As the last of the major U.S.-listed Chinese solar companies to report Q1 2011 earnings, LDK Solar’s (LDK) earnings report didn’t contain many surprises, especially after the company pre-announced results a month earlier. Still, LDK’s earnings release is significant to review in detail because it contains perhaps the most accurate and up to date assessment of the current state of the solar industry. With many peers reporting first quarter earnings in early to mid-May, prior forecasts made by other solar companies did not accurately reflect the current pricing dynamics within the industry. Across all four major verticals in the crystalline photovoltaic industry from polysilicon to modules, average selling prices (“ASP”) dropped by as much as a third in just the past month.
Many factors played a role in the recent drop in selling prices. The most notable is an oversupply situation brought about as many companies expanded to meet last year’s demand boom. A large part of the new capacity brought online was fragmented in that many companies only operated in one vertical and became dependant on the success of each corresponding participant in the entire value chain. For example, silicon wafer-only producers sold to solar cell-only producers who in turn sold to module manufacturers. In an undersupplied environment, gross margin was high enough across each vertical for everyone to sustain healthy gross profit.
In an oversupplied environment, gross margins compress squeezing out smaller scale or less efficient manufacturers. The situation was made worse for single verticals as large scale giants such as Suntech Power (STP), Trina Solar (TSL), and Yingli Green Energy (YGE) also expanded aggressively at a much higher level of integration. Even formerly large scale but less integrated producers such as Canadian Solar (CSIQ) and China Synergy (CSUN) also embarked on higher integrated expansion in the past couple of years. Fragmented single verticals became more and more dependant on other fragmented peers instead of large scale industry leaders.
In a seasonally weak first quarter, which was compounded by a policy concern in a key solar market, Italy, slow demand hit fragmented manufacturers first. As highlighted in my recent Trina Solar article, many single verticals experienced sharp monthly declines in business starting in March and extending through April. With industry demand channeling through top tier suppliers first, many smaller players were forced to curtail production or simply liquidate inventory in order to generate cash flow as noted in Suntech‘s Q1 earnings conference call.
By May, inventory dumping fed on itself and prices across all verticals dropped sharply during each week throughout May. Just in the past two weeks alone, polysilicon and wafer ASPs dropped by almost 20% and 15% respectively according to PVInsights. Thus, LDK’s second quarter guidance given in its Q1 2011 earnings conference call shouldn’t be too much of a surprise for those following the industry closely.
For the first quarter however, LDK Solar still performed extremely well considering earnings pre-announcements by higher integrated peers such as Trina Solar and Yingli Green Energy. As noted in my earnings revision for LDK, the company’s lowered revenue was more a function of product mix as the overall level of gross profit was only guided down marginally.
To be sure, Italian policy concerns did ultimately delay module shipments in the first quarter as overall module shipments declined by over 20% sequentially. While revenue was lost, LDK’s absolute level of gross profit did not decline significantly because the company only lost revenues from a low margin business unit. LDK’s Q1 2011 gross profit totaled $241.6m, or down only marginally from the $251.4m in gross profit recorded in the seasonally stronger fourth quarter. As a result, gross margin expanded from 27.3% to 31.5% quarter over quarter as a higher ratio of LDK’s shipments were at a higher level of integration. With most operating and non-operating expenses sequentially in line on a percentage basis, LDK’s pre-taxed earnings declined slightly to $179.7m in Q1 from $189.8m in Q4.
Due to a higher than expected realized tax rate of almost 25% vs. the 15% 2011 tax rate noted in the company’s recent annual report, first quarter net income declined to $135.4m, or .95 in earnings per share (“EPS”) which beat Wall Street consensus of .86 in EPS, but fell short of my 1.01 EPS estimate. Under a 15% tax rate, LDK’s net income would have been $152.8m which would have beat my $145m estimate.
Many Chinese solar companies, including LDK Solar, have been granted “High and New Technology” preferential tax rates of 15% vs. the standard uniform 25% rate. In LDK’s case, its preferential rate expired last year and may be currently under consideration for renewal. As Trina Solar noted, renewal can take months but normal tax rates must be used in the meantime. If or when granted, the lower tax rate would retroactively be applied to previous quarters.
Prior to 2010, LDK Solar was primarily a silicon wafer supplier with upstream polysilicon capacity slowly coming online. Had the company maintained this strategic positioning, the recent ASP declines would have left it extremely vulnerable like many fragmented single verticals. However since early last year, LDK started to transition from primarily a single vertical to a fully integrated supplier from polysilicon all the way down to systems integration. According to its 2010 annual report, LDK’s targeted capacity by the end of 2011 is as follows: 25,000 metric tons polysilicon, 4.2gw silicon wafers, 1.8gw solar cells, and 3.5gw pv modules. In essence, by the end of this year, LDK’s capacity and vertical integration level would rival the largest tier one peers such as Suntech, Trina Solar, and Yingli.
As a result, it would be incorrect to view LDK Solar as just a silicon wafer supplier. According to management, LDK expects by Q3 2011 solar module revenues would be a larger portion of the company’s revenues than its traditional core wafer business. It is LDK’s ability to shift from fragmented verticals as pricing spreads collapse to a fully integrated product mix that should save the company’s profitability levels moving forward. While additional capital is being allocated to complete and de-bottleneck LDK’s polysilicon plant as well as retrofit the company’s current wafer capacity, nearly all of LDK’s new capacity additions this year expand its cell and module capacity. If this integration of capacity rather than expansion of capacity continues next year, by the end of 2012 LDK could completely transform into a fully integrated photovoltaic systems provider on a scale larger and more integrated than any current peer in the industry today.
Until this integration process is complete however, LDK is still vulnerable to recent pricing declines more severe in middle verticals than the end module vertical where brand and bankability can differentiate companies. LDK’s core wafer vertical could see ASP declines over 20% on a sequential basis in the second quarter. Gross margin on a per watt basis for silicon wafers could get cut in half relative to the first quarter. With likely half of the company’s revenues still dependant on its wafer division, overall gross margin should decline namely due to per watt gross profit contraction in wafer sales.
A few buffers
Luckily for LDK Solar, the company does have a few buffers. The first is as noted, higher levels of module shipments. With a second quarter module shipment guidance of 200-220mw, or up from 118.7mw recorded in the first quarter, overall revenues should stay relatively stable despite large declines in wafer selling prices. In addition, a higher ratio of module shipments will also be at higher integrated levels as new cell capacity comes online. For the second quarter, LDK guided internal cell production to increase from 45.8mw in Q1 to 120-130mw for Q2. By my estimation, this should increase LDK’s module gross margin from 15.6% in Q1 to roughly 22-23% in Q2.
Secondly, LDK’s polysilicon plant is progressing extremely well. Production for units already online is near full ramp and LDK expects to produce 2650-2750 metric tons of polysilicon in the second quarter. More importantly, production costs continued to decline in the first quarter. With polysilicon spot pricing in Q1 no higher than 85/kg, and polysilicon procurement costs as reported by peers also no higher than this range, LDK’s 58% polysilicon gross margin implies production costs declined further to 36-37/kg from about 39-40/kg seen in the prior quarter. This production cost estimate is further reinforced by the company’s stated blended silicon consumption cost of 46.2/kg vs. prior quarter’s inventory carrying cost of 56.4/kg at similar volumes to LDK’s Q1 polysilicon production.
Although declines in blended polysilicon costs will only have a small effect in LDK’s Q2 earnings, every small increment is critical given the recent pricing environment. Continued minor sequential processing cost reductions across multiple verticals should also help buffer last month’s ASP collapse. For the second quarter however, cost improvements as well as changes in product mix will unlikely negate recent ASP declines across all verticals. For the second quarter, LDK’s official guidance is as follows:
LDK Q2 2011 Guidance:
Gross Margin: 22-26%
Wafer Shipments: 500-550mw
Module Shipments: 200-220mw
Internal Cell Production: 120-130mw
Polysilicon Production: 2650-2750mt
Based on this guidance and coupled with numerous other metrics LDK has stated in its earnings release and conference call, a second quarter earnings estimate has been compiled.
LDK Q2 2011 Earnings Estimate:
Core Wafer: 520mw @ 0.68/watt = $354m
OEM Wafer: 5mw @ 0.40/watt = $2m
Module: 200mw @ 1.50/watt = $300m
OEM Module: 20mw @ 0.45/watt = $9m
Polysilicon: 500mt @ 64/kg = $32m
Total Revenues: $732m
Cost of Goods:
Core Wafer: 520mw @ 0.51/watt = $265.5m
OEM Wafer: 5mw @ 0.30/watt = $1.5m
Module: 70mw @ 1.30/watt = $94.5m + 130mw @ 1.06/watt = $138m, $232.5m total
OEM Module: 20mw @ 0.40/watt = $8m
Polysilicon: 500mt @ 37/kg = $18.5m
Total COGS: $556m
Gross Profit: $176m
Gross Margin: 24%
Operating Expenses: $48m
Operating Profit: $128m
Net Interest Expenses: $32m
Government Subsidies/Other: $2m
Net Income: $73m
Diluted Share Count: 150m
Potential Net Foreign Exchange Gain: $5m
Potential Net Foreign Exchange EPS Impact: .03
As usual, the estimates above represent only operational earnings and exclude most non-operational gains or losses not previously announced. Net foreign exchange translations have been separated and is based on the recent exchange rate between the euro, usd, and rmb, relative to each other. The net foreign exchange gain estimate above assume LDK does not alter its currency exposure from the prior quarter and is mainly affected by a roughly 3% appreciation in the euro vs. the usd since the start of the second quarter. With a month of currency trading left, the ultimate net foreign exchange gain or loss is still open ended and dependant on where the euro vs. usd vs. rmb end the second quarter.
In addition and in contrast to more aggressive estimates made last year during the industry’s boom cycle, LDK’s Q2 earnings estimate above portray a more conservative estimate. A number of factors could cause reported EPS to be higher. First, overall shipments could be higher outside of unexpected events such as Italian policy uncertainties as well as a major natural calamity in Japan, which caused push outs in Q2 shipments. Secondly, average selling prices could be higher on a blended level. Pricing was still relatively strong in April and did not start to drop rapidly until mid-May. In LDK’s case, if pricing stabilizes in mainly the wafer vertical at recent levels, blended ASP could be higher than estimated above. Lastly, a lower 15% tax rate could be applied and possibly retroactively applied to Q1 earnings.
While on a sequential basis, LDK’s profitability should decline significantly, Q2 is likely an earnings trough for the company in 2011 as long as pricing in the more fragmented verticals stabilize. The main argument for pricing stability is an end of inventory liquidation. Very simply, many companies cannot operate on a sustained basis at current pricing spreads. Smaller or less efficient manufacturers may liquidate inventory at losses, but it would be extremely unlikely they continually operate at losses. History in the industry suggests inefficient capacity shuts down factories rather than continue to operate at greater losses.
Price stability would not cure LDK’s lower profitability levels in the second half of 2011, but it would at least set a floor. LDK is large enough and efficient enough to operate even at liquidation levels for most the industry, albeit at much lower levels of profitability. As long as profit margins can stabilize in the company’s key wafer vertical, continued increases in module shipments can make up the gross profit difference.
In the company’s Q1 earnings conference call, management noted a dramatic pickup in demand since June. Module shipments have increased each month in the second quarter and is expected to increase monthly for the third quarter as well. This implies Q3 module shipments could be in the range of 300mw, up from the 200-220mw guided for Q2. In addition, internally produced cells in Q3 can be as high as 200-250mw. The added gross profits from incremental module shipments at higher levels of integration combined with stabilized wafer pricing should cause Q3 profitability levels to rise sequentially. As noted, in this event, Q2 should be LDK’s earnings trough for this year.
Finally as noted with other U.S.-listed Chinese solar companies, sequential slowdowns should be viewed in the proper context. Many are already trading at historically low valuations. In LDK’s case, once Q1 2011 earnings are applied, trailing twelve month EPS would increase to 3.12 per share. In other words, LDK’s 40% share price decline in May to around $7.00 a share values the company at slightly over 2 PE. Earnings in Q1 2011 were slightly down sequentially from seasonally strong Q4 2010 levels, but still up significantly year over year. Due to issues detailed above, Q2 2011 earnings are likely to also decline sequentially, but would still represent annual growth over Q2 2010 levels. If Q2 is indeed an earnings trough for LDK, 2011 annual earnings would also likely post growth over 2010.
Disclosure: I am long LDK, TSL, YGE.
Additional disclosure: No Position in STP, CSIQ, CSUN.