The market dynamics of the PV industry have been a rollercoaster ride for more than five-years. A quick and easy example is the fact that not one company has been able to stay the market leader more than a year and this year looks likely to be no different.
At the forefront of the industry growth has been a huge infusion of capital across the supply chain that, if someone was brave enough to collate, must surely top US$100 billion or more. Freely available capital led to massive capacity expansions from new entrants as well as established PV companies.
While the focus has been on capital expenditure, gigawatts of capacity and billions of dollars of annual revenues for the top tier players, it is remarkable that little attention has been given to what should be another key metric: R&D spending.
Delving into the depths of annual SEC financial filings for the tier 1 PV module manufacturers to extract some of the data provided in this report highlighted how much emphasis and page space companies give to their research and development activities.
According to Suntech Power Holding’s 2011 annual report;
'We will need to invest significant financial resources in research and development to keep pace with technological advances in the rapidly evolving PV industry and to effectively compete in the future.’
However, Suntech goes further and provides a clear reason why R&D matters;
‘Our objective is to be the global market leader for the development, manufacture and commercial scale installation of PV products and systems, and to spearhead the movement to achieve grid parity. We will continue to develop new technologies and to design more advanced equipment to manufacture, cost-effectively and on a large scale, PV cells with higher conversion efficiencies.’
Sentences, paragraphs and entire reams of similar statements typically run through such documents for a large number of stock market listed companies.
R&D spending trends
Using the latest market leader rankings from Solarbuzz for 2012, we plotted the annual spending on R&D of the top 10 ranked companies (see figure 1), since 2007. However, due to Sharp Corporation not providing a breakout of its R&D spending specific to its PV operations, we replaced Sharp with Norway’s REC, which is ranked at number 11 by Solarbuzz for 2012.
First Solar heading the pack
Looking at chart 2 it is immediately apparent that there is a major disparity between the R&D spending behaviour of First Solar and its competitors. However, it should be noted that there are some unique differences between First Solar and the others that does go someway in explaining the disparity.
The most obvious factor of course is that First Solar is using CdTe thin-film technology, while the other four are crystalline silicon based technology users. The thin-film market leader also went on an aggressive capacity expansion plan on three continents and also used its form of ‘copy exactly,’ the manufacturing mantra of semiconductor behemoth, Intel Corporation.
Employing an all enclosed linear process from start to finish, as well as a history of pushing conversion efficiency improvements on all manufacturing lines - as part of the capacity ramp process also partially explains the inherently higher spending ratios, compared to its crystalline silicon rivals through to 2010.
However, in 2011 the continued increase in R&D spending, according to the company, was focused on its components segment. This segment, due to its major focus on its utility-scale PV project business added an extra spending burden. Also in the mix was the way First Solar accounted spending on further process developments and record the depreciation of equipment as research and development expense.
Neither charts that relate to R&D spending as a percentage of revenue and the number of employees working in the R&D job specification category, respectively reveal anything at odds with First Solar’s overall spending levels. However, the fall in employee levels within R&D in 2011 was due to halting work on CIGS technology and realigning R&D purely on CdTe.
REC at the peak
When we first analysed R&D spending we focused attention on the top 5 module manufacturers. In widening the scope, REC which is ranked in 11th position by Solarbuzz was the surprise party.
The pioneer of the fully-integrated business model had consistently outspent its c-Si rivals, including Suntech of the 5-year period analysed.
A major part of the spending had been related to its work on the Fluidized Bed Reactor (FBR) polysilicon technology but also having been a player in the solar wafer market had also been developing both mono and multicrystalline wafer technology.
In the last few years, emphasis had also been placed on high-efficiency cell and module technologies since 2009, which has resulted in the company maintaining a high spending commitment.
Unlike some of its rivals, R&D efforts have been undertaken where its various manufacturing facilities have been located, rather than centralised. The US has been home to its FBR developments, while Norway and Singapore have focused on wafer and cell developments respectively.
REC does not breakout its employees by functions so we were unable to analyse its headcount dedicated to R&D activities.
Although REC is well respected for its high-efficiency and quality modules, its R&D spending, second only to First Solar hasn’t provided the company with the revenue success of its rivals.
Suntech’s long-term R&D commitment
Even before 2007, Suntech Power Holding can rightly claim to have a history of R&D spending above its rival China-based crystalline silicon module manufacturers. As chart 2 reveals, Suntech increased R&D spending in 2008 through 2010 while its Chinese rivals continued to keep spending well below its spending habit.
The major spending push was due to its focus on ‘Pluto’ cell technology. The company has had various targets for the technology since its early days but is now aiming to reach increased efficiencies of around 21% for monocrystalline silicon and 17.5-18.5% for multicrystalline silicon-based solar cells. Although actually ramping the Pluto technology has proved slower than expected, most notably due to the perennial problem of process cost issues, Suntech has been able to achieve around a 0.5% efficiency improvement per annum for both wafer types.
Yingli Green catching-up fast
To say that Yingli Green didn’t really do R&D back in 2007 is probably an understatement. As figure 1 shows, Yingli Green was a serious laggard, spending only US$2.4 million on R&D that year. R&D spending as a percentage of revenue was only 0.4% and employed only 130 people in its research activities.
However, that all started changing in 2009 as the company started its PANDA technology push. Remarkably, Yingli Green actually spent more in R&D in 2011 (US$45.3 million) than Suntech (US$38.6 million).
Although Yingli shows an erratic spending behaviour as a percentage of annual revenue, the company has actually spent higher in percentage terms than Suntech on two occasions in the last 5-years.
But it doesn’t stop there. Yingli Green also recruited over 700 people into R&D activities in 2010, reaching 1,080 at the end of the year compared to only 259 the year before (see figure 3). That trend continued in 2011, when the figure reached 1,330.
With respect to project PANDA, Yingli Green collaborated with the Energy Research Centre of the Netherlands, and Tempress Systems, a wholly-owned subsidiary of PV equipment specialist, Amtech Systems. In 2011, Yingli reported an average cell conversion efficiency rate of 19.0% on PANDA production lines, and a record cell conversion efficiency rate of 20.0% on the PANDA trial production line.
Improvements were also seen with its multicrystalline cells, claiming average conversion rates of 16.2% in 2009, 16.5% in 2010 and 17.0% in 2011.
Trina Solar late developer
In 2011 Trina Solar more than doubled its annual investment in research and development from US$18.6 million in 2010 to US$44.1 million in 2011. The company has focused on solar cell manufacturing processes, which include passivation and metallization techniques. Like Yingli, emphasis is recent years has also been to develop monocrystalline capabilities such as its N-type SPARC solar cells with back-printed contacts.
Trina Solar had achieved mono and multicrystalline cell conversion efficiencies of 20% and 19.6%, respectively in 2011.
However, as chart 1 clearly shows, it was only in 2010 that Trina Solar really got serious about its R&D commitment. A late developer, Trina Solar is tracking Yingli Green’s increase in R&D spending, surpassing Suntech for the first time in 2011.
Yet the significant shift was the steep climb in spending as a percentage of revenue, which is now on the same trajectory as that of First Solar and rocketed past its Chinese rivals.
This catch-up is also reflected in the number of employees dedicated to R&D activities. Trina Solar went from on 48 people in this sector to 560 by the end of 2011, surpassed only by Yingli Green.
Canadian Solar needs ELPS
One of the most surprising revelations from the analysis has been how little emphasis has been placed on R&D by Canadian Solar. Although this has increased, they have been lagging its nearest rivals for the last 5-years.
It is hard to believe that Canadian Solar only spent US$1 million on R&D in 2007, while this equated to only 0.3% of revenue, the lowest of the top 5 producers. Even accounting for its increased spending and employee numbers, R&D investment as a percentage of revenue has yet to exceed 1%.
Jinko Solar’s minimalist approach
Arguably, Jinko Solar’s minimalist approach to R&D spending in relation to its ranking position is the hardest to explain. The company is the youngest of the pack, having only started manufacturing operations in 2006. Jinko Solar only created an R&D department in 2008 that included just 17 researchers.
However, in 2010 its research department increased significantly from 43 in 2009 to 343 in 2010. The company didn’t explain in its filings what sparked the increase and only a year later the headcount had fallen to 148, again not explained. The company also stands out for having the lowest R&D spending of all the major module manufacturers.
JA Solar’ slumber party
The low R&D spending at Jinko Solar is not the exception. JA Solar has also kept spending to a minimum over the 5-year period under review. In JA Solar’s defence the company has successfully transitioned away from being a wafer producer to that of a major module manufacturer. R&D spending through most of this time would have related to its wafer operations rather than on solar cells and modules.
However, without a significant increase in both spending and manpower allocation to R&D, maintaining its market position could prove challenging.
Having the world’s highest cell and module conversion efficiencies after more than 15-years of concentrated R&D activity, SunPower’s R&D spending shouldn’t come as a surprise. Over the period covered the company has consistently been the third-largest spender behind First Solar and REC.
What was surprising was that SunPower has out spent Suntech. With the acquisition of the company by oil firm, Total R&D efforts have intensified in 2012.
However, SunPower has had to spread its R&D activities wider than that of Suntech as it has been developing its tracker systems, concentrated solar and PV power plant technologies.
Maximising its return on R&D investments will become increasingly important. Yet its leading technology position is being boosted by Total’s involvement and development of ultra-thin wafer technology and advanced module assembly techniques should keep spending levels in the higher ranks for the next few years.
Hanwha SolarOne to watch
The acquisition trail of Solarfun and more recently Q-Cells has catapulted the Korean-owned company into the top 10 rankings. Hanwha SolarOne management have been emphasising its focus on technology over the last 18-months as it fights to create differentiated products.
The company has yet to differentiate itself in the R&D spending category but the recent Q-Cells purchase could be a catalyst for change. Currently, Q-Cells is operated as a subsidiary of parent company, Hanwha Chemical and little has been said by management on what course the companies will go.
Potentially, however R&D efforts previously made by Q-Cells could be a major benefit to Hanwha and its previously stated efforts to become a top player in the industry.
The analysis of the top 10 PV module manufacturers R&D spending activities points to a manufacturing sector that it still in the development stage. In the semiconductor industry, R&D spending as a percentage of revenue has historically been above 10% and often close to 20% during cyclical downturns.
However, R&D spending should be seen as a forward indicator of a company’s ability to compete against rivals, should R&D projects prove successful. Despite previous problems at First Solar, the CdTe thin-film leader continues to invest in its future and has already made a turnaround.
Yingli Green, Trina Solar and Canadian Solar have all woken-up to R&D as an essential aspect of their future health and wealth, yet Yingli Green seems the most aggressive in its actions at the moment.
We will be updating our charts and commentary when 2012 R&D figures are officially announced as part of annual financial reporting, and guidance given on spending for 2013.
Indeed, 2013 could prove to be a critical year as next generation process technologies, pushed back by many companies as focus centred on cost reductions should emerge. Recent R&D spending by the leading companies could provide a leading indicator of near-term business success.