Similar to Renesola (SOL), Daqo New Energy (DQ) reported first quarter earnings before most other US-listed Chinese solar companies. Despite reporting earnings in early May, almost two weeks after SOL's earnings, Daqo’s report still came before a large industry price adjustment, as detailed in my recent Renesola article.
In fact, Daqo’s main business vertical of polysilicon production was among the last major crystalline photovoltaic vertical to hold average selling prices (“ASP”) firm. In an industry where pricing among each vertical is so highly correlated, polysilicon ASPs finally gave ground after remaining high for several months, even as downstream vertical pricing started to decline. In the three weeks after Daqo’s Q1 2011 earnings report and conference call, spot market polysilicon prices fell as much as $20/kg from the mid- to high- $70s/kg to under $60/kg.
Prior to May, the polysilicon vertical was perhaps the most constrained among the four major verticals that comprise the crystalline photovoltaic market. High capital costs and long lead times in constructing polysilicon plants led to relative under investment compared to other downstream verticals. When overall industry demand skyrocketed in 2010, other verticals were more able to expand capacity than the polysilicon vertical. The capacity imbalance was clearly witnessed in the second half of 2010 as spot market polysilicon prices rose from around $50/kg to $85/kg on average, with high end quotes exceeding $100/kg. Since polysilicon production is Daqo’s main business, the company’s margin and profitability also significantly expanded.
Polysilicon pricing continued to remain strong throughout the first four months of 2011, and Daqo’s strong first quarter earnings reflected this market environment. Revenues in the first quarter grew sequentially to $87.3m from $81.9m despite ASP declines for polysilicon. Most of the company’s realized polysilicon ASP decline from $76/kg to $71.4/kg was the result of 40% contracted volume at lower prices in Q1 2011, whereas the prior quarter saw no volume at long term contracted pricing. Higher polysilicon shipments, along with an increasing downstream product mix, also helped expand revenues despite general pricing declines in all operational segments.
However, given Daqo’s generally flat production costs, declining ASPs did compress the company’s margins. Gross margin declined to 51% from 55.9% quarter over quarter. Again, since nearly 90% of DQ’s revenues were still derived from polysilicon sales, lower polysilicon ASPs were the main reason for margin compression. As the company ramps newer wafer and module capacity, higher volumes of lower margin wafer and module business will also dilute overall gross margin moving forward.
Overall corporate profitability was more or less comparable on a sequential basis. Consolidated gross profit only declined 2.8% to $44.5m in the first quarter as higher revenues made up for most of the gross margin contraction. An unexpectedly higher realized government subsidy of nearly $3m as well as a reversal of a small foreign exchange loss helped increase net income to $35m in Q1 2011 from $32.8m recorded in Q4 2010. On an earnings per share (“EPS”) basis, Daqo posted first quarter EPS of 0.99, which was well ahead of Wall Street expectations of 0.85 per share.
As Daqo progresses through 2011, it will look increasingly like a smaller scale version of Renesola and LDK Solar (LDK). All three companies have significant scale in the polysilicon vertical, which is very different from larger peers more concentrated in downstream cell and/or module verticals. Unlike SOL and LDK, however, DQ’s wafer and module division is new and extremely small in scale. By the end of 2011, for example, Daqo expects to have 250mw of wafer and 200mw of module capacity, with no cell capacity in between. In contrast SOL expects to have 1900mw, 240mw, and 600mw of wafer, cell, and module capacity respectively while LDK’s scale is even greater at 4200mw, 1800mw, and 3000mw of wafer, cell, and module capacity respectively.
Despite Daqo’s much smaller downstream scale, any incremental revenue from the company’s new wafer and module division should help buffer recent downtrends in polysilicon pricing. DQ’s second quarter guidance was modest, and thus still achievable despite heavy pricing declines witnessed in the three weeks following its conference call.
For the second quarter, Daqo expects to ship just 4.5mw of wafers at an implied ASP of just .64/watt which was already much lower than the 0.75-0.80/watt range seen at the time of its guidance. Recent spot market wafer pricing have fallen below 0.60/watt but this may be more a reflection of inventory liquidation by struggling lower tier producers than a sustainable trend. At least on a blended level for the entire quarter, Daqo’s wafer ASP guidance is already lower than implied levels by both SOL and LDK.
On the module business side, DQ expects to ship a total of 16.5mw in the second quarter, of which 14mw will be company branded. At such a small scale, visibility should be high enough such that there should not be too much volume risk. Again, on the ASP front, Daqo gave extremely conservative guidance which implied pricing at around 1.50/watt or possibly lower which is the low end ASP range relative to guidance given by other US-listed Chinese solar companies.
For 2011, Daqo hopes to ship 125mw of total module shipments, implying second half volumes roughly tripling levels expected in the first half. Whether the company can achieve such an accelerated level of downstream business expansion remains to be seen, especially given the industry’s general oversupply of lower tier and/or newer entrant capacity. Shipment volumes recorded in the first quarter as well as that guided for the second quarter are still too low to draw absolute conclusions regarding the company’s sales channel strength. If such targets can be met, then incremental gross profits from this business unit would help cushion gross profits lost as polysilicon ASPs continue to decline to more normalized level.
In addition, Daqo already has 125mw of wafer capacity online as of the second quarter. As noted, a total of 250mw should be operational by the end of 2011. As a more natural extension of the company’s core polysilicon business, wafer shipment targets should be more easily achieved than for modules. Granted, if the current inventory liquidation influenced pricing environment persists, gross margin may be extremely slim at 25-50% of recent or historical levels, but any volume would represent incremental gross profits for the company. That is key for DQ as any incremental gross profit from any new business segments could help offset polysilicon gross profits lost under ASP contraction.
Until newly built downstream business segments ramp to higher utilization levels, Daqo’s gross profits should contract on a sequential basis. Wafer and module shipment expectations for the second quarter are still too small to make up for a 25% top to bottom decline in quarterly ASPs for polysilicon. Due to blended effects, reported polysilicon ASPs for the second quarter should not reach levels seen in the spot market recently, but should be lower than the company’s prior guidance which implied average polysilicon prices around $67-68/kg. Below are estimates for Daqo’s second quarter which are mostly based on the company’s statements although a lower blended polysilicon ASP is assumed.
DQ Q2 2011 Earnings Estimate:
Shipments: 1000mt polysilicon, 7mw wafer, 14mw module, 2.5mw oem
Asps: $64/kg polysilicon, 0.64/watt wafer, 1.50/watt module, 0.45/watt oem
Unit Cost: $29/kg polysilicon, 0.51/watt wafer, 1.30/watt module, 0.40/watt oem
COGS: $29m + $3.6m + $18.2m + $1m = $51.8
Gross Profit: $38.8m
Gross Margin: $38.8m / $90.6m = 42.8%
Operating Expenses: $4m
Net Interest Expenses: $2m
Net Income: $27.5m
Diluted Share Count: 35.3m
The estimates above represent only operational earnings and exclude unannounced gains or losses from non-operational items. While Daqo should record a small net foreign exchange gain based on currency levels seen as of today’s date, the amount should be very small and thus excluded. In addition, the company may also record additional government subsidies. These two non-operational gains could contribute a minor $1m increase to overall net income, or roughly .03 in additional EPS.
With blended polysilicon ASPs still estimated in the low $60s/kg for the second quarter, Daqo is still very vulnerable to additional pricing declines in its core vertical in the second half. Spot market pricing could return to levels seen a year ago at $50/kg, or possibly even lower if overall solar industry demand doesn’t grow fast enough to encompass much higher levels of polysilicon capacity being brought online in the next several quarters. In short, operational metrics DQ witnessed in the past three quarters were not sustainable and the company’s normalized level of earnings which assumes more sustainable pricing trends is much lower.
For many who valued Daqo based on unsustainable metrics will likely be disappointed. For others who base forward analysis on more normalized trends, DQ’s prospects haven’t changed too dramatically. In my first corporate overview of Daqo, I noted the possibility of polysilicon pricing being unsustainable and DQ should be analyzed based on more normalized metrics such as $50/kg polysilicon ASPs even well ahead of recent declines in pricing.
Based on assumptions ASPs would drop towards $50/kg in the second half, I estimated DQ should still be able to earn 2.75-3.00 in annual EPS. With the first quarter already posted and assuming the second quarter comes in roughly as estimated, first half 2011 EPS would total approximately 1.75-1.80 leaving 1.00-1.25 left in the second half to meet this annual earnings target. Just as a rough preliminary estimate which is based on $50/kg polysilicon ASPs combined with moderate increases in wafer/module shipments and small cost improvements, Daqo should be able to post 0.55-0.65 in quarterly EPS in the second half of 2011.
Of course, one has to value DQ based on forward annualized earnings and not trailing numbers. While Daqo should still post 2011 annual EPS near 3.00, its normalized annual earnings based on $50/kg polysilicon is lower at 2.20-2.50 in EPS. By my early estimation, DQ must operate at full wafer and module utilization based on current capacity targets to maintain a quarterly EPS of around 0.50 per share if polysilicon pricing continues to decline towards 40-45/kg. As a result, annual EPS of 2.00 is still very achievable for Daqo if the company can execute well even factoring in more realistic longer term metrics.
Lastly, as with other US-listed Chinese solar companies, recent share prices may already factor in lower earnings expectations. Daqo slightly differs from some peers in that forward annualized EPS would most likely be lower than trailing numbers as a result of capacity constraints and unsustainable pricing levels realized in the past three quarters; most downstream peers experienced margin pressure in recent quarters due to higher upstream procurement costs such as polysilicon which resulted in a benefit for DQ. However, even based on lower expectations, DQ’s forward earnings power of about 2.00 in EPS values the company at only 5x sustainable earnings.
Disclosure: I am long LDK.