopulation or overabundance. Each crisis leads to a new beginning and an opportunity to do it better this time around – the natural process we refer to as evolution. Darwin observed that amazingly, even the simplest creatures respond to such natural and ecological threats by multiplication and diversification – somehow intrinsically understanding that change is a natural ingredient of survival. We alone as humans, tend to cling to the old order when faced with existential threats.
Even though it is an economic and business process, when analyzed and mapped, it more closely resembles how the brain works through a series of neural networks to expand upon initial thoughts and continuously combine those with other expansions of adjacent thinking (see below):
The larger the problem, the greater the opportunity to use a great diversity of ideas and approaches to find a possible solution. By definition, change creates market gaps or opportunities that can be seized by this process of explosive ideation.
Unfortunately, the initial problems often don’t lend themselves to instantaneous solutions. As a result inventions such as the printing press, the steam engine, the automobile, the personal computer, the telephone, the television or even solar panels are often more the result of an accidental thought that is then carefully nurtured into a larger inventive process, often by far more than the person who first came up with the idea. The problem is that commanding such disruptive inventions to occur upon demand is unlikely to produce positive returns.
More often, we are good at continuously improving upon the first implementation of such a breakthrough concept, continuously refining it until it fully meets market needs and is rapidly adopted by users. Each of these discovery, refinement, adoption and maturation processes can be depicted as an S-curve in that they start slowly, reach an inflection point, grow rapidly to a certain level of saturation and then slow down. As a result, technological process is often depicted as a consecutive series of these S-curves (see chart below):
SOURCE: 2008 California Green Innovation Index
But depictions like the chart above, ignore what happens between the flattening of one S-curve and the start of the next, and, more importantly, how difficult it is to start an entirely new set of curves from scratch.
Most of us are intimately familiar with the concept of the Gartner Hype Cycle (see figure below). But Gartner speaks more to market adoption trends and investor excitement than the underlying realities of the creative destruction process that is occurring amongst the companies vying for leadership in a given market segment.
For an investor, the ability to make a small investment at the point of “technology trigger” and to then liquidate that investment at the “peak of inflated expectations,” represents the optimal seed and venture capital investment model. If that investor actually has to wait until the “plateau of productivity” for liquidity, the cash on cash returns may still be attractive, but the IRR has suffered meaningfully. Worse, if during the “trough of disillusionment” the development of that technology required significant additional dilutive capital, then virtually all gains may be erased for early investors, particularly if they don’t have the deep pockets to retain their ownership share as the company proceeds through the Gartner stages.
III. Competitive Carnage. We see the “trough of illusionment” in a different light. We see it as the point at which the enthusiasm built up during the “peak of inflated expectations” realizes the market simply isn’t ready to adopt the new technology at the pace needed to support the hundreds of companies built and funded to exploit that technology.
As a result, they begin to compete very aggressively for what little market adoption there is, lowering prices and driving each other out of business. As they do so, the lower prices begin to attract more early adopters, but the market typically turns upward too late to save most of the companies first created to exploit it. In fact, typically only a handful of companies remain viable to benefit from the rapid growth phase of the new market.
So instead of a picture that looks like the chart below (“A Next Wave?”), we get a series of curves that look more like a Gartner curve but depict two very different things, first, the number of competitors in the market and second, the rate of product adoption.
Here we see how a fundamental technological breakthrough tends to spur an outburst of company formations (the “Explosive Ideation” phase), seeking to perfect that initial invention and effectively bring it to market, followed by the competitive decline forced by a not yet ready market (the “Competitive Carnage” phase).
By focusing on A (the number of years it takes between reduction to practice of the critical invention, the height of entrepreneurial enthusiasm and company building and the end of the competitive crush that inevitably follows), B (the number of companies created to pursue the perfection of that fundamental invention), C (the point at which the new technology has actually been adopted by 1% of the U.S. population (or relevant target market) and D (the time it takes to get from 1% adoption to approximately 70% market penetration (where adoption often begins to level off), we can compare a series of such inventive cycles around technologies now familiar to us and begin to predict where other fundamental technologies are on this same development scale.
Here, for example, is the set of curves and points A, B, C and d applicable to the automobile:
Note that in this instance both World War Two and The Great Depression slowed the automobile’s progression toward 70% or greater market penetration. Importantly, we today tend to forget the important part of automotive history that was the roughly 250 car companies that competed alongside Ford, General Motors, and Chrysler for domination of this nascent market. In fact, in the early years it was far from clear who would dominate and in fact Chrysler didn’t enter the picture until 25 years after the automotive “gold rush” began.
As importantly, the 1% market penetration point occurred only after most of the competitive carnage was over and the field had been reduced to a relatively small number of competitors. In all, the rise and fall of the entrepreneurial wave of automobiles took about 35 years to run its course. But once that 35 years was over, the growth rate in car sales was exponential and the few winning companies, represented extremely attractive long-term investment opportunities, even if they did anything but that during the prior 30 years. Further, for the investor “playing the field” the average returns would have been miserable, as the vast majority of the competitors failed. Although we have not studied it in detail across technologies, our initial research indicates that the winners tend to represent odds of 5-in-100 or less and, although over the long run they represent outstanding investments, the IRR for investors betting on those same companies after the competitive carnage is over (i.e. in 1912) is far better than that for those who actually picked out the Ford Motor company back in 1903.
More recent history for most of us is the development of the personal computer (although for many of today’s investors their memory only covers the subsequent software and internet periods, not the more difficult hardware development phase of information technology).
A few investors were prescient enough to bet on Apple Computer by 1980, but the truth is that you really could have waited all the way until 1985 and still have gotten most of Apple’s computer gains (and until 2006 to get in ahead of Apple’s next technological breakthrough – the iPhone). Once again, betting the field on personal computer hardware manufacturers would have produced very disappointing overall results during the 1975 to 1985 time period.
Generally, the difficulty with most of these hardware breakthroughs is that they require significant amounts of capital to a) reduce the technology to practice, b) build a factory to scale up production and c) produce in enough volume to bring prices down to high levels of market acceptance.
Of greater relevance to the “6th wave” discussion on clean energy, sustainability and green chemistries, is the same set of curves as applied to the solar industry (see below):
For solar, we are nearing the completion of the competitive carnage phase. We have seen massive drops in production costs and, as a result, panel costs (see chart below). Solar PV now represents more than 1% of US energy capacity, but has already crossed the 10% penetration levels of eligible households in several states (like Hawaii, California, Massachusetts, Arizona and Colorado) that were early adopters. Conforming to the historical pattern, investing across the field of PV manufacturers was, on average, a very poor bet. Interestingly, one investor, Europe’s Good Energies turned a roughly $400M initial investment across a range of German, U.S. and Chinese solar companies into a roughly $6B paper valuation at the height of the competitive frenzy, but (in large part because liquidity was unavailable) rode those valuations back down to less than $1B by the time the carnage was over.
Nonetheless, today’s First Solar and SunPower probably represent the investment equivalents of Ford or Chrysler, Dell or Apple in terms of their long term prospects as the solar industry continues its long climb toward 70% penetration of eligible households.
What about the rest of the so-called “CleanTech” industry? Wind has already undergone its boom, bust and recovery phases, but it is largely a business-to-business or utility-scale solution and thus didn’t generate the same competitive frenzy as most consumer product technologies (of which solar PV is really the first energy entrant). As we have seen, biofuels and biochemical are well into their competitive carnage phase and even if we have seen the end of the carnage, there are few new entrants in those fields at the moment. On the other hand, battery chemistries are all the rage today and we may still be on the upswing of new company creation. Will overbuilding of lithium ion factories for car batteries in the face of much lower oil prices spur the beginnings of competitive carnage in batteries; that will rapidly bring down prices and, as a result, speed up adoption?
This pattern has repeated itself over and over again with regard to a wide-ranging series of technological hardware breakthroughs (including the telephone, radio, washing machine, microwave, VCR, color television, etc.), each having somewhat differing levels of entrepreneurial gold rushes, competitive carnages and years from start to peak to trough. But the pattern of proceeding rapidly from those first percentage points of market adoption up to nearly 70% of households generally holds true across this range of consumer goods (see chart below).
Another notable factor with respect to the automobile, personal computer and solar PV industries is that they represent a paradigm change in a) personal mobility, b) personal information access and c) personal energy production as those compare to their legacy counterparts of a) buses and trains, b) enterprise computing and c) utility scale energy. The parallels are further extended when you factor in the importance of storage (the gasoline tank and the gasoline station, computer memory and cloud storage, and energy storage whether in the form of batteries or the grid) and a supporting network (the road and highway system, the Ethernet and Internet, and the “Smart Grid”). The power of the new consumer technology, affordable storage and networking technologies together represent a new era in energy as much as cars and personal computers changed mobility and information technology.
Inflection Points. History shows us that with these “hardware” technologies, once they are adopted by just a few percentage points of the applicable population, they very rapidly grow (at exponential rates) until they have been adopted by somewhere between 70 and 90 percent of the population.
What these curves generally don’t reflect particularly well, is how many years a new technology crept along at the less than 1% adoption level. Pure math tells us that with 100% per year growth rates (exponential doubling), it takes equally long to get from 0% to 1% as it takes to get from 1% to 100%. As a result, investors are often ready to give up just when the noticeable inflection point of more than 1% adoption really happens. As a result, knowing when to invest, how long you are likely going to need to hold that investment and how much capital you might need to protect your position through liquidity become critical success criteria (see chart below):
Investing early in these breakthrough hardware technologies is neither for the faint of heart nor for the shallow of pocket book. On the other hand, waiting too long and not investing in the leading companies once they have emerged from the competitive carnage means leaving large sums of money on the table.
For incumbent corporations being disintermediated by these changes, the situation can be even more severe, as they cannot expect to successfully acquire the winning technology once that player has established a dominant position, yet their organization is structurally handicapped with respect to aggressively internally developing a winning response to an emerging disintermediating threat. We have previously published a “One Pager” on this challenge – see “A Repeating Growth Pattern”
The good news for most investors is that fundamental hardware breakthroughs are just the first part of the story. For each such breakthrough there follows a succession of new business models, service industries, and software applications and solutions that are enabled by the hardware/infrastructure breakthrough and that generally represent far shorter and therefore more attractive paths to economic returns from investment.
Insightful Recombination. The other very important pattern of entrepreneurial behavior enabled by these technology invention and adoption curves is the ability to combine technologies across markets together with new concepts to address entirely new market opportunities. Good examples of this type of recombinant invention for the resource productivity sector are 1) Tesla Motors, Nest Labs, Solar City, and 4) Uber.
Tesla involved taking advantage of technological progress in the electric motor segment, the lithium ion battery segment and the parallel processing computing segment and applying the best available technologies to an entirely new segment – automotive transport. Ultimately building first the tesla Roadster and then the brilliant Model S were still daunting challenges, but imagine if Tesla had had to develop its own engine, its own battery and a novel way to process charging and discharging all by itself. Instead, it leveraged the work of others by applying it to an entirely new market segment.
Similarly, Nest Labs took the known capabilities of a thermostat and overlaid the learnings of having built the iPhone (as well as its global supply chain) to rethinking what a thermostat might do. In so doing, Nest combined the best attributes of two industries that hadn’t yet fully intersected on their own. Solar City didn’t develop its own PV modules. It relied on the hard work of others to do that. But it borrowed business models from the housing industry to rethink rooftop installation, lending models from banking to provide innovative financing tools, google Maps to provide a quick look at rooftop eligibility and innovative marketing models to change how energy is sold to a consumer market.
Uber, in turn, has completely disrupted the Taxi industry by borrowing a series of Internet inventions developed by others (Google maps again, routing software algorithms, handheld devices and the ability to use those devices for billing and payment, and a series of other tools developed for the Internet industry, but applied them to a business theretofore not disrupted by these technologies.
In retrospect, each of these companies’ approaches seem rather obvious, but they obviously didn’t occur to the many, and those to whom they did occur rapidly leveraged the preexisting technologies into market dominance.
IV. Secondary Expansions and Network Effects represent the overlay of new services and business models and applications software on top of the underlying hardware technology: Microsoft to the PC, iTunes to the iPhone, Netflix to the VCR, Social Networking to the Internet, etc. These both leverage the difficult work put into the underlying hardware technology and tend to add even more recombinative effects, thereby allowing for far more rapid growth trajectories than the underlying inventions they could not have existed without.
Unlike the prior category, most of these businesses involve purely software or service models and are thus even easier to build and market test than the foregoing category of recombinant insights. Much of what we have seen in social networking and online business models represents this phase of the technological evolutionary process. It is also often the last to occur as it seeks out smaller niche markets that can rapidly be tested, exploited and then expanded upon. From an investment perspective they tend to both be “capital light” and have fairly short maturation times, meaning one can “fail quickly and cheaply” and use capital to consolidate success, often producing stellar rates of IRR. But remember, that these business largely have arisen only in the last 5-10 years of the IT revolution and would not have succeeded or even been possible in the first 20-25 years of that IT revolution.
Understanding these patterns of failure, challenge and success across the range of investment opportunities that occur in a new wave of innovation is a critical aspect of investment success. Although complex, we believe the chart below sums up the steps companies need to take on their way to an outcome and thus explains some of the risks involved and the reasons why investment returns are harder to achieve along some paths as compared to others:
Most of the Internet investments that have produced enormous venture returns over the last five years represent companies that are able to proceed along the far right hand side of the chart. A few hardware companies in the “CleanTech” sector, notably Tesla and Nest Labs have leveraged third party hardware advancements and manufacturing partnerships to successfully travel up the middle path. For those progressing from the bottom left, it has been a slow, arduous and challenging journey and has rewarded very few investors.
Therefore, from an investor’s perspective, deciding where and when to invest depends heavily on where a particular company started its journey and how far it has progressed. For the most part, companies on the bottom left of the chart remain very challenging investments. For many of them, successfully gaining access to the horizontal bar of infrastructure/project finance capital becomes a key determinant of the future success. At the same time, each group of companies that reaches the top right from the bottom middle and left of the chart opens the door to an entire generation of new software-based applications and solutions companies that can travel up the right hand side of the chart and thus represent much more attractive private equity and venture returns.
What does that mean for investment prospects in the clean energy arena? In our view it means that one should pay significant attention to the maturity level of the underlying technology. It also means that there are later stage and/or publicly traded companies that represent the scarce survivors of the competitive carnage and that history suggests will represent 20-30 year-long attractive investment holdings that no matter how expensive they appear today, will if they remain a category leader, be yet far more valuable over those many years.
We also see a very attractive convergence between four technology trends that are all maturing rapidly, two of them already having completed most of their shakeout phase. Those four technologies are (i) solar PV, (ii) electric vehicles, (iii) batteries for PV and EV storage; and (iv) the overlay of big data analytics and the internet of things on the electron supply and management to and from PV, the EVs, households, businesses and the grid. The intersection of these four is spawning a host of new sectors, many characterized more as applications and solutions businesses, which we believe represent the equivalent of the 1995 to 200 wave of information technology investment. In particular, we see the following subsectors as of particular interest:
As shown in the chart, many of them will benefit from an overlay of infrastructure finance to facilitate their more rapid deployment into the market. These, in turn, will represent attractive yield opportunities for investors who understand the underlying technology and deployment risks and support the companies progressing their technologies to ever-greater market acceptance.
The foregoing chart omits much of what has been characterized as the “CleanTech” investment landscape. We believe that as success models are borne out in these sectors, other sectors will both mature their fundamental technologies and will benefit from the success patterns developed in these sectors.
As a result, we remain very much focused on the places where we believe investment returns can be produced today, both in terms of category leaders and the new applications and solutions sectors they have enabled.