Investing Hobo

Investing Hobo
Contributor since: 2010
There was a editing glitch. The last paragraph should have been the first.
There was a editing glitch. The last paragraph should have been the first paragraph.
When comparing costs for electricity especially across various competing formats, it is best to use LCOE as I described above. The cost per watt varies greatly between coal, oil, natural gas, hydro, nuclear, wind, solar, etc. Some have higher fixed costs but little operational costs, such as solar. Others have low fixed costs but higher operating costs due to constant fuel consumption. LCOE is the best method of comparing costs because it simply gives the cost of electricity over a period of time usually the expected operating lifespan of the power plant.
The point of this article was simply to show what utilities are willing to pay for electricity of various forms in California. Specifically, it is to show renewable energy particularly solar PV has reached a cost point directly comparable to natural gas. It would take much longer to describe the situation in other parts of the world because there are so many varying factors. As noted in a reader comment, most of Hawaii's electricity is from power generated from imported oil and thus is much higher cost than natural gas plants in California. That is also why retail electricity rates in Hawaii are also significantly higher in nearly all other parts of the USA.
There are also many other factors which are not discussed in great detail in this article. Some examples include as also mentioned in a reader comment, solar power is distributed power generation. This mean it can be built where it is needed, often saving on distribution costs. In addition, solar produces electricity during peak hours when electricity rates are generally higher. These factors for example have not been factored into basic calculations above.
Solar PV is not viable everywhere nor can it replace other electricity power generating sources. In certain regions depending on various factors, solar PV's costs can be economically viable without subsidies and thus be utilized in conjunction with other sources. This critical mass point has at least been reached in California and that is the point of this article.
Non-silicon module conversion costs have roughly 10% labor, 10% utility, 10% depreciation, and 70% raw materials. These percentages are further lowered since polysilicon still makes up 25-30% of a module's unit cost. The process is highly automated and anyone can check out demonstration videos at company websites. Wages haven't been disclosed by if you take companies at their word, it's much higher than "minimal" because operating advanced equipment requires slightly more skill than assembling toys for example.
As for utility, Chinese companies pay industrial rates up to 75% higher than residential rates, a flip contrast to western countries. One can argue if these usd $0.11-0.12/kwh industrial rates are too low but you might also want to research what polysilicon producers in the US pay for electricity. Take my word for it, it's considerably lower.
As for depreciation, that's per US GAAP standards the same as any US public company.
If anyone wants to really know why China is so efficient, dig deeper into its pv raw materials supply chain vs any US/Japanese/Western peer. A superior raw materials supply chain is simply the key reason why its manufacturing base is far more efficient than most other countries.
I can't and won't comment on how the market values any company at any given point in time. As I mentioned, I have discussed PWRD in previous articles so please refer to that. As for the price adjustment ex-dividend, that always occurs - stocks drop by the same amount as the dividend amount. You can research dividend terminology to understand how it's done.
While management has stated they do not look at its share price, they have proactively taken steps they believe are in the best interest of shareholders. In the past three years, they have repurchased almost 13m shares, reducing the share count from around 60m to 47m. Of course recently they did a $95m dividend payout. Had they not bought back any shares in the past, this payout would have been closer to 1.50/share instead of 2.00/share. They indicated that annual dividends would repeat and be dependent on the company's earnings. Thus if earnings remain stable and dividends repeat, holding PWRD stock would yield about a 12-14% return, regardless of how its actual share price performs.
Lastly to answer your question, the company bought a building about 4 years ago. It's a fixed asset that will likely be more or less neutral towards earnings since it's unlikely to be sold. In other words it's more a capital investment than the way you described it being more an expense.
You can read some of my previous articles on PWRD. At least from my point of view that's taken from management's consistent comments in the past, the company is not as concerned about maximizing near term profitability. They are only looking at building a leading brand not only in China but globally. In order to achieve this, they are leveraging the company's extremely rich cash flows and investing it in assets which may return yields in the future. Of course until all these investments yield results, PWRD's earnings will look weaker than peers which spend less, hence the drag on its relative share price. I sense management will not change their strategy so its stock may not be ideal for short term investors. At least they have been proactive lately and rewarding longer term holders with fairly high dividends.
It would take many pages to answer your questions in full detail which I will not do here. You do raise very valid questions and what I have tried to do is explain the industry through my articles.
The vast vast majority of installed solar power is due to subsidies. Without subsidies, solar would not have been economically viable at the costs installed in the past however many years you want to go back. That is a fact and I do not dispute it. Subsidies are not new to any industry and the point was to help the solar industry achieve high enough scale to reach critical mass in terms of costs. You can either do the research yourself or take my word for it, costs have reached levels where subsidies are no longer required in many markets. Subsidized FIT rates falling well below grid in Germany is only one example.
This applies to FSLR which is among the industry's cost leaders for total installed project cost. The specific project you are referring to had terms arranged several years ago. If planned today, I'm sure prices would be much lower. Utililties tend to pay higher rates for renewable power because the government at various levels requires an increasing portion of its portfolio come from renewable sources. It's also often not discussed but pricing is often equalized for environmental effects such as pollution. Look up carbon credits for an example. This makes installing new renewable capacity just as attractive as shutting down high carbon emitting coal plants. Once again research the rate coal plants are being shut down across the country. Thus at $0.15/kwh by itself seems subsidized, above market, losing money, etc whatever you want to label it. Other factors including, utilities which are profit generating entities generally do not look to do bad business. Lastly you also have to consider this rate is fixed over 25 years. What is a higher rate TODAY may be a much lower rate in a few years. Want for details, research how to calculate adjusted levelized costs when inflation is considered. On average over 25 years, the utility has planned on generating an overall degree of profits from this power purchasing agreement.
I hope I gave enough examples for people to understand things a little better. It would take many pages to fully describe it in methods the casual viewer could rationalize.
I have tried to make a point of not replying to comments in my articles but on this issue and especially when facts are portrayed inaccurately I have to take exception.
Buffett's solar farm has a ppa of $0.15/kwh for 25 years. That is not 3-5x market rates in California. If you live there you would know rates average around $0.12/kwh but can go much higher during peak hours. Because solar power is generated mostly during peak hours, comparing costs to average rates can be misleading.
As FSLR has shown, solar can already be cost effective domestically. My argument was not to support Chinese solar companies because if it could be done within the US at a cost effective level, of course that's best for Americans. My argument was to show a large portion of the value chain does not go to Chinese companies. Chinese solar companies are simply more efficient manufacturers due to logistics, vertical and horizontal integration - steps many US manufacturers tend to overlook witnessed by SPWR's horrid supply chain of yesteryear which transversed across 4-5 different countries. As history has already PROVEN, solar industry trade between the US and China has been to the advantage of the US. Any non-market disruptive influences may penalize US solar industries which have effectively competed with global and Chinese counterparts prior.
Whether the US wants a major role in the solar industry is up to us Americans. It's clear solar installations has expanded by large degrees in all parts of the world including China. By politicizing the issue even with token tariffs, the US would only marginalize its already successful solar companies on the world stage. In the long term it would only mean the US becoming another laggard in another booming industry.
According to the conference call, the money will be distributed in early/mid April. From there the nasdaq sets the ex-dividend. My guess, in about 4-6 weeks. If you bought in Feb and hold it through the ex-dividend date, $2 per share will show up in your account. Keep in mind on the flip side the stock usually drops about the same amount as the dividend. This is only meant to reward long term investors with periodic income. Short term oriented traders may not care about the dividend and probably should exit the stock especially on gains prior to the ex-dividend.
Not really. DQ's earnings included almost $4m of government grants which I did not include into my estimates. Without that, the net income line would have been similar. As I state in all my estimates, I calculate only operational numbers. Other non-operational figures that cannot be predicted are excluded. I'm not saying DQ ins't cheap, because it is. However realize its earnings are and will continue to trend down as I have stated in previous DQ articles.
The point I was trying to make in the article was not to endorse any particular company. CSUN was used as an example because it was the first among the group to literally trade below 1x trailing PE. Had another company achieved that milestone first, the article would have named it instead. What companies I own or do not own may be a better representation of my sentiment if it's not already implied within the article.
It's tough to pin any number for any country or region since climate can be fairly diverse. I'm sure southern France is much different from northern France. Even for California which gets fairly good sunshine everywhere, there can be a big difference. STP's 150mw project in California for example is expected to yield 2.3kwh/y.
At first I was a bit skeptical when reading Chinese local government claiims on their region's annual sunshine hours, but the Dunhuang project demostrates close to 3kwh/y is definitely possible. Contrast that with costal China and you get something more on the order of 1.5-2.0kwh/y. An FIT rate of rmb 1.15/kwh would be overkill in the western desert regions, but would clearly be required for costal China at least if IRRs are expected.
Read all my articles and disclosures regarding my estimates. I do not speculate outside of information provided directly or indirectly by the company. If companies I make estimates on due to previously disclosed information make guidance adjustments then most likely my estimates will also have to be adjusted. In this article I cited reasons to adjust estimates on comments the company previously made due to information provided by the market. However I still kept other metrics within the parameters provided by the company.
It's a sentiment observation no more or no less different than technical obervations traders make. Perhaps you didn't notice but I did differentiate tsl from being a potential fraud vs other companies which have been targeted more recently. In any case I've held tsl since 2007 and don't plan to sell in the event it might appreciate due to any perceived "pumping" articles including my own.
"I think some folks definitely need more education before they post."
You might want to take your own advice "beststockpicker." Just review not only the sheer number of your comments, but your statements regarding CSUN and DQ in my articles. I don't make calls on stock price activity, but perhaps you should have paid more attention when I heeded caution on gauging companies on unsustainable metrics.
DQ is likely to produce more than 7mw of wafers in Q2. I am assuming like you are a certain portion of produced poly will be used for internal wafers which will either be sold or tolled into cells. In the estimates I made, I am assuming around 8mw of additional internally produced wafers will be tolled. The remaining cells would be purchased.
As for Q3 and Q4, it's too early to know the exact product mixes other than the vague reference DQ provided. Generally I believe per watt gross margin will be slim in both wafer and module verticals so when I calculate consolidated gross profit, I'm purely looking at gross profit potential at each vertical then adding everything up. However internal production is either consumed or sold should ultimately add up to roughly the same per watt gross profits at each vertical capacity. Based on what management stated, the blend between tolled/purchased cells might be around 50/50 ratio. As I noted, I think Q3/Q4 EPS will fall roughly in the .55-.65 range assuming 50/kg poly asps.
No idea on their second plant in Xinjiang. They just started construction so it's way too early to estimate anything. DQ's general guidance is better than anything I can guess at. I'm guessing the ultimate capacity will be similar to their first plant however at around 4400mt once at full utilization.
Read all my articles regarding LDK if you want my position. I have detailed my views in full. Of course, they are just my opinions and interpretations. I'm sure there will be many who will continue to take opposing views so I'm not going to try to convince anyone otherwise. If you really read what I wrote carefully, you will find the answer to all your questions. From the content of your comments, I believe you have already made up your mind however.
In past articles I've already tried to differentiate PWRD's model vs. other Chinese online gaming peers. PWRD's approach on basing games based on popular historical or fiction novels is not entirely new to the industry and success cannot be implied just from this factor. Their branding and marketing approach has also been much more creative than peers, but if you have followed the company since their ipo several years ago, novel marketing concepts also have spotty success rates. My general point is to let viewers know PWRD's business model is highly leveraged on even a single new game hit, but until this is realized through earnings reports Wall Street has generally been more cautious in discounting incremental revenues from new sources. In the meantime, continued profitability and low valuations offer some downside support.
The realized diluted share count is normally based on weighted average diluted shares during the quarter. As a result, days prior to the finalization of their recent secondary, you use about 135m shares. Days after the secondary you use 149m shares. These are approximate figures and you can get the exact number from their annual report. So on a weighted average basis, it works out to (x days * 135m + y days * 149m) / (x + y aka days in the quarter). As for your other number, I haven't seen it so I'm not sure where you got it. The final number may be a little higher due to option grants, but I don't think it goes as high as 158m.
LDK did not sell 18.whatever% of the polysilicon division earlier this year. They sold convertible notes which when converted will represent that percentage stake. It's not the same type of transaction as in late 2009. As for your other opinions, it's a free country.
Whenever LDK is ready to announce a date, they will do so with usually 1-2 weeks of lead time. Until then, don't pay attention to any dates out there. They are all unofficial and essentially guesses which often turn out false. If anyone tries to make a big deal over reporting dates, then it's usually because they have other motives.
Companies have 90 days to report prior quarters numbers. That means LDK could report as late as the end of next month. Last quarter they reported in mid March, or about 9-10 weeks after the quarter ended. So far we are around the 6th week after Q1 ended. Sometimes companies report a little earlier, and sometimes a little later. It's as simple as that and there may be no particular reason whatsoever. So my advice is just to wait and not pay attention to any other sources of information until they finalize the date and report. There isn't a need to speculate why, good or bad.
There really isn't a clear cut breakdown regarding who is tier 1 or who is tier 2. It's just a generalization on which companies are closer to the core of the industry and as a result less vulnerable to down cycles than others.
Generally tier 1 companies do business with other leading companies in the industry. In the case for module producers, tier 1 companies would sell to customers who have more direct exposure to module consumption.
For example, a utility company may already have more than a year of backlog of solar projects to be built out. Whoever is supplying that utility company have their sales volume more relatively secured. Other tier 1 module consumers may include EPC project developers fulfilling work for investors seeking a certain rate of IRR, generally from FITs. Large distributors may also be classified as high tier customers although generally not as reliable as the previous two mentioned examples.
An example of a tier 2 or tier 3 module consumer may be some small regional distributor. If a certain market freezes up, smaller distributors have less avenues to redirect their inventory and thus may hold back on further module purchases. That would obviously affect the module producer. The module producer who have higher exposure to categories of this nature may be considered more as tier 2 or lower.
There are also many other factors which differentiate who may be tier 1 or otherwise. Cost is also a key factor because the lower cost producers can place their products more competitively. Larger companies with scale or financial backing may also be considered higher tier because they are more bankable in nature and thus institutions financing solar projects would more likely finance certain brands over others.
These are just a few examples but I hope you can get an idea what I mean.
For LDK in particular, I wouldn't classify them in the highest tier yet because of their lack of history with end markets. STP, YGE, and TSL have been doing business for years and have developed extremely strong sales channels with leading downstream consumers of their products. LDK does have avenues to place their own products through their own systems division, but the strength of that portion of their business is untested imho. Due to scale, cost, and financial backing, LDK perhaps right behind the big 3 listed above among Chinese companies. If their systems division develops into a capable arm, it would put them right near the top. However it's too early to say although LDK's module performance this year will give a very good idea on their progress.
I own a number of stocks but am only required to disclose positions of companies I discuss in each article. Prior to last year solar companies only represented around 10% of my overall portfolio. For many years TSL was the only solar company I owned. Currently solar companies represent over 30% of my overall protfolio, reflecting my forward bullishness due largely to valuations. JKS in particular is a new position and a relatively small position relative to my other solar positions.
I'm not negative on SOL nor do I let personal bias affect my objectivity despite whatever anyone may think. In the end SOL is trading at around 4-5x earnings. To be negative at such valuations when their level of earnings should stay relatively stable over the course of the year would simply be wrong in any context of investment philosophy.
For those who have followed me over the years knows how much emphasis I put on management teams. Maybe it sounds too simplistic of a generalization but I have often found it true. Managements that execute well tend to continue to execute well even during negative business cycles. Managements who are more disconnected with their company or industry tend to increase the risk variables in their company's performance. My track record should be well known regarding corporate performances on managements I have been critical of in the past.
The point of the article was not if SOL's management is right or wrong about where module asps end up in 2011. In fact lower tier providers who may be forced to liquidate may very well sell at such lower pricing, but it's not reflective on the major players within the industry. However by defending their position where SOL can maintain expected margins while their customers all fail has been historically wrong. It should call into question management's judgement on the industry and quite frankly their own business. Whatever the business metrics end up being towards the end of the year isn't the point and I didn't try to predict or argue it beyond the current quarter.
My views were apparently also shared by several other analysts as they repeatedly asked if SOL truly believed in their statements.
All of my solar articles talk only about the companies and news that may affect their businesses. I have never speculated on how LDK "the stock" would perform. That is up to each reader to decide for themselves. My track record goes well beyond the brief period my articles have been circulated on SeekingAlpha, and many will confirm the accuracy of my company analysis over the years. If you read enough of my articles and compare them to other "bloggers" who may use hype or scare tactics in their opinions, you might see the difference.
So I'm not sure what you are referring to when you state "our author." If you are referring to this article, I am only outlining how a possible ipo spinoff LDK's units might affect their balance sheet. From there it is up to the reader to decide for themselves if this event is a postive or a negative. In my other LDK articles analyzing the company's earnings potential, I only analyze the items which may affect the absolute level of the company's earnings. I make no statements on price targets as a result of any earnings or if any stock is "cheap" or "expensive" other than indicating a stock at a 5pe is on the lower side of historical means.
To be perfectly honest, some of the other companies you suggest are much more speculative in nature. In the end good investments are not about hype or promise about any industry or how our future is headed. It will be and always will be in the end related to the level of earnings any company can produce. Renewable energy will most likely be a key industry moving forward but along the way there will be just as many failures as successes within the industry.
Credit Suisse holds CSUN shares because they underwrote their convertible bond offering in 2008. Almost 4.5m shares were issued and loaned to Credit Suisse in order to allow bond purchases to short shares as a hedge. Their real net exposure is probably closer to 1.5m shares based on the 6m shares you referenced.
You are entitled to your own views about the company. It's my impression that your views are not objective and bias since you choose to look at the potential positives without weighing the related negatives. Looking solely at cash position without considering debt is a good example of one dimensional opinions.
In fact I assumed 350mw of internal wafer production which is even above the numbers you quoted. I accounted for non-internal produced cell module production as cell tolling agreements. You should be able to tell because the unit cost is well below the cost of purchasing cells in the open market and apply module processing costs. Of course they could sell more wafers and buy cells instead, but the gross profit impact would be similar despite higher stated revenues. You do agree that 350mw is full utilization or even above, correct?
You are welcome to send your inquiries to KPMG. As a big4 auditor, I'm sure they will take any accounting irregularities you see very serious. As it stands, they did already audit the way LDK accounted for the transaction and apparently it confirms to US GAAP accounting.
As for the other issues, I've already discussed my opinions with you in the past. I do not have insider information on LDK's operations nor have I ever seen their books on an internal level. I can only discuss what is known and trust reputable auditors are doing their jobs. I rather not speculate on issues I do not have detailed information on. What you appear to be implying is that KPMG is missing out on some details which you as an outsider who I'm willing to bet have never had access to LDK's internal books knows more information.
I probably should have addressed the shareholder equity issue you raised in the article but I can briefly describe the events here. Keep in mind I do not have access to their books so what I describe may not be an accurate account of what happened. It is what I believe took place or at least to a similar degree.
Look at LDK's Q4 2009 earnings report. It was also the quarter LDK sold 15% of their larger polysilicon plant for a sum around 220m if I recall correctly. In this quarter, LDK showed a shareholder equity gain of 200m, despite posting minor losses for the quarter. This is what may have occurred. The way LDK accounted for the 15% stake sold was off the books in some fashion. I don't know the exact details. However the 200m+ LDK recieved from the divestment was added to LDK's balance sheet as assets. Hence you have a net shareholder equity gain of the capital injection (minus whatever fees or items that may have been the difference in the sale price and change in shareholder equity).
Now look at their Q4 2010 earnings report. It was also the quarter they bought back the same 15% stake in the larger polysilicon plant, although the repurchase wasn't reported until early Jan. Now look at how shareholder equity changed. Combine the change in shareholder equity plus the net income LDK earned in the quarter and you get slightly over 200m loss in shareholder equity. Why? For the exact reverse reason as explained in the prior paragraph. When LDK repurchased the stake, the 200m+ was taken off the balance sheet to buy back the stake. As a result, 200m in shareholder equity was taken away from prior quarter levels.
Apparently the way LDK accounted for this minority divestment was acceptable under US GAAP accounting. Remember LDK is a major company by international standards and not like many smaller Chinese firms who may have more questionable practices. LDK uses a big4 auditor, KPMG and the 15% asset sale has already been audited as acceptable. My guess is LDK may have issued some form of convertible note to the purchaser which didn't immediately reflect the divestment but could have under certain conditions. For all purposes, it was like owning 15% of the polysilicon plant because they still got minority payments but technically no real assets were transfered.
No offense since everyone is entitled to their own analysis, but some of the numbers I've seeing do not match company statements. Yingli for example has not reached 5.3g/watt silicon consumption. It was 5.8g/watt ending 2010. Yes their Panda silicon consumption may be 5.3g/watt but not their blended metric. Assuming half Panda capacity by year end, the blended rate might be something in between 5.3g/watt and 5.8g/watt, not anything towards 5g/watt as suggested.
That's just one example as I saw other minor points in the conversation between the two of you. Some metrics have been stated outright, so I'm not sure why other numbers were used. In Trina's case, they stated their gross profit for in house production at .66/watt, so I'm not sure why .63/watt was used. Stuff like that so maybe some figures need to be doubled checked. It's good that everyone is starting to dig deeper into the dynamics of the industry on their own however.
In the example I detailed, I actually assumed "42%" of the existing base would be sold as new shares. That was the maximum rounded off number I calculated and it's slightly ahead of the 40.66% you noted. However after reviewing what I wrote again, I realized I made an error in summarizing my example. 10.5 our of 42 is 25%, not 20% I stated would be ldk's selling position.
The 31.5% number is perhaps a bit misleading. Perhaps I should have clarified that it is the largest portion ldk could sell on its own without issuing new shares. This may still be the case since my scenario was only meant to be a reference example. Generally ipo spin offs of other Chinese companies I follow involved issuing new shares so that's why I decided on an example with this structure.
According to company statements, they show no intention of fully vertically integrating. My opinion from following the company since their ipo in 2007 is that they truly believe in their operating model. Their downstream module integration is only to suit certain customer requests. Their continued commitment to large long term wafer supply agreements suggest they are only willing to take upstream wafer integration to a partial level.