Year-end thoughts: Solar and the long tail
First and foremost, we pretty much have to start with solar.
According to numbers from an AP/ Dow Jones report (with figures from the NVCA), solar VC investments reached $67.7M in the first three quarters of 2005, as compared to $31.4M for all of 2004. The AP and Dow Jones also report that solar made up more than a third of all energy VC investments so far this year. Regardless of the specifics of the numbers, solar is without a doubt the hottest cleantech investment area right now (if you'll pardon the pun).
Relative to the billions-of-dollars-per-year market that solar power is fast becoming, the amount being invested is still a small amount. But it's important to recognize that almost every investment in solar technology (with some rare exceptions) is someone making an investment in a proprietary generation technology, all from the same basic source of energy - the sun. A natural assumption would be that for any given energy-generation situation (ie: rooftop panels), there will be one single most economic technology. With some solar investors declaring that they are "looking for the next Google," it is tempting to think of solar technology as a zero-sum game, with one winning technology (perhaps already invented, perhaps not) and a lot of lesser approaches that will either be subsumed or that will outright fail.
Consider that this year there were at least 10 different solar-technology venture fundings (probably a lot more, just did a quick scan), plus other IPOs, plus the presence of well-established major players such as GE and BP Solar, etc. If solar is indeed a zero-sum game, it makes it a bit tough to understand the level of investment in solar right now. Is there going to be one "home run" and a lot of money-losers in the segment? Is there a bubble in solar?
First, let's examine the idea that solar is truly a winner-take-all market. We tend to think of solar as being a uniform market -- solar panels on rooftops, whether commercial or residential, whether new-build or retrofit. But in fact, there are a number of differentiating factors that can mean different technologies are best-suited for different niches.
For example, the standard economic measurement of efficiency for solar technologies is cost per peak watt. One would assume that the "winning" technology would simply be whatever achieves the lowest cost on that measure. But installation and application are key variables that can affect the answer, and in fact create different needs.
For a residential system, the end user often has a relatively small energy load, is usually unable to feed excess power back into the grid, and doesn't have storage. Thus, the load is the limiting factor for the size of the system, and there is often more than enough space on the rooftop for a number of different technologies to be used, including technologies that achieve low cost per peak watt, but also lower amounts of energy generated per square foot. In other words, it can sometimes be more attractive to go with a cost-efficient technology that requires a lot of square footage, because you have the space.
However, in speaking with installers who target large commercial buildings, you often hear a different story -- the limiting factor of the system isn't the load, it's the available roof space. In which case, these installers argue, it's often best to go with a higher cost-per-peak-watt system if you can squeeze more watts on a roof (this can also be affected by the specific form of any subsidies).
The real-life trade-off here, currently, is thin-film versus high-end silicon-based systems. No unconcentrated technologies beat high-end silicon in terms of watts per square foot. But many emerging thin-film technologies, because they avoid the use of silicon, can achieve significantly lower costs per peak watt. Which one will be best-suited will thus depend upon the specifics of the situation.
And so we can already see that the market environment which may seem monolithic in fact has different niches that different technologies can best address. Other similar market schisms are created by trade-offs such as:
- Centralized grid power vs. decentralized off-grid power
- Building-mounted versus other uses (portable systems, electronics, LED lighting, etc.)
- Cost of system versus lifetime of system (for your rooftop system you probably want at least 20 years of warranteed lifetime, but for your solar-powered radio you can probably be happy with a more rapidly-degrading 5-year lifespan if it means lower cost)
- Form of mounting (rooftop panels vs. concentrators vs. BIPV)
- Power only or co-generation (ie: solar heating and/or hot water)
Perhaps one way to think about this is to borrow the much-overused phrase from other venture investing areas, the "long tail," which is really just another reincarnation of the old 80/20 rule. The concept here is that, if one is to think about an overall market for solar, there will be some large key segments that make up the lion's share of the market (e.g., 20% of niches make up 80% of total market demand), and a multitude of much smaller niches that fill out the rest of the market (e.g., the other 80% of niches make up 20% of total market demand). True long-tail thinking says that the ratio may actually be 80/30 or 80/40 (and available market data suggests that for the solar industry it's probably pretty concentrated in the big grid-connected rooftop markets, so our original guess of 80/20 may be most appropriate), but the basic concept still holds regardless.
In other technology markets, the idea is to invest in technologies that enable efficiently tapping into the long-tail market opportunities (ie: eBay, Amazon.com, etc.), because the larger market niches have already been somewhat saturated, but if you can aggregate all those individually tiny niches you can add up to a pretty big market as well. The same phenomenon appears to be happening in solar, albeit at an early disaggregated stage.
Secondly, the other factor at play is the timing question: What will win today, versus what will win in ten years. Silicon-based technologies are well-understood and well-warranteed. But as silicon prices continue to rise along with demand, and emerging second- and third-generation PV technologies continue to mature, will non-silicon technologies achieve pre-eminence in the coming years? Investors are playing both sides of this question.
Finally, even within specific niches and time-to-market windows, there is always room for more than one player to potentially succeed -- in some niches it will be winner-take-all, and (more often) in other niches there will be several permanent players. The management team's ability to execute thus becomes key -- technological advantage is not enough.
This is all a long way of getting to a very basic conclusion -- that there is a lot of room for a lot of different bets in solar, it will remain a very un-concentrated market for some time. Silicon-based solar saw a wave of investments a few years ago, recently thin-film technologies have seen a lot of bets, and perhaps third-generation nanotechnology-based PV and solar concentrators will see the next wave, but the biggest conclusion is that there is room for all -- at least for now.
There are some other conclusions investors should draw as well:
- Investors need to be aware of whether the technology they are backing is an "80%" technology or a "20%" technology. Does the technology only address specific niches in the long-tail, or in the main market? Is it a platform technology that can address multiple niches, or is it really just suited for a couple of specific applications?
- Very soon, the market is likely to see significant consolidation. Right now, there is so much demand that everyone has as much business as they can handle, and the focus is on simple execution. But during the inevitable slower parts of the growth cycle, core competencies such as branding, distribution, etc. will drive companies to try to aggregate niches by aggregating technologies, and thus achieve economies of scope and scale. Some companies, for example Carmanah, are already targeting the long tail by aggregating different applications and market niches. Others will find the advantages of being able to "single-source" for installers a full suite of technologies addressing their different needs, and will look to acquire those technologies.
- Given the above, we are probably passing from a period where investors will be funding startups with an IPO exit in mind, and moving into a phase where trade sale is a relatively more likely exit. This should affect valuation expectations accordingly.
- While never unimportant, the quality of the management team and other aspects of execution are going to be increasingly important relative to the specific technology being developed.