Daqo New Energy performed relatively well in the second quarter, which was particularly difficult for most within the crystalline photovoltaic industry. Daqo was able to maintain a robust level of profitability evident in its second quarter earnings report, which was in line with slightly lower revised guidance given a month prior. Unlike other crystalline verticals, DQ’s core polysilicon business witnessed relatively better pricing and average selling price (ASP) declines lagging downstream counterparts in both magnitude and timing.
For the second quarter, Daqo posted revenues of $70.7 million, which was down 19% sequentially due mostly to a 12% ASP decline for its polysilicon product. While the company is currently expanding downstream, polysilicon sales still generated 90% of its revenues in the second quarter. In comparison, polysilicon’s direct neighboring downstream vertical, silicon wafers, saw ASP declines of 21% as reported by ReneSola (SOL) and LDK Solar (LDK). With nearly double the magnitude in ASP declines as polysilicon, most wafer suppliers either lost money or barely remained profitable in the second quarter.
Nevertheless, a 12% decrease in polysilicon ASP was enough to compress DQ’s gross margin from 51% to 46.6% on a sequential basis. As a result, Daqo’s gross profit declined by almost 26% from the prior quarter. Fortunately for the company, its core polysilicon business has relatively fixed operating costs, since it’s primarily a raw materials business. Downstream verticals such as solar cells and modules require more research and development, as well as brand marketing expenses. On a US GAAP basis, Daqo earned $25.7m in the second quarter, or 0.73 in earnings per share (“EPS”).
While DQ’s Q2 0.73 EPS vastly surpassed Wall Street consensus estimates of 0.54, it was generally in line with my revised estimates. Within Daqo’s second quarter earnings was a non-operational income of $3.8m from local government grants and incentives. While the company has mentioned possible income from government incentives, most earnings estimates excluded this figure due to its unpredictable nature. For example, in the most recent earnings conference call, management indicated annual incentives may range in the $5 million to 6 million area, but the company has already recorded $6.7 million in government incentives in the first half alone. Excluding this figure which added 0.11 in EPS, Daqo’s second quarter earnings would have been 0.62 per share vs. my 0.61 EPS estimate.
Looking forward, Daqo’s earnings are likely going to continue contracting. Unlike many downstream peers that still have the ability to reduce cost of goods sold, DQ’s polysilicon production cost should remain fairly constant at its current installed capacity. For its current 4,300 metric ton polysilicon capacity, which is running at full utilization, management has guided production costs to generally range between the $28-30/kg level, depending on seasonality of electricity costs. With small variance in costs for its core polysilicon business, and the continued decline in ASP from the $63/kg reported in Q2 to the range of $50-55/kg in Q3, Daqo’s gross margin will most likely contract further.
As mentioned in Daqo’s first-quarter review, the company is currently in the process of expanding downstream. Relative to its polysilicon scale, the company’s downstream wafer and module capacity is very limited and dwarfed the integration levels of its closest US listed counterparts, ReneSola and LDK Solar. Nevertheless, successful downstream expansion even on a limited scale should add incrementally to gross profit and potentially mitigate portions of the gross profit loss from its polysilicon segment in a declining ASP environment.
However, with new downstream capacity coming online during a period where the entire industry is in a general state of oversupply, success is far from guaranteed. Daqo’s second-quarter earnings revision was entirely related to greater than expected weakness in its new downstream wafer and module businesses. Even at very small volumes guided for Q2, DQ still wasn’t able to meet its original guidance.
As a new player where branding is especially important, Daqo may have to use pricing as an incentive to gain market share. Many fragmented single vertical wafer companies even at much larger scale than DQ lost money in the second quarter. Even some higher tier branded module producers warned on lower than expected shipment volumes for the quarter. Any market share gain from these new business segments would be extremely difficult in the current market environment.
Unlike many in between single verticals, Daqo does have an advantage with its relatively low cost polysilicon production. With polysilicon gross margin still high relative to downstream peers, the company can in essence subsidize its downstream expansion. That is what appears to have occurred in the second quarter.
From the either stated or implied cost metrics, Daqo’s cost of goods sold for its polysilicon and wafer segments totaled $29.7 million out of the company’s overall blended $37.8 million cost of revenues. This means approximately $8m of its revenue cost was attributed to its module segment. However, in the company’s earnings report, DQ stated it only generated $6.8 million from its module businesses. Daqo lost about $1.2 million, or 17.6% negative margins from its branded, OEM, and module processing business segments in the second quarter. Not only did Daqo miss original volume targets, but had to conduct business at a loss just to gain little traction in this new module product line.
Daqo’s tiny 1.3MW of wafers sold in the second quarter was also essentially subsidized by its high-margin polysilicon business. From polysilicon sales alone, the company generated around 0.20/watt in gross profit during the second quarter. At extremely low utilization and hence higher depreciation costs, DQ stated in its conference call wafer conversion costs were in the mid-0.30s/watt. With an additional 0.18/watt in polysilicon costs internally consumed in its wafer production, the company’s wafer cost may have been over 0.50/watt in Q2. At an ASP of 0.69/watt, the company generated no additional incremental gross profit in conducting its new wafer product line.
In the longer term, if volumes can continue to ramp to full utilization, these two new business segments should add to gross profits albeit however small. Establishing a brand and gaining market share may be more difficult in today’s solar industry landscape where smaller or less cost effective participants are being forced out by larger scale, fully integrated producers. Daqo’s established polysilicon business could subsidize growth in these new segments, but the results may be slow to materialize until further industry consolidation corrects recent oversupply issues.
The more pressing concern for Daqo is if polysilicon pricing continues lower and at a faster than expected rate. Currently spot market pricing ranges between $50-55/kg for solar-grade polysilicon and has been stable in recent weeks after the initial correction from over $80/kg levels seen prior to May. Looking forward as more capacity comes online, further pricing pressure could force polysilicon prices towards or even below $40/kg as indicated by many industry leading peers. With production costs unable to be further reduced by a similar magnitude for its current polysilicon capacity, Daqo may continue to experience gross margin pressure for its core business looking forward.
To combat the probably compression in per-kilogram gross profit for polysilicon sales, Daqo is in the process of expanding capacity. Phase 2 of its polysilicon capacity will add 3,000 metric tons annually, if not more, once fully optimized at a targeted cost of $20-25/kg. Lower production costs for this new capacity is possible due to newer equipment that is not only cheaper to purchase but operate more efficiently as well. In addition, the new plant is located in a more remote area, Xinjiang. Thus labor as well as electricity costs are expected to be much lower.
Daqo also recently announced plans to adopt new processes for its current 4,300 metric ton polysilicon plant that is expected to more than double output to 9,000 metric tons. With the total costs expected to be just $135 million, depreciation amounts for the additional output will be significantly lower. The combined effects of these initiatives should bring the company’s blended polysilicon production costs to the $23-25/kg range by the end of 2012 when full utilization is scheduled. If successfully implemented, this will not only increase output, but also help minimize margin compression if polysilicon ASPs decline lower.
In the nearer term, Daqo guided Q3 shipments to range 975-990 metric tons for polysilicon and 10MW each for its wafer, module, and module processing business segments. Added volume for downstream verticals should add more incremental revenues than gross profits. As a result, blended gross margin should continue to contract. As usual, an earnings estimate for DQ’s third quarter has been compiled using the company’s guidance as well as stated cost and pricing metrics without further speculation beyond company comments. Again, as usual with all my estimates, it excludes non-operational items such as net foreign exchange translations and more specifically to Daqo any additional government subsidy income.
Daqo New Energy Q3 Earnings Estimate:
Revenues: $71.5 million
Shipments: 980mt polysilicon, 10MW wafer, 10MW module, 10MW OEM
Asps: $52/kg polysilicon, .55/watt wafer, 1.15/watt module, .35/watt OEM
Unit Cost: $28/kg polysilicon, .42/watt wafer, 1.10/watt module, .35/watt OEM
COGS: $27.5 million + $4.2 million + $11 million + $3.5 million = $46.2 million
Gross Profit: $24.3 million
Gross Margin: $24.3 million / $71.5 million = 34%
Operating Expenses: $4.1 million
Net Interest Expenses: $2 million
Tax: $2.6 million
Net Income: $15.6 million
Diluted Share Count: 35.3 million
EPS: 0.44
As a final note, recent share price declines among US-listed Chinese solar companies such as DQ may already reflect declining earnings. Daqo is facing margin compression and thus reductions in net income, but from very high levels. Even at third quarter’s estimated earnings on an annualized basis, which is half the run rate experienced just a few quarters ago, DQ could still earn anywhere from 1.50-2.00 in EPS. At under 6.00 per share recently, the company’s share price much like many direct peers is trading in the low/mid-single digit price to earnings ratio.
In fact, valuations in the sector have gotten so depressed such that both peers most closely related to Daqo on a structural basis, ReneSola and LDK Solar, have recently announced share repurchase programs. A $100 million share repurchase plan was announced by ReneSola last week, and LDK Solar revealed in its Q2 earnings conference call that it has already fully executed its $110 million share buyback since announcing it a couple months ago.
While Daqo’s near-term earnings may still be unstable due to an extremely fluid pricing environment, over the course of the next year additional polysilicon capacity at lower production costs should help offset potential further ASP declines. In addition, once the current liquidation by uncompetitive peers exiting the industry plays out, pricing spreads between major crystalline verticals should eventually normalize to more reasonable levels. At that point, DQ’s downstream capacity should add incrementally to gross profits. These two factors should, through the end of 2012 keep gross profit levels on average from declining much further relative to Q3. Thus most likely given the dynamics known today, Daqo’s earnings for next year should still range between 1.00-1.50 in EPS. At the company’s current stock price range, valuations would still be in the mid-single digit earnings ratio.
Disclosure: I am long LDK.