Last Friday, the LDK Investor Group held its third interview with LDK Solar's (NYSE:LDK) Chief Financial Officer, Executive Vice President and Secretary, Jack Lai. We had sent him a large number of questions, covering everything from order book, financials and plans for the poly plant. (These questions had been contributed by many of the group’s members.) We were restricted to 30 minutes at this time, and therefore followed up with a second interview on Monday.
In light of the recent analyst reports and the article from Barron's, we decided to focus on the three main issues as outlined below.
In Q4, the inventory has risen far faster than revenues. The implication from Barron's is that this rise is evidence that LDK is stockpiling or sitting on unusable inventory. We wanted to ascertain if this is the case and establish why the rise in inventory has happened. During the conference call, some analysts said they could not reconcile the value attributed to the Inventory with the quantity in stock and we also addressed this issue.
2) PolySilicon Plant Progress and Supply of Poly Feedstock
Another major issue raised by the analysts is that if the larger polysilicon manufacturing facility is delayed then LDK will suffer severe poly shortages in 2009. This would result in the company being unable to fulfill their contractual obligations and/or reducing the margins substantially.
We focused on the implications of this event occurring and questioned Mr. Lai on the backup plan in the event of a delay.
The last issue raised and addressed was the ability of LDK to finance their new plant and maintain cash flow throughout the year.
On the whole, we find that LDK tries hard to answer all questions and concerns within the confines of competitive information and public disclosure. They need to intensify this effort with the people who really matter – analysts and money managers – to overcome what at present appear to be double standards; for example:
1) Rising inventory is seen as a major issue by Lazard while at the same time they are criticized by Piper Jaffray because secured poly supply (ie inventory) is not sufficient to meet 2008 needs.
2) Analysts are concerned about margins falling despite the fact that LDK already have the highest margins in the multicrystalline industry and are projecting further a increase to 42% to 50% gross margin for 2009, up from 30% in their most recent results.
3) Piper Jaffray is willing to value First Solar (NASDAQ:FSLR) at a lofty 40 times 2010’s earnings, yet LDK is only valued at 14 times this year's projected earnings, despite similar earnings growth prospects over the same time frame.
4) Analysts have sometimes downgraded the stock while simultaneously raising their expectations. One would expect that higher expectations would reflect a higher share price so it is hard to reconcile these positions.
These confusing messages are likely the result of LDK providing insufficient information to the analysts covering the company. As LDK continues to execute on its business plan, they will need to focus more efforts on addressing such concerns.
To participate in the Investor Group, you can join the group on Facebook, or send an email to either Hakan Telenius (htelenius[at]shaw.ca) or Conor Shaw (conorshaw1[at]gmail.com).
What are the inventory components?
Having received several requests for additional information, the company is putting together a more detailed schedule of the inventory components and will include this within the presentation for the Morgan Stanley conference. The presentation will be published on their website.
In the meantime they have roughly broken down the inventory for us as follows:
- 1,600 tons are presently “in possession”; of which
- 856 tons is raw poly in the warehouse
- 752 tons are in transit (mostly recyclables; valued at $121m. This is the component that some analysts were missing when they attempted to reconcile the inventory value with the amt of raw poly in the warehouse.)
Mr. Lai explained that most of the increase from Q3 to Q4 is inventory that has been recently purchased and is actually still in transit.
It is coming in from suppliers all over the world and so takes time to arrive.
The remaining inventory ($350m less $121m) is made up from the raw poly in the warehouse; work in progress, finished goods, and other inventory required to produce their end product such as Crucibles, Wiresaws, Slurry etc.
All of this polysilicon will be used up during 2008, as described later.
On the $29m non-current inventory, Mr. Lai explained that this product will all be put into use in 2009; once the larger poly plant is completed for 2009 the product can be mixed with their virgin stock. This inventory [~150 tons] composes only 1.5% of the poly requirements for 2009. Importantly, because the inventory will be used, it should not be written off.
We asked Mr. Lai if the auditors [KPMG] are in agreement with the way the company is accounting for inventory and the numbers given. He confirmed absolutely that this is the case.
It is a fact that LDK relies heavily on sourcing feedstock from recyclables for feeding its growth and keeping costs down. This is a strength, not a weakness, of the company’s strategy, a fact that somehow has been lost in the present debate. However, with the company’s auditors onside [and the major review led by an external team that took place last year, exonerating the company], we believe that the audited financial statements in the upcoming annual report will be without concerns from the auditors on this issue.
Is inventory growth too large?
The inventory rose from $225m in Q3 to $380m in Q4 - faster than revenues have risen. The unstated allegation was the growth proves that the company is sitting on unusable feedstock. This issue was possibly the biggest red herring following the earnings release, and should have been put to rest right away.
The company requires 4,400 tons of feedstock to manufacture its pre-sold production for 2008. We know that LDK did not [and still does not] have the full feedstock requirement fully secured at this point. It is therefore absolutely critical that they acquire material when they can – anything less would be highly damaging to the company’s prospects.
Mr. Lai provided the following overview of the present feedstock situation for 2008:
- 1,600 tons of poly are either in the warehouse, or in transit to LDK [see above]
- An additional 1,600 tons are already secured from suppliers to be delivered during rest of year
The above total 3,200 tons, or 73% of total needs for 2008
- Add some internal production projected: 350 tons; makes 81% of the 4,400 tons required
- Thus, the remaining to be secured is at least 850 tons (19%). We gather that the following amounts are under negotiation on price, and are expected to fill the remaining gap.
- 600 tons virgin
- 800 tons recyclable
Put simply, the inventory is rising because they have been aggressively buying more in Q4.
Mr. Lai said that, in an environment of increasing feedstock prices, they needed to secure poly as early as possible.
If we don’t commit to the prices now, it is very likely that later in 2008 we will need to pay more, and that’s if we can buy at all.
Finally, it should be noted that the vast majority of the increase from Q3 to Q4 ($121m) was not in feedstock sitting at the warehouse, but rather in the amount of newly purchased and incoming material, still in transit.
2. Poly Plant Update
Mr. Lai confirmed that the construction of the 15,000 ton polysilicon plant remains on schedule (which includes the production of TCS) and that it is due for mechanical completion at the end of 2008; at that time with an annual production capacity of 6,000 tons. In addition to providing updates on the website, they will update shareholders regularly (certainly every quarter), and we were invited to check in with them as needed, including visiting the site.
"Mechanical completion at year end": this means that the plan calls for all equipment and machinery including TCS production and recycling systems to be installed and ‘ready to go’ by the end of the year.
Given that the TCS production and recycling element of the plant is the most complex and most likely to experience a delay we questioned Mr. Lai on the impact of this happening and asked if a fully operational system was priced into their guidance range for the year. He confirmed that the 42% to 50% margin guidance assumes in house TCS production. However, should they experience a delay in this area they will initially buy in the gas. (They already have an agreement with a local company to supply gas for the smaller 1000 ton poly factory, which should be in production late 2008).
The company has been guiding for a production of 5,000 - 7,000 tons of polysilicon in 2009. This represents an extremely fast ramp-up to production; however Mr. Lai has strong confidence that this can be achieved and pointed to DC Chemical and MC Tech of Japan who, he says, have both ramped up in less time than the LDK plan calls for.
Over the coming weeks, we will continue to explore how [why] the company believes that it can achieve this level of production in 2009. However, LDK is putting in extraordinary efforts on the project; and is working with credible companies. In our view, it would be premature to conclude that the company cannot produce anything in 2009, given the importance placed on this expansion, as well as the company’s records to date. Further, customers such as Q-cells and Hyundai, who are literally banking on LDK’s ability to deliver in 2009, have done their own due diligence and come away with a favorable impression:
"Having visited LDK Solar's polysilicon plant in Xinyu City, China several times, we are confident that both their wafers and new manufacturing facilities will meet the highest industry standards," commented Kwon-Tae Kim, Hyundai Heavy Industries Executive Vice President.
“LDK has been a highly reliable partner and we are pleased to deepen our business relationship with the company,” stated Anton Milner, CEO of Q-Cells AG.“This 10-year contract with LDK ensures that we will continue to receive their high-quality multicrystalline wafers in support of our growth plans.”
3. More on Poly Supply for 2009
In 2009, wafer shipments are expected to double from 2008 to around 1.1GW. An estimated 9,000 tons of poly will be required during the year; that is, 2,000 – 4,000 tons more than they have guided to produce in-house. Further, several analysts currently predict that the in-house production will be considerably lower than guided; that LDK will be unable to secure the required poly for the year and actual results could be as low as half of guidance. We spent time on this topic, and will continue the discussion, to understand the company’s options and the impact of various scenarios.
Mr. Lai said that during 2009, approximately 2,000 – 3,000 tons of polysilicon will be coming from LDK customers who themselves have access to poly – this covers about 30% of the production needs. We think there should be a reasonably high likelihood of this material be forthcoming (customers tend to have long-term contracts; and are essentially the same as in 2008). Further, this source of polysilicon has, at least in some cases, pricing mechanisms such that if the customer-sourced poly price is higher, the wafer price also gets set higher. (This feature was also discussed in the Q4 conf call, explaining why the average selling price was higher than anticipated).
In addition to customer-sourced poly, Mr. Lai said at least 800 tons will be coming from incumbent polysilicon vendors such as Wacker and Chinese suppliers via long term contracts.
Mr. Lai also said that they can keep the recycling factory going at full capacity into 2009, as this is currently producing 2,000 – 3,000 tons per year. This is important as the plan currently calls for the poly plant to replace the recycling factory so they have in effect an in place backup plan which can cover 50% of the factory’s capacity, should it be delayed.
The small 1,000 ton facility has the higher likelihood of reaching its production goal: it is at least 6 months ahead of the larger plant, and will not depend on TCS being produced on site.
Combining the above numbers would suggest that against a requirement of 9,000 tons, LDK can obtain between 5,800 and 7,800 tons of feedstock during 2009, should the larger plant completely fail to deliver. Given the margin protection mentioned for some of this supply, and the relatively low cost of poly from long-term contracts and recycling scrap, it would seem likely that this portion [64% - 86% of 2009 needs] is somewhat secure in both supply and pricing.
In addition to the above sources, Mr. Lai said that they expect the significant increase in worldwide poly production coming online in 2009, will make it easier to source poly, and reduce the spot price.
Based on the above summary, in a bear case scenario where the poly plant cannot produce the targeted range, we believe there is a good chance that LDK can still deliver a majority of the 1.6 GW through external sourcing of the missing amount (they have at least one year to find such backup supplies). Thus, the company should be able to substantially increase its profit in 2009, even in the “bear” case scenario. (This is because most of the production is presold, and is essentially twice the 2008 levels). Margins would obviously be impaired, but it is highly likely that the 2009 profit would, nonetheless, well exceed the 2008 estimate of $200m, a year when margins are tight and wafer sales are only half of what they are expected to be in 2009.
In contrast, if the company is able to ramp up production and reach its initial cost target of some $60/kg for poly for the year, we estimate the upside for 2009 to be substantially higher at the top end of guidance, currently 42% to 50%.
4. Capital Needs
LDK has two credit facilities of significance: one $100 million 5-year facility; and a $500 million line of credit. The latter is a 1-year revolving facility (which is why the debt on this line of credit must be put under the "current debt"). Though the line of credit keeps revolving, it is in our view not necessarily the ideal facility to finance a large capital project. In any case, they have drawn about $260 million on that facility, as of December 31. Therefore, basic math tells us that they have at least $300 million still available from the two facilities.
They estimate that they will put to use about $600-800 million in capital during 2008. However, Mr. Lai said that, while cash flow remains “tight”, they are managing for now and that customer prepayments remain their preferred method for financing the capital projects.
He reiterated the following round numbers for 2008:
- They have about $100 million in the bank
- They currently expect about $300, perhaps $400 million in customer prepayments during the first half of 2008. Another $100 million in the second half.
- They project a net profit of $200 million for 2008 (that is $1.786 per ADR, by the way)
- ...and they have about $300 million available in the existing loan facilities.
The above makes a total of at least $1 billion in cash available for 2008 expansion. Whilst this seems likely to be enough for stated 2008 capital needs, it makes for a tight equation and it is clear that the line of credit is less than ideal as the fill-gap solution. Interest expenses already amount to some $1.5 million per month (assuming a 6% rate on $300 million) and – more importantly – the line of credit is on a 12-month revolving basis. LDK needs to have either another [longer term] loan, or issue equity to reduce the reliance on the line of credit. When asked on this, Mr. Lai said that they are considering their options but that the market conditions at present clearly make equity issuance expensive. (A convertible bond may be most appropriate, if they have to raise cash with a depressed share price; our note.)
5. Status of Class-Action Lawsuits
These appear to be following a standard path. We gather that the 8 or 9 separate actions are currently in the process of being consolidated into one suit. This should be completed within a few weeks. Once that has taken place, LDK can be expected to file a motion of dismissal; and a decision on such a motion can take some 5 weeks. Hence, there should be an update on this coming out in around 2 months from now.
6. Investor Relations
Mr. Lai will be meeting with analysts and shareholders over the coming days, which hopefully will enable everyone to get more firsthand answers to questions raised. This is of great importance: the fact is that LDK has an outstanding growth story to tell, and that nobody really doubts that they will be able to increase their profit [over 2007 levels] both in 2008 and in 2009, especially since most of the production is already presold. The debate is how much more profit they will make, but we believe that the company guidance should be a strong indicator of where they will end up, at least for 2008. Thus, a net result of $200 million (on sales of a billion) for the year represents a 40% increase over 2007 levels. And 2009 should see further increase: even with a much lower-than-guided in-house poly production for that year, their profit increase should be substantial, as noted above.
We believe the real story of LDK is one of growth and upside potential based on smart sourcing of feedstock and rapid expansion. It is a shame that the company finds itself bogged down in defending its business practices, though the ultimate responsibility for improved communications rests with LDK alone.
Disclosure: Author holds a long position in LDK