Speculation on the roll-out of new finance models, the next big emerging market and even whether manufacturers can match this year’s unprecedented demand have created additional uncertainty.

Among a chorus of predictions on market size and development, hotly-tipped technologies and estimates on the fall out of trade tariffs, Finlay Colville attempts to set the record straight and get the industry focussed on the big issues ahead. Mark Osborne asks the questions.

Back at the start of December 2013, NPD Solarbuzz forecast the 2014 end market to be 49-50GW when all other research firms were at 40GW and some even below this mark. Eight months into the year, how do you see the market in 2014 ending up at?

Things still look very much as though the December 2013 forecast is holding up well. Current estimates are close to the 49-50GW level, despite all the changes across the different markets globally.

Back in December, we adjusted our demand methodology significantly at NPD Solarbuzz. This was done to address the changes in the PV industry over the past couple of years. It also recognised that the manufacturers and developers were playing a much stronger role in shaping the overall market size in any given 12 months.

So, looking at the key leading indicators now, we are still seeing a 49-50GW end-market for 2014, with several metrics in place that provide lower and upper bounds to the original December 2013 market forecast, but don’t fundamentally alter the original methodology applied by us.

There have been some reports recently about a panel shortage in the PV industry. What are your thoughts on this?

Well, my initial thoughts are that this prognosis is somewhat na�ve and missing the point on how the solar PV industry operates today. Simply to have a basic supply/demand model in place grossly underestimates the role of multi-national producers and developers to ultimately frame the size, or demand, of the PV industry.

Basically, once you accept that the end-market, or demand, is being driven by government-allocated quotas and supply-driven dynamics, then there is in effect no panel shortage. If you decouple demand from supply, then of course you can arrive at panel undersupply conclusions, but this is missing the point on how the solar PV industry is operating today.

Some bold forecasts have been made about the US market being the biggest globally in 2016. Would you agree with that?

There are many different aspects of US centric optimism these days. On the specific question of the US being the largest market globally in 2016, this is really a red herring in terms of market analysis.

Firstly, it is relative, and unless we know what the proponents are referencing against, it cannot become a meaningful claim. Furthermore, if Beijing wishes to have the largest PV end-market in 2016, then have the largest end-market it will. China has the bandwidth to create this outcome, more than any other country. The rationale for this may be more political than environmental.

What is more relevant is the question of what the US market will look like in 2016, not what size it is relative to other countries. Will the leasing model remain attractive and spread to states where limited deployment has been seen? Will the utility segment sustain the strong growth potential seen in the past few years? How much of an ITC rush will be seen? These are more pertinent issues to contemplate.

Recent estimates of SolarCity’s planned expansion of Silevo have put the cost at close to half a billion dollars. What are the chances of this happening?

The numbers may be somewhat immaterial and of secondary interest. The key issue from an analyst perspective is quite simple: does the technology work? If this was anyone other than SolarCity, the announcement would have probably not garnered headline coverage.

During the past 10 years, there have been dozens of Silevo’s, or other start-ups advocating differentiated c-Si based technologies. Several others surfaced in the US, many across mainland Europe, and some in Asia Pacific.

We still return to what we know about the PV industry and the track-record of advanced c-Si concepts in being able to scale to mass production. Only two companies have managed this in 20 years: SunPower and Panasonic-Sanyo. Each spent almost 10 years learning the recipe. Each accumulated vast IP resources that drove tool suppliers to develop equipment specific to their needs.

But more importantly, each of these companies evolved within a PV industry that was very small, growing at relatively small increments and could accommodate higher cost manufacturing approaches. New companies seeking to scale advanced c-Si concepts today are faced with a completely different set of challenges.

So the question is not how much money is needed to be thrown at Silevo, but is there really enough known about the technology to merit the capex projections? Of course, in practice, SolarCity’s initial financial exposure is likely to be minimal, based on ramping small capacity levels and potentially having other state level incentives in play to minimize any risk implications of a GW fab.

That’s not to say the GW fab won’t materialise, but it will likely be the result of many milestones and targets being hit one after another over the next 3 years.

The US trade case against China and Taiwan is currently the subject of much discussion. What is the short-term impact on the PV industry going to be over the next 12 months?

It has been interesting to see so much focus on adding 2012 or proposed 2014 duties and coming up with spread-sheet scenarios about the cost competitiveness of product shipping from China or Taiwan that does not meet the non-duty parameters.

The 2014 AD case has been approaching for 18 months like a freight train in the distance, moving ever closer at snails-pace. To think that the market leaders had not put contingency plans in place against the worst-case scenarios may be somewhat na�ve. Of course, not all would have been so proactive, but most had the hindsight of previous US and EU rulings to learn from.

But does China really want to fund the US through duties? Or even it some did, for how long? Is the point of duties really to create a level playing field?

Certainly, to a large portion of China and Taiwan, paying duties is simply not an option. Yes, the US market is important globally, but it is not the only game in town. So, is play by China in the US market essential, or politically required as part of a bigger long-term objective? Working that out may be more important than simply getting out the slide rule and doing a matrix of theoretical duties that may be paid.

The other big issue missed by many is that adding a duty and getting cost-equivalence to a brand-established supplier is not really how the market works. Regardless of the country or the application segment, there is a delta in prices for Chinese supplied modules and brands like REC Solar or SolarWorld. You don’t simply add a duty, get to price equivalence with the Western brands, and still have a market to sell into.

While Taiwan was the previous beneficiary of the 2012 US trade case with China, are there likely to be any new countries that end up with opportunities they had not foreseen?

Yes, this is probably the more pertinent question in the short-term, and this is likely to explain why weighted averages for module pricing will increase to the US market. Simply by having less Chinese brand modules being sold will move blended average selling price (ASPs) higher.

Straight away, we have seen REC Solar grab business, and it is likely that other brand-established suppliers will also see opportunities in the near term. However, the situation for Korean-based suppliers may be less obvious, as many of them have chosen to restrict shipment volumes and to stay in higher margin residential markets.

But the other knock-on effect of the US trade-case could be in markets such as Japan. If Chinese based shipments see a downside of several hundred megawatts during the end of 2014, then other markets are likely to absorb this. One part of this equation is obviously the second half shipment quota within the domestic Chinese market, but the effects could also extend to Japan.

Increased shipments to Japan from China may further act to erode the ASPs in Japan, where multi c-Si modules have increased share and the value attached to local brands can be seen to diminish on tighter project margins.

Moving onto technology, there has been a lot of talk about PERC cells being the next big thing in cell technology. Are there any other technologies that could compete with R&D capex next year?

Rear passivated cells are certainly emerging as a viable candidate for the next round of cell upgrades. It has been fascinating to see this upgrade route move from blue-sky research 20 years ago to being near the top of R&D activity and upgrade strategies for so many Taiwanese and Chinese cell makers.

The success of shifting to dual/double printing on the front surface has been a major factor here, as it gave Asian cell makers the confidence that upgrading cell lines is a viable roadmap route.

PERC is somewhat more challenging, as it needs several process changes and multiple tool supplier participation. But compared to some of the more esoteric cell technologies, it may be the best option available for the next couple of years.

In contrast to the front printing changes that were cost reduction driven, PERC gives a major boost to efficiencies and should be applicable for both mono- and multi- substrates.

So, at the top level of PV technology, where do we stand on the mono versus multi debate? Has anything changed recently?

Beyond a shadow of doubt, the simple question of mono versus multi is now the dominant question for PV technology roadmaps in the next five years. This dwarfs the debate on premium c-Si cell plans of Silevo and TetraSun, and is also far more relevant than n-type versus p-type discussions. Knowing how much p-type will be made on mono or multi will be the defining benchmark factor for all PV module manufacturers in the near to mid-term.

No-one thought five years ago that a 60-cell p-type multi c-Si based module would be approaching 275W panel ratings by the end of 2015. Nor that the best-in-class Asian manufacturers would be producing this type of module with silicon and non-silicon costs south of 50 US cents per watt.

So far, the big push has been from optimising the directional solidification furnaces, with multi quality simply getting better and better every quarter. Adding incremental cell improvements, such as dual/double printing and PERC, and the combination could be game-changing.

The improvements from p-type multi c-Si manufacturing have also been a factor in accelerating efficiency upgrades at First Solar for example. But the impact on n-type and p-type mono ingot capacity is more important, and whether next-generation crystal pulling can raise the bar beyond what the legacy semi-borrowed ingot growth technologies had achieved in the past.

Hanergy is still being very bullish on thin-film, claiming the technology will grab strong market share gains in the next five years. How do you see Hanergy’s strategy evolving as they seek to ramp up CIGS capacity?

Had Hanergy just unveiled a thin-film strategy that centred on CIGS, and was focused on one acquisition, say Solibro, then the overall synopsis of Hanergy’s thin-film prospects may be more transparent. But we have the history of the attempted multi-GW fab built-out of a-Si based technology across various sites in China. There were three CIGS acquisitions, all with different technology approaches. And there is also a diverse range of downstream strategies, such as project development and trying to sell rooftop kits on IKEA shop floors.

Because the overall strategy is so different to anything else in the PV industry, it becomes almost impossible to comment on just how it will evolve. But how many other organisations would be able to bankroll so many approaches, and continue acquisitions, in the absence of strong market-share gains or tangible end-market revenue streams? For now, we can only wait and see.

Finally, looking at project developer’s changing strategies, hold co’s and yield co’s are all the rage today. What’s behind this, and should we expect to see more or less going forwards?

This topic is particularly interesting today for a whole bunch of reasons. First, from a visibility perspective, it is fascinating, as it has brought downstream financing operations to the upstream manufacturing segment – something that was largely decoupled before First Solar, SunPower and SunEdison shifted gears from legacy upstream activities.

Unless you have supreme confidence that 40% gross margins will return to component supply – whether you make poly, wafers, cells or modules – then there is no other source of revenue streams from the PV industry unless you participate in system selling or long-term ownership, residential or utility based. The exceptions to this of course are some pure-play Asian manufacturers that may be satisfied with low single-digit margins, so long as the volumes are large enough. Or some of the largest poly suppliers that can retain attractive margins, and avoid the midstream ‘squeeze’ effect when upstream supply levels change quickly.

But whether you have a build, hold or yield strategy is both company-specific and project-specific. Where you take risk in the project is also different: from simply being a green-field developer taking a site to shovel-ready with the intent of flipping, to looking for short-term buys of projects completed by other EPCs to add as assets to portfolio capacity targets. The risks, returns and rewards are different, and can vary from one project to another quickly.

The other interesting part though is that the revenue streams, or confidence in future revenues, has shored up what was becoming a massive problem with simply being a high-volume manufacturer within an industry where the price curve is only trending one way in the long term. Having to constantly fight on costs, just to stay afloat, is not a particularly attractive model to most investors or shareholders.

So, for many reasons, visibility on the role of large-scale solar farms purely as assets, or the role of secondary market activity, is actually providing a great educational tool for many on the upstream that had not been required to consider the whole value-chain proposition of solar PV in the past.