A Preview of LDK Solar's Possible Polysilicon Division IPO

This is part two of LDK Solar s (NYSE: LDK) recent financing activities. Much of LDK's financing rush described in part one may be linked to the company's persistent efforts on keeping their corporate plans intact. Perhaps one of their ultimate goals is to ensure a successful HK IPO where they hope to raise $1 billion USD later this year. Since the IPO of their polysilicon division could dramatically reshape LDK's finances, possible scenarios drawing from recent industry history will be examined in this second part.

The first historical reference which involved the Hong Kong market and polysilicon production was GCL-Poly's purchase of Jiangsu Zhongneng for $3.4 billion USD in the middle of 2009. The structure of the deal involved GCL issuing over 2.8b USD of new common stock. The amount raised from the HK equities market is staggering considering many solar related deals completed in the US in the past raised a tenth or even less as much.

In comparison, most profitable solar juggernauts such as First Solar (NASDAQ: FSLR) and Trina Solar (NYSE: TSL) raised around 300m and 100m in their IPOs, respectively. The two newest US listed solar IPOs, Jinko Solar (NYSE: JKS) and Daqo New Energy (NYSE: DQ) both raised under 100m USD each. While LDK's intended $1 billion USD HK IPO is large by US market standards, larger deals such as GCL's purchase of Zhongneng's polysilicon division have been completed in Hong Kong.

This may be the reason LDK is turning to Hong Kong, and GCL's deal may be the example LDK is following.

From an operational standpoint, there are some parallels as well. At the merger date, Zhongneng stated an 18,000 metric ton polysilicon production capacity by the end of 2009, the same as LDK's target for the end of this year. However, in terms of actual production, LDK's 1900 metric ton polysilicon production guidance for Q4 2010 is already ahead of the roughly 1200 metric ton GCL produced in the middle of 2009.

If LDK achieves a 40/kg production cost, it would also mirror Zhongneng's 39/kg production cost at that time. LDK's polysilicon operating metrics will likely improve by the time they attempt their IPO later this year. From a value proposition, LDK's intended polysilicon division IPO is no less attractive than GCL's purchase of Zhongneng where Hong Kong investors were willing to sink over 2.

8b USD of new capital into the company.

With this historical reference in mind, LDK's $1 billion HK IPO of their polysilicon division appears reasonable to what Hong Kong investors might be able to digest. If LDK is successful with their IPO later this year, how might it be structured?

While the actual structure of the IPO may vary, it's most likely LDK will want to remain a majority shareholder in the new IPO. A controlling stake is important for LDK if the company wanted control on how polysilicon production is allocated in the future. With 18.46% of the polysilicon division already divested in an IPO, only a 31.5% stake is left for distribution if LDK wanted to remain a majority shareholder.

In a simple scenario, LDK might divest a third of its 31.5% limit, or 10.5%. The remaining portion would then be sold by the polysilicon division ("LDK-Poly" for short) in the IPO. The maximum amount of new shares issued by LDK-Poly would be approximately four times the amount LDK sells in the IPO.

As a result, LDK would be selling 20% of the IPO shares while LDK-Poly would be selling the remaining 80%. In a $1 billion capital raising IPO, LDK would generate $200 million themselves while LDK-Poly would raise $800 million for their own corporate needs. LDK would still own slightly more than 50% of LDK-Poly while about 37% of the spin off would be owned by the public.

In this example, LDK-Poly's IPO size would be slightly above $2.7 billion USD.

Why would LDK-Poly's IPO size be smaller than the price GCL-Poly paid for Jiangsu Zhongneng? Although LDK's polysilicon unit today is well ahead of Zhongneng in mid 2009, LDK may want to attach much of the division's debt in its IPO. By doing so, LDK would raise the enterprise value of LDK-Poly's IPO.

LDK may try to package up to 1.8b of the polysilicon plant's construction costs in the IPO. In comparison, GCL's solar division debt was under 600m after their purchase of Zhongneng's polysilicon unit. In this scenario example, 1.5b of LDK's debt is transferred to LDK-Poly, bringing the enterprise value of its IPO to 4.

2b or very close to the enterprise purchase price GCL paid for Zhongneng.

While GCL's purchase of Jiangsu Zhongneng's polysilicon division serves as a template on how LDK-Poly's IPO may be structured, LDK would still have to present an attractive proposition for new Hong Kong investors to accept a large IPO especially one that carries a high debt load.

Perhaps LDK's first pitch attempt is an unusual polysilicon supply agreement signed to BYD late last year. This is unusual because it's LDK's first polysilicon supply contract. Because LDK still relies on polysilicon procurement for their own internal needs, selling some of their own production makes little sense.

In addition, instead of signing their first contract with a well known solar industry peer, LDK signed their first supply agreement with a car maker with intentions of entering the solar pv field? While BYD is perhaps a total unknown in the solar industry, it's a very high profile Chinese auto producer.

BYD is also listed in Hong Kong and is perhaps best known after Warren Buffett's Berkshire Hathaway took a 10% stake in the company in late 2008. LDK could now claim most if not all of LDK-Poly's polysilicon production is not only secured by its LDK parent, but also by a far more well known company BYD.

Secondly, LDK recently transferred stakes in its polysilicon division from local to national level investors. Earlier this year the company announced they repurchased all of the 15% stake in their 15000 metric ton polysilicon plant sold in late 2009. At the same time, LDK issued new bonds which when converted grant holders an 18.

46% stake in their polysilicon division. In the press release, LDK announced that investors included both China Development Bank as well as China Construction Bank, two of China's big four banks. The company also stated other major banks were involved, but did not disclose names. With China's largest banks as investors, LDK-Poly's IPO may be viewed as having strong financial backers.

Lastly and perhaps most importantly, what would LDK-Poly's finances look at like at the time of its IPO later this year? At full capacity which LDK-Poly should be close to achieving by the end of this year, the unit could produce about 18000 metric tons of polysilicon annually discounting any over nameplate production already experienced by many peers.

LDK-Poly would also be able to claim a supply contract to BYD along with its parent LDK. Even assuming 55/kg average selling price, much lower than the 70-90/kg currently seen in the spot market today, LDK-Poly could generate 25/kg in gross profit assuming production cost targets were met. As a result, $450 million in gross profits could be generated from polysilicon sales.

Unlike other verticals, operating expenses for polysilicon production is very low because costs are fixed and are not a function of revenues. A good example is evident in a polysilicon pure play Daqo New Energy (NYSE: DQ) who guided for relatively flat operating expenses for the year. Assuming 1.5b of debt is transferred from LDK to LDK-Poly and assuming a 6% annual interest rate on this debt, total operating and interest expenses would total approximately $100 million.

Because LDK-Poly only recently generated profits, the division may still be under a tax holiday which normally exclude taxes for the first two years of profitability. LDK has not broken down the tax rates between its divisions so it's unclear what tax status LDK-Poly is under. The maximum rate however is the 12.5% rate indicated in LDK's annual report.

In terms of cash flow, the outlook is even better. With nearly all of the installed capacity already in place, LDK-Poly would require very little capital expenditures on its existing capacity. Because depreciation represents around 12/kg of production costs, the cash generated in gross profit would increase to 665m using the same metrics as in the example above. Excluding financing/investing activities and capital expenditures, cash flow generated from operations would be approximately 200m above net income reported.

Thus, even when assuming conservative metrics, LDK-Poly could generate slightly over 300m in net income at full capacity. In the 2.7b IPO size example described above, IPO investors would be paying roughly 9x price to earnings. Without further expansion and assuming the same metrics, LDK-Poly could generate around 500m in cash flow annually, or enough to pay the $1.

5 billion of debt (excluding the $800 million raised from the IPO in this example) transferred from LDK in about three years of operations. In comparison, GCL's price to earnings ratio is about 25x their Q3 2010 annualized net income. The Hang Seng Index, Hong Kong's stock exchange, trades at roughly 16x earnings on average currently.

In conclusion, LDK has already taken steps to present a strong case for their polysilicon division by linking it to high profile investors as well as customers. Even under fairly conservative assumptions, the valuation on LDK-Poly's IPO in the scenario presented above would still represent a discount to a very similar direct peer and to the overall Hong Kong market in general. If LDK's IPO plays out similar to this example, it would greatly restructure the company's debt.

The new LDK-Poly IPO would have $800 million in cash raised from the offering, while taking on $1.5 billion of LDK's debt linked to the cost of its polysilicon plants. LDK would have its debt reduced by $1.5 billion to around $400 million, plus generate $200 million from selling shares in the IPO. Based on financial ratios reported in LDK's last earnings report, it would leave the company with a net cash balance of over $500 million.

Ultimately, the numbers will shift depending on market conditions at the time the IPO is priced. Overall ratios may be different from the example presented, but the end result should help restructure LDK's debt in a meaningful way. Debt restructuring may be LDK's primary motive as they look to recapitalize in order to compete with large rapidly expanding industry peers such as GCL-Poly.