In the past month since we recommended taking profits on our Tier One Chinese Solar trade, the sector has been hit heavily - largely driven by margin erosion and a generally less than encouraging earnings season. The key question from here is whether or not we are once again at prices which offer a buying opportunity. The answer is probably not quite yet.
click to enlarge
The chart above shows the percentage change in three Chinese tier one solar stocks plus the solar ETF TAN in the period since our last buy recommendation on Nov 28th of last year (see the original article here). In what continued to be a very volatile period for solar our basket rallied heavily and by the time of our recommendation to take profits on Feb 10th Suntech Power (STP) was up +82.5%, Trina (TSL) +67.4% and Yingli (YGE) +54.9% (see our original take profit recommendation here).
That happened to be the high of the year for those three stocks and they have fallen significantly since. Indeed, at the time of writing, STP has fallen -27% from its February highs, whilst TSL is down -32% and YGE has declined by -37%.
So is it now time to buy once again? The very volatile price action in the market reflects a genuine and very intense debate with regard to the extent to which 2011's very damaging supply imbalance has been eroded. Moreover, this level of volatility is likely to be with us for a while as the industry continues to undergo an intense period of creative destruction as it consolidates and transitions to a more stable environment.
So where are we in this transition process? To feel comfortable that we have a solid buy opportunity once again we need to see progress on five main factors:
- The clear emergence of visible new demand, particularly in China and the US, to replace the demand the industry is losing in Europe.
- A halt in capacity expansion plans amongst tier one solar players and clear evidence of significant capacity shedding amongst tier two and three players.
- A resultant stabilization of average selling prices (ASPS) across the supply chain.
- Further progress amongst module makers in blending in lower polysilicon costs. Clearly, the preponderance of long-term contracts in the industry has meant that actual poly costs to module manufacturers have lagged behind the significant fall in spot polysilicon prices.
- Similarly, further progress in lowering non-silicon costs in order to secure gross margin stabilization.
Somewhat obviously, significant progress on all of these factors will gradually bring more balance into the industry and will probably be enough to see both gross margins stabilize in the low teens and profitability return for those tier one low cost players who survive the current transition. With Grid Parity ahead in many countries by 2015 you then have a solid environment for the winners in this intense process of industry consolidation.
So how much progress has been made on each of these issues? First of all, in terms of demand, the numbers now stack up fairly well. Full year 2011 demand was probably around 26 GWs globally. This year demand in Germany is likely to fall significantly from a 7 GW handle to around 3.5 GW. However, demand in the US should pick up from 2 GW last year to 3 GW this year. Meanwhile, demand out of China looks likely to expand rapidly to around the 5 GW mark. Japan and India should also be growth areas. Overall, despite the sharp drop in demand out of Germany, there is good reason to see a small net increase in demand in global terms. So far, so good.
In terms of global capacity, one of the reasons that last November we recommended getting long Chinese tier one solar was that we were beginning to see major players finally cutting back on plans to increase capacity - for example see here. This was encouraging. More recently, Suntech Power has confirmed the company's intention to hold its global capacity at 2011 levels - 2.4 GW for modules and 1.6 GW for wafers. Much has also been made of the decision by Korea's main players to withdraw from their plans to push aggressively into the module market.
However, somewhat disappointingly, other major module suppliers in China have recently re-affirmed plans to expand capacity despite the over supply in the industry as a whole. For example, during Yingli's recent earnings call, the company confirmed that it will add 750 MW of module capacity to reach 2.45 GW this year. Similarly, Trina has stated the company's intention to add 500 MW of capacity to reach 2.4 GW.
This of course means that reaching something close to supply-demand balance in the industry will require much more outright destruction of capacity amongst tier two and three players. Supply in this area is much more difficult to get a handle on. However, what market intelligence we do have suggests that adjustment is certainly underway. For example, Suntech Power CEO Zhengrong Shi made the following point in the Q&A section of his company's latest earnings call:
'Well, what we have been told and through some survey in China for Tier-2 and Tier-3 manufacturers, more than 50% have reduced their production volume; and if we look at the bidding process in the recent China market, you will also see the number of participants actually has been reduced at this time'.
Meanwhile, in the Q&A section of Trina's recent conference call, company Chief Commercial Officer Mark Kingsley stated their understanding that 'in China in general, there have been more than 50 companies that stopped making solar and we expect that to continue'.
Nevertheless, this will clearly be a slow process of attrition, with the government in China showing no inclination towards forcing the pace of consolidation amongst solar players. The result in the short-term is likely to be that many of the struggling tier two and three players will continue to dump modules on the market at uneconomic prices in order to raise cash, thereby putting further downward pressure on average selling prices and industry margins.
This brings us to our third factor - after falling heavily in 2011, ASPs are likely to fall further this year. Final module selling prices for a company like Trina fell into the high 80 cents area in Q4 of 2011. Over the course of this year, a continuation of the over supply situation in the industry could push those margin levels down into the low 80s. Margin stabilization is therefore going to require considerable progress on getting costs down.
Turning to polysilicon costs first, most producers should benefit over the year as contracts continue to be renegotiated and the blended poly costs actually faced by module producers fall closer to spot levels. Trina has suggested, for example, that their current blended poly cost is around 'a $45 per kilo range' - or a contribution of about 30 cents per watt to module costs. Alternatively, an industry average of 5.5 grams of polysilicon per watt would imply a cost of 25 cents per watt.
As blended poly costs fall further over the course of the year, converging on spot levels by year end, optimistically poly costs per watt could conceivably fall to 15-20 cents - a not insignificant improvement.
Meanwhile, non-silicon costs continue to fall as module manufacturers continue to squeeze out production efficiencies. STP has indicated, for example, that their non-silicon production costs fell by 7% in Q4 to some 74 cents per watt. Both Yingli and Trina reached non-silicon costs of 64 cents per watt. Trina and Yingli expect to get that number down to 60 cents and 56-58 cents respectively by the end of this year. STP conservatively estimates 65 cents per watt by year end. However, this also reflects a higher efficiency mix of module output.
What is clear from these numbers is that the Chinese tier one players have a roadmap which could conceivably keep costs sustainably below ASPs, thereby stabilizing margins. For example, total costs in the low 70s per watt and ASPs in the low 80s would obviously provide gross margins in the low teens.
Canadian Solar (CSIQ) is already talking about total costs including polysilicon moving towards the 55 to 60 cent range per watt. However, that would require a further fall in poly costs to around $20 per kilo - an event which may well pull prices down across the supply chain, leaving gross margins for module producers little better than in the low teens anyway.
Nevertheless, the above certainly represents a roadmap which could well offer a post-consolidation positive story for the low cost producers who will be the winners from the current difficult process. On this basis we have little doubt that there will be a time and a price at which to buy the Chinese tier one players once again.
However, in the meantime the current quarter looks likely to see a continued supply-demand imbalance and further downward pressure on both prices and margins. STP has for example provided guidance looking for both a seasonally weak Q1 in terms of shipments and further margin contraction into the 3 to 6% range. That implies further considerable losses.
In conclusion, the price action probably looks like it will have to make further progress on the downside before all of this is priced in. In the meantime, it continues to look like a time to keep your powder dry and invest another day.