LDK Solar's Earnings Normalization Continues in Q4 2010 Report

Includes: DQ, LDK, SOL, SPWR, TSL
by: Investing Hobo

LDK Solar (NYSE: LDK) continued to show progress in reaching normalized earnings with its recent fourth quarter 2010 report. Higher levels of capacity utilization relative to the company’s installed fixed assets should continue to produce incremental gross profits. In addition, higher levels of integrated capacity should also improve LDK’s gross margin. While LDK Solar’s recent earnings report showed both of these characteristics, the company was still only half way towards achieving full capacity utilization.

Among U.S. listed solar companies, LDK posted the highest sequential gross profit increase on an apples to apples comparison. U.S. based Sunpower (NASDAQ: SPWRA) posted a large absolute gross profit gain due to uneven revenue recognition for its system sales, while Daqo New Energy's (NYSE: DQ) 70.2% sequential gross profit gain came from polysilicon average selling price spikes rather than from operational improvements. LDK Solar increased its fourth quarter gross profit to 251m, up from 150m or a staggering 67.6% sequential increase. Trina Solar (NYSE: TSL) came in second in terms of sequential growth at 26.6%. Other peers who had already reported posted flat gross profit growth relative to the prior quarter. LDK’s closest competitor in terms of product mix, Renesola (NYSE: SOL), only posted a sequential gross profit increase of 2.2% in the fourth quarter of 2010.

In terms of gross margin, LDK Solar also showed a meaningful sequential improvement over the prior quarter. For the fourth quarter of 2010, LDK recorded gross margin of 27.3%, up from 22.2% in the third quarter. Again, only DQ was able to post higher magnitudes of gross margin expansion due to large average selling price increases in its core polysilicon product. With metrics already normalized, many direct peers posted flat or even declining gross margin in the fourth quarter of 2010 vs. the prior quarter.

With much of LDK’s costs relatively fixed, gross profit gains trickled down to the bottom line. Most of the company’s interest costs were already accounted for, although portions may still be capitalized for inactive segments of its polysilicon plant. While operating costs should move higher as the company expands, as a percentage of revenue it should remain flat if, not decline through economies of scale. LDK’s Q4 2010 operating costs did increase to an unusually high level, but the company was still able to expand net income margin from 13.8% to 15.8% on a sequential basis. Generally operating costs posted in the fourth quarter may not reflect normal rates since end of year activities such as balance sheet adjustments and annual bonuses typically occur in the quarter.

LDK’s net income grew to 145.2m in the fourth quarter, up from 93.4m posted in the prior quarter. On an earnings per share basis, the company reported a U.S. GAAP 1.09 in EPS-- which easily beat Wall Street expectations of .92 in EPS, but fell short of my 1.14 EPS estimate on mainly a higher than expected tax rate.

Although the recent news flow surrounding LDK Solar was so dynamic, only aspects regarding the company’s earnings prospect, as well as operational statements made by the company in its recent Q4 2010 conference call will be further discussed in this article. For more balance sheet related analysis, please refer to my recent articles highlighting LDK’s recent financing activities, as well as the company’s potential IPO of its polysilicon division.

Since the start of 2011, LDK Solar made a major adjustment in its efficiency assumptions for the core wafer product. When the company IPO’d in 2007, the assumption was that each wafer produced 3.7 watts. In 2008 and 2009, the assumption was increased to 3.8 watts. Moving forward starting in 2011, the new assumption will be 4.1 watts for its main multicrystalline wafers. Since LDK and other wafer producers sell wafers by the piece, the change in output assumption due to cell efficiency improvements does not alter the gross profit potential for its installed wafer capacity.

However, this change will alter every reported metric regarding LDK’s wafer business. With a higher wattage output assumption, LDK’s stated wafer capacity will go up while its wafer related costs and average selling prices will go down. As a result, Q1 2011 metrics will look dramatically different from Q4 2010 metrics, which used lower wattage output assumptions. It will appear the company is selling more wafers at lower prices but in fact it is selling the same per piece volume at the same prices, assuming actual metrics did not change on a quarterly basis. This move is similar to Renesola’s assumption adjustment announced a year ago.

In contrast to LDK’s aggressive expansion practices in the past, the company appears to be taking a more tempered approach regarding capital expenditures in 2011. Much of its new expansion will add to cell and module capacity. Along with four cell lines at its main Xinyu facility, LDK seems to be adding 17 new cell lines at the new Hefei manufacturing center. At 60mw per line, the expansion will give LDK 1.26gw of new cell capacity from zero at the start of 2010. It is also important to note that the Hefei cell and module expansion is fully financed by the local government, thus saving LDK fund raising efforts.

With all of LDK’s new capacity already financed, most of the company’s remaining capital expenditures this year will be allocated towards upgrading its existing capacity. In addition to completing the final stages of its 18,000 metric ton polysilicon plant, LDK announced debottlenecking efforts to increase its output capacity. The company hopes that by the end of this year, debottlenecking can increase the polysilicon plant’s capacity by 7000 metric tons to 25,000 metric tons in total.

If LDK can successfully increase the capacity at its polysilicon plant, it would have a significant impact operating metrics. Not only would the company be able to sell or internally consume higher volumes of polysilicon, but it would realize a lower cost of production as well. Higher yields should lower production costs since a portion of the costs are fixed. One of the key production cost metrics is depreciation. With yields increasing by 38%, depreciation on a per volume (kilogram) level should decrease by 28%. In essence, just through increased production yields, LDK should be able to shave 4-5/kg off its polysilicon production cost. This would help the company lower its fully loaded polysilicon production cost to around 25/kg, which is more in line with leading polysilicon producers today.

LDK Solar is also taking steps to upgrade its core wafer capacity instead of adding new equipment this year. Through equipment upgrades, LDK hopes to augment its existing wafer capacity from 3gw at the end of 2010 to 3.65gw by the end of 2011. Just as in the case for polysilicon debottlenecking, higher wafer production yields using the same equipment base should lower realized production costs as well. In combination with the company’s new wafer wattage output assumption, LDK should be able to increase its overall wafer capacity to approximately 4gw by year end.

Normally a seasonally weak quarter, the first quarter of 2011 should be slightly more challenging for most solar companies. On the cost side, supplies at the polysilicon and wafer verticals remain tight. On the sales side, module ASPs have dropped to reflect lower subsidy levels in key markets. Although actual performance will vary from company to company, generally most downstream producers with limited upstream capacity should feel the highest margin pressure. As a result, most companies that fall into this category should post first quarter results flat or declining from fourth quarter 2010 levels.

However, LDK Solar should be insulated from the seasonally weak first quarter. Its core wafer vertical is currently still experiencing strong pricing. On the cost side, LDK also has polysilicon capacity which is slowly being higher utilized. Whether LDK consumes or sells its own polysilicon production, the net result will generate incremental gross profits for the company. Its capital expenditures are focused on improving existing capacity efficiencies as well as further integrating the entire value chain. As a result, LDK Solar should have enough catalysts to shield it from short term industry hiccups in order to maintain continued sequential improvements.

Based on the company’s Q1 2010 guidance, this appears to be the case. Although some metrics such as shipments reflect a seasonally weak first quarter, LDK still guided gross margin to expand to 27-29%. The company’s revenue guidance suggests roughly an 8-9% sequential drop, but over the past four quarters LDK has beaten its original revenue guidance for each quarter by an average of 14%. However, a sequential revenue drop for LDK should still be the likely scenario due to lower selling prices as well as changes in its product mix. Gross margin expansion should be enough to make up for potentially lower revenues such that overall gross profit does not decline. On an absolute level, LDK should generate a higher net income on a sequential basis, but due to a higher weighted average diluted share count because of its recent secondary, bottom line earnings per share may remain flat vs. the prior quarter.

Below is an estimate of LDK’s first quarter earnings. It assumes the company is operating at full capacity given statements the company has made about its business condition as well as the relative pricing environment for its main products. There may be changes in LDK’s actual product mix, but the overall absolute gross profit potential should not deviate too far from this estimate since it reflects the company’s earnings power potential per unit of installed capacity.

LDK Solar Q1 2011 estimates:

Core Wafer: 625mw @ .85/watt = 531m
OEM Wafer: 125mw @ .48/watt = 60m
Module: 130mw @ 1.7/watt = 221m
Module OEM: 8mw @ .45/watt = 3.5m
Polysilicon: 1000mt @ 65/kg = 65m
Total Revenues: 880m

Cost of Goods:
Core Wafer: 625mw @ .58/watt = 363m
OEM Wafer: 125mw @ .28/watt = 35m
Module: 80mw @ 1.58/watt = 126.5m + 50mw @ 1.15/watt = 57.5m, 184m total
Module OEM: 8mw @ .37/watt = 3m
Polysilicon: 1000mt @ 40/kg = 40m
Total COGS: 625m

Gross Profit: 255m
Gross Margin: 29%
Operating Expenses: 45m
Operating Profit: 201.5m

Net Interest Expenses: 25m
Forex Gain: 5m
Government Subsidies/Other: 2m
Tax: 29m
Minority Interest: 6m

Net Income: 157m
Diluted Share Count: 143m
EPS: 1.10

As with all my estimates, they only reflect operational earnings with estimates of non-operational items factored in. Any unannounced or unexpected charge related to business write downs or gains due to various non-core business activity are obviously not factored in.

Disclosure: I am long LDK, TSL, DQ.

Additional disclosure: No position in SOL, SPWRA.