Yingli Green Energy Gains Market Share

| About: Yingli Green (YGE)

In what was a very difficult quarter for the solar industry, Yingli Green Energy (YGE) once again outperformed many peers on a number of metrics in the second quarter. Oversupply continued to loom over the industry during the second quarter, which caused many less competitive, less bankable, and fragmented manufacturers to reduce or shut down production. Yet Yingli’s shipments rose 36.6% sequentially to record levels as revenues grew 63% on an annual basis. Declines in average selling prices (asp) across all major verticals in the crystalline segment of the photovoltaic (“PV”) value chain crippled the profitability of most peers to the point bankruptcy of high profile companies have already been seen. Yet Yingli’s gross profit as well as net income both rose on a sequential and year over year basis. Q2 of 2011 was far from a perfect quarter for Yingli Green Energy, but it was another quarter that increased the separation between the winners and losers in a rapidly evolving solar sector.

Since Yingli revised its second quarter guidance prior to its Q2 earnings report, there were few surprises. The raised shipment expectation increased overall revenues relative to my original estimate, but as explained, lower than originally expected module asps reduced gross margin slightly. As a result, I only estimated a $1-2m increase in operational earnings above my original figures.

YGE’s Q2 earnings before taxes and non-operational items came in at $58m, below my $64m estimate mostly due to higher than expected operating and interest expenses. However, on a US GAAP net income level, Yingli posted 0.36 in earnings per share (“EPS”), which topped my revised 0.29 estimate and Wall Street consensus of 0.28 in EPS. YGE’s upside surprise came mostly due to higher than expected net foreign exchange gains, lower than expected minority interest payouts, and a one time purchase gain of $8.5m. Without this non-operational one time gain, YGE’s Q2 EPS would have been 0.31 per share.

For investors, Yingli’s second quarter results shouldn’t be the primary focus. With Yingli’s as well as other US-listed Chinese solar stock prices so depressed and industry news flow so negative with perceptions of pricing declines causing sector wide bankruptcies, it’s easy to lose focus on the wider picture.

The solar sector boom in 2010 caused an oversupply situation since the turn of 2011. Many companies that had no business expanding so aggressively increased “bad” capacity to unsustainable levels. "Bad" capacity is defined by capacity which can only operate profitably under extreme or ideal market environments. In the crystalline segment of the photovoltaic industry, this capacity is generally represented by single or fragmented verticals. In an under-supplied environment as experienced last year, margins are high enough that even single verticals can thrive. As demand crosses under-supply, margins contract to the point where only higher integrated manufacturers can generate enough gross profit to operate profitably. In essence, the fragmented middle men get squeezed out.

The issue isn’t with the overcapacity in manufacturing capacity. The rapid pricing declines witnessed in the past several months was more a factor of inventory liquidation by uncompetitive companies which fed on itself rapidly. As explained in Yingli’s sister peer Trina Solar’s (TSL) Q1 earnings review, no company will operate for greater losses. Uncompetitive capacity is simply shut down as market prices fall below production costs.

Below $2.00/watt PV module prices, many higher cost “western” manufacturers began experiencing problems. US based Evergreen Solar (ESLR) and Energy Conversion Devices (ENER) began shutting down manufacturing lines as shipments declined despite a record growth year for the industry in 2010. As pricing dropped towards the $1.50/watt level in prior quarter, more cost thresholds became breached which eventually resulted in Evergreen’s Chapter 11 filing. More recently much higher profile Solyndra filed for bankruptcy despite over $500m in US government loan guarantees.

It ultimately comes down to cost; manufacturers with higher cost structures can quickly become vulnerable. It doesn’t matter if it's a higher cost crystalline manufacturing peer such as Evergreen Solar or thin film variant such as Solyndra. All suppliers can become affected, as recently witnessed when a more distant relative to silicon based PV, concentrated solar power, lost a massive 500MW project due to recent cost declines in crystalline-based PV alternatives. Lower cost silicon based PV is winning vs. higher cost alternatives.

This is the general environment investors in Yingli Green Energy should understand above operating results in any given quarter. Yingli is among the top three largest PV solar module suppliers in the world and one of the most cost effective producers. Current market pricing has dropped to levels where many smaller and/or more fragmented companies in YGE’s peer group cannot continue to operate for extended periods. As these uncompetitive companies reduce or shut down production, inventory has been liquidated at prices market leaders such as Trina Solar identified as “non-sustainable reinvestment” levels. Until this situation has been completely absorbed through demand, pricing would likely remain under pressure. Despite conditions that are ravaging most within the industry, Yingli is still able to operate profitably and gain market share.

In the near term, the pricing environment will negatively affect Yingli’s quarterly earnings. However, margin pressures at YGE as described in my prior quarter review should be more short term in nature. Yingli’s procurement costs are lagging current end market pricing. Part of the problem as also seen with Trina Solar’s Q2 is a delay in contracted polysilicon pricing declines due to less periodic price adjustments rather than at the moment spot market pricing. As a result, blended polysilicon costs only start to decline after contracted prices are adjusted at pre-scheduled dates. Since YGE sources almost two-thirds of its polysilicon needs through long term contracts, its blended costs will get reduced much more slowly than peers such as Jinko Solar (JKS), which sources all of its polysilicon through the spot market.

Secondly, as mentioned in Yingli’s Q1 earnings review, the company’s inventory remained at fairly high levels. Although YGE reduced inventory down from $461m at the end of Q1 to $396m at the end of Q2, inventory turns remained at almost a full quarter. This compares to Trina Solar’s roughly five weeks of inventory. As a result, it will take Yingli over a quarter to fully blend its inventory costs down to current market pricing. While blended silicon costs should trend lower in the remaining quarters of 2011, it may not be until early next year before YGE’s blended silicon costs reach current spot market levels of near $50/kg from the high $70s/kg realized in the second quarter.

These factors will cause Yingli’s polysilicon cost component in its module production to remain higher than current market levels for the remainder of the year. Since polysilicon comprised almost 40% of YGE’s module production cost in the past quarter, lag periods in realizing current procurement cost levels will put short term pressure on Yingli’s gross margin. Based on the metrics used in my estimate for YGE’s Q3 earnings, this time delay would dilute the company’s per watt gross margins by approximately 0.10/watt from a potential 0.33/watt. If YGE’s polysilicon costs in the third quarter fully reflected current market prices, Yingli would post 25% gross margin instead of the 17.4% gross margin calculated in my estimates.

Again, investors should not concentrate on what Yingli’s margins are in any given quarter. Instead, it is more important to realize the normalized earnings power of each company assuming reasonable market parameters. As Yingli has already shown in the second quarter and will likely continue to show in the third quarter, it can still operate at a reasonable level of profitability even at metrics which would force many industry peers out of business. As competition leaves the field, and as the resulting inventory liquidation is absorbed by demand, earnings metrics for companies able to gain market share under consolidation periods should revert to “sustainable reinvestment” levels as described by Trina Solar.

Due to these factors, Yingli’s third quarter earnings are likely to decline on a sequential basis based on the company’s general shipment and margin guidance. These metrics have been compiled in YGE’s Q3 earnings estimate below. As with all my estimates, all metrics used were either stated or implied by the company without speculation beyond those parameters. If companies alter its forward guidance, then my estimates would have to change accordingly since it only reflects information provided by the company. As usual, these estimates only reflect operational earnings and thus excluded non-operational items such as foreign exchange translations which cannot be properly estimated prior to the quarter’s close, or obviously unannounced gains or charges recorded by the company.

YGE Q3 Earnings Estimate:

Revenues: $725m
Shipments: 535MW
Asp: 1.32/watt
Unit Costs: 535 x 1.09 = $583m module + $16m others = $599m total
Gross Profit: $126m
Gross Margin: 17.4%
Operating Expenses: $68m
Net Interest Expense: $25m
Tax: $5m
Minority Interest: $2m
Net Income: $26
Share Count: 160m
EPS: 0.16

Lastly, there are two issues worth noting in Yingli’s longer term earnings potential. The first is a dramatic reduction in its minority payments during the second quarter vs. prior quarters. When YGE went public on the NYSE in 2007, Shanghai listed Tianwei Baobian was a 24% minority holder in Yingli’s main China based subsidiary. As the company expanded, new subsidiaries were formed independent from YGE’s original subsidiary Tianwei Yingli which Tianwei Baobian continues to hold its stake. As a result, sales from new subsidiaries are thus not subject to minority payments to Tianwei Baobian.

Although minority payments were gradually reduced on a quarterly basis since 2009, its impact wasn’t extreme until the past quarter’s earnings report. After Yingli clarified this issue in its Q2 earnings conference call, the parallel between Yingli’s capacity and shipment growth and inverse reduction in minority payout percentages became more evident after review. Moving forward, YGE’s minority payments may drop to the 4-5% of after tax net income levels instead of the 24% payouts originally reported initially after the company’s US IPO. Through expansion, Yingli has effectively reduced dilution caused by one of its main early minority investors.

Secondly, Yingli’s higher efficiency Panda line, which exhibited a shaky start in the first quarter, is starting to reveal its potential. Although Panda was minimally accretive to YGE’s margins in the second quarter due to its slow initial ramp up and under utilization, the premium line’s higher asps, which Yingli states at roughly 10% above its normal multicrystalline module line, did increase the company’s overall blended module asps. At roughly 1.59/watt in the Q2, Yingli did report a blended asp premium over US listed Chinese solar companies which reported module asps anywhere from 5-10% lower.

In Panda’s early ramp up stage, costs were reported higher than at levels the company expects at full utilization and optimization. Currently Yingli states Panda modules costs roughly 10% more to produce vs. standard multicrystalline modules, but expects the cost difference to reduce to 5%. If these operational conditions are reached in the future, Panda module production which could make close up half of YGE’s output at full utilization may give the company a 3-4% gross margin advantage vs. peers solely using normal multircrystalline production technology. While this small gross margin advantage may not sound interesting, at Yingli’s annual revenue level in the $2-3 billion and growing, it becomes incremental gross profit that fall almost directly to the bottom line.

Disclosure: I am long TSL, YGE, JKS.