For the first three quarters of 2011, Daqo New Energy (DQ) was able to remain extremely profitable, despite the majority of its crystalline photovoltaic solar industry peers suffering intense pricing pressure. Although pricing for its core polysilicon business had been steadily declining since the start of last year, margins remained very high.
As the solar industry has shown in past cycles, no individual vertical within the sector's value chain can retain a disproportionate share of profitability for any extended period of time. As downstream average selling prices for silicon wafers, solar cells, and modules collapsed below the manufacturing costs of most suppliers, pricing for the last holdout accelerated downward in the final four months of last year. By the end of October 2011, spot market polysilicon was quoted as low as $25/kg versus prices as high as $100/kg a year prior.
For the third quarter, Daqo's earnings report still showed relative strength compared to many direct peers. Although overall revenues declined 15.7% sequentially to $59.6m, due to pricing declines for its core polysilicon product, gross margin was still a very healthy 33.3%, despite a $3.7m inventory provision taken in the quarter. Excluding this charge, consolidated gross margin declined from 46.6% in the second quarter to 39.5% in the third quarter. In contrast, many US listed direct peers posted low single-digit gross margins at the better end, while much of the industry witnessed negative margins. Daqo's third-quarter gross profit prior to the inventory provision was $23.6m, slightly below my $24.3m estimate.
While polysilicon volume increased sequentially to 1,022 metric tons from 1,001 metric tons in the second quarter, Daqo's newer downstream segments posted shipments much weaker than the company originally guided. The company had expected 10MW of wafer and 20MW of module shipments, but only delivered 6.5MW and 11.3MW, respectively. It was the second quarter in a row the company's newer downstream segments delivered lower-than-expected volume. I cautioned in my Q2 review Daqo's entry into these new verticals could yield dilutive results, given the state of the overall industry. Based on the company's implied approximate $24/kg polysilicon gross profit, DQ once again sold silicon wafers and solar modules at a loss of almost $1m on a combined $6.6m in revenues.
Overall, Daqo's third-quarter net income totaled $12.1m, or 0.34 in earnings per share (EPS). Excluding two non-operational items[inventory provision and subsidy income], DQ's adjusted operational earnings came to $14m, which was shy of my $15.6m estimate. The difference was attributed to a slightly higher-than-estimated operating expense due to Daqo's new polysilicon expansion, as well as higher-than-expected dilution from its downstream business segments.
Unfortunately for Daqo, industry pricing dynamics went heavily against the company in the fourth quarter. While continued weakness in DQ's newer wafer and module business was expected, pricing for its core polysilicon business collapsed. The company posted average polysilicon asps of $52/kg in the third quarter, but by early fourth quarter, spot market pricing was quoted as low as $25/kg as noted by peers such as Canadian Solar (CSIQ), JA Solar (JASO), and Suntech Power (STP).
As indicated in Daqo's Q2 earnings review, the company's polysilicon production cost was fairly fixed in the range of $28-30/kg. In just one quarter, Daqo's robust 46% third-quarter polysilicon gross margin may completely vanish in the fourth quarter.
Although Daqo provided fairly detailed fourth-quarter guidance of 800-850 metric tons of polysilicon, 16MW of silicon wafer, and 14MW of module sales, the company may not generate a consolidated gross profit, based on metrics given in its Q3 earnings conference call. While DQ has not suffered from inventory cost blending lag like many larger peers, since its polysilicon production cost was already at a normalized state, a small but meaningful increase in its silicon wafer and module inventory may cause higher degrees of dilution than witnessed in prior quarters.
Daqo has apparently been running its newer wafer and module capacity at a fairly high utilization, despite recent shipment weakness. Inventory, which was indicated as mostly wafers and modules, nearly doubled in the third quarter to $29.1m. On an absolute basis, DQ's inventory level was extremely low compared to many peers, but compared to just $6.6m in third-quarter revenues generated from these segments, it was disproportionately high. As a result, the company's guided fourth-quarter run rate may cause higher gross loss and/or force another inventory provision.
An estimate for Daqo's fourth quarter has been compiled below using metrics stated or implied by its management and do not speculate outside those ranges. With the exception of a small $1.5m subsidy income estimate, all other unannounced non-operational gains or charges have been excluded. As noted, weaker than expected sell through of its newer wafer and module businesses may cause another inventory provision.
DQ's Q4 Earnings Estimate
Shipments: 825mt polysilicon, 16MW wafer, 8MW module, 6MW oem
Asps: $30/kg polysilicon, $0.35/watt wafer, $0.95/watt module, $0.35/watt oem
Unit Cost: $30/kg polysilicon, $0.42/watt wafer, 1.00/watt module, $0.35/watt oem
COGS: $24.8m + $6.7m + $8m + $2m = $41.5m
Gross Loss: -$1.5m
Gross Margin: -$1.5m / $40m = -3.8%
Operating Expenses: $4.5m
Net Interest Expense: $2m
Subsidy Income: $1.5m
Tax Benefit: $0.5m
Net Loss: -$6m
Diluted Share Count: 35.2m
In the near term, there appears to be little upside for Daqo's prospects. While polysilicon was the last of the four main crystalline verticals to fall in pricing, it is also the most commoditized and potentially the most over-supplied based on capacity targets by the top five global producers. Polysilicon pricing around $30/kg may continue for the foreseeable future, especially when much larger producers have streamlined costs in the low $20s/kg. With Daqo's main plant already fully utilized and with costs already normalized, it is possible the company may not be able to generate much, if any, gross profits from its polysilicon business.
The good news is despite being a relatively small supplier, DQ's production costs are competitive enough to maintain operations because its cash cost of production excluding depreciation may be as low as $20-$22/kg. Thus despite potentially generating no positive margins, the company's polysilicon plant is still a positive cash flow generator as long as the production can either be consumed and/or sold. Combined with relatively low operating expenses, DQ's polysilicon segment as a stand-alone unit is still a viable business under current industry dynamics.
As indicated in DQ's Q2 review, the company is in the process of expanding its polysilicon operations, which by the end of this year targets up to 12,000 metric tons of capacity versus the company's current 4,300 metric ton capacity. In addition and more importantly, the improved scale is expected to reduce overall manufacturing costs towards the low $20s/kg down from $28-30/kg currently. If successful, DQ's polysilicon business could rival larger peers and if pricing can stabilize at current levels, the company could generate similar levels of gross profits booked in the third quarter of 2011.
This scenario is still several quarters away and given the current state of the industry represents perhaps an optimistic view. With Daqo's limited scale downstream wafer and module expansion slow in gaining traction, the company's fortunes are becoming increasingly dependent on a single vertical. As a result, the company's success may be dependent on more factors than fully integrated peers. While current market valuations may already discount Daqo's future earnings potential, compared to larger more integrated peers, DQ remains a higher risk play in the solar industry.