The Solar Sector: Tomorrow’s Winners Today

I am not any sort of self-proclaimed market guru, nor am I an overpaid analyst riding the markets coattails rather than lighting the way for investors. I, like you, look for the fundamentals whether trading or investing…and frankly I’m a bit confused right now about the current trends in the solar sector. Recently, while researching an article, I created a spreadsheet, a snapshot of the solar sector, to try to get a picture of what is going on with the sector as a whole, and a couple companies in particular. What I found is interesting:

LDK Solar (LDK) emerges as the clear value, with higher EPS, stronger booked revenue and stronger guidance than any other company in the sector. Trading at the lowest P/E, one would think this was a mature stock in the materials sector rather than one forecasting 100% annualized growth through 2009 and already the most profitable by any metric. Further digging reveals LDK Solar is the only fully vertically integrated player in this sector, with expansion under way that will place them as the market leader with a strong competitive advantage; a competitive advantage that will only increase as new facilities come online growing already dominant margins.

While other players in the sector demonstrate great promise for long term growth, much of that growth is already priced in. Trading at P/Es ranging from 185 to 640 First Solar (FSLR), Yingli Green Energy (YGE) and SunPower (SPWR), respectively, will have to deliver on forecasted revenue growth for several years to justify their valuations. When a company is earning $0.20 per share and trading at more than 600 times earnings, I’m not sure how any analyst can keep a straight face while asking you to buy his shares of that company and sell him your shares of a company earning $2.76 per share and trading at only 13.5 times earnings.

I’ll leave the speculation on Wall Street’s motivations for guiding you in the sector for another story. I can tell you one thing though, on Main Street that’s not how we pick’em.

Disclosure: Author has a long position in some of the above-mentioned securities.

Comments

[solar jim:]

Actually, LDK very recently, four to five weeks ago, was at $76.75/share. Using that metric LDK had a market cap of a little over $8billion at $76.75/share and a p/e of 59.9 on 2007 consensus earnings estimates. And roughly 40 times 2008 consensus estimates. At todays close (Nov 1) the stock is roughly half what it was a month ago. The reason it dropped was the concern of the allegations that were made by a departing financial controller. Whether true or not, that more than anything else is the reason for the drop to the $40 level.

Second, I'm not sure of your estimates for the companies you mention. At least two of them seem to be in error. SPWR currently has a consensus estimate for 2007 of $1.23 and 2008 of $2.03 for 2007 and 2008 p/e of 101 and 61 respectively on todays close of $126.16/share. FSLR has consensus estimates for 2007 of $.64 and 2008 of $1.42 for p/e's of 242 in 2007 and 109 for 2008 based on todays close of $155.35.

And where does the LDK number of $2.76/share in EPS come from? Is this for 2007? 2008? 2009? Consensus estimates for LDK in 2007 are currently at $1.28 and 2008 of $1.88.

I quickly looked at the others and they seem to be incorrect in your chart as well from consensus estimates.

Every EPS and p/e is wrong in your chart. LDK seems inflated by a good amount over consensus estimates while all the others are shown with quite a bit less than consensus estimates. Could you be long LDK???

While I am long two of the ones you mentioned (FSLR and SPWR) I don't need to manipulate or misrepresent numbers to make a bull case for either of them.

2007 Nov 01 06:00 PM

[User 86999:]

my gosh, so much fud? does this happen in other posts on this site too? are these people paid for this propaganda misleading effort?

2007 Nov 01 08:37 PM

[solar jim:]

It may be an honest mistake, but I've seen other somewhat misleading claims as to LDK margins in comparison to other solar companies that was in error as well in another article he posted. If not misleading then very sloppy, at the least, in actually doing any math to find out the actual percentage difference between companies before he states unequivacably that LDK has margins twice that of their closest competitoror, which is NOT true. When you see this it can be an indicator for someone who is long and wants out and is down from their buy point. If you can't use real data to make your case, or show/tell/explain where you got it then its useless and meaningless.

2007 Nov 01 10:29 PM

[Michael Sha...:]

In checking his numbers against Google Finance and Reuters it appears he is correct. While you may be long on two of the higher priced equities in question, falsely attacking on in order to help you case does no one any service. I checked the authors site...he posted a correction. Google Finance erroneously calculated EPS based on float rather than total...Mr. Shaffer posts the corrected .29 along with 31.1% profit. According to Reuters the P/E is correct. Rather than quickly looking, as you state, you should take a little more time before you attack and informative article. One would think you have an agenda.

2007 Nov 02 11:17 AM

[solar jim:]

Also, TSL is definitely vertically integrated as well - not just LDK.

2007 Nov 01 06:03 PM

[Ellen:]

Thank you for keeping the posts clean and fair. This is not a place we want a long-holder to be misrepresenting facts to drive us into buying into a stock based on incorrect data, please!
thank you Solar Jim.

2007 Nov 02 04:09 AM

[Michael Sha...:]

I just checked Reuters and Google finance and they backed up his numbers. You guys sound like you have a short agenda where LDK is concerned.

2007 Nov 02 06:59 AM

[solar jim:]

And actually, I have no agenda with LDK other than seeing that someone is pumping it up with false information, which TO ME, looks like someone is uderwater on LDK and trying desperately to pump the stock to get out at a better price. I have zero short positions and have no interest at all what LDK does as a company or which direction their share price goes. I happen to own two of the companies this guy has posted on and the information he is using was extremely erroneous. So then I checked his LDK stuff, as it seemed he was long the stock from all the cheerleading going on. Again, I found the earnings information to be erroneous, although with this one it was inflated instead of all the others being deflated. Of course it may be a simple matter of not being able to do simple addition, subtraction, division, and a little mutliplication. Or he simply took the chart from somewhere else without doing any kind of verification of the numbers he was using. Either way, sloppy, sloppy, sloppy.

2007 Nov 02 03:10 PM

[solar jim:]

Again, I have no idea where you guys are getting your numbers. Are you talking about trailing 12 months earnings? Last fiscal years earnings? This years earnings? Next years earnings? One quarter of earnings? The numbers I used are consistent whether you look at google finance or yahoo finance, and actually you know, click the link called earnings estimates. Tough to do, I know, but worth it. If you actually do that, you see that FSLR is selling at 102 times NEXT years earnings and 200 on THIS years estimates.

Hey, here's an idea... actually check the last four quarters of reported earnings for SPWR. Lets see... 4th qtr Dec 2006 $.18, 1st qtr 2007 $.29, 2nd qtr 2007 $.25, 3rd qtr 2007 $.33. So again, where is this $.20 EPS number for SPWR coming from????

First off, maybe you don't understand how p/e ratio's are actually calculated. You take a years worth of earnings, like SPWR's last 4 qtrs, or $1.05 in the last twelve months. Then you divide that sum, or $1.05, into the share price. Using the price at 11:54am (Pacific Time) SPWR's P/E on trailing 12 months earnings is 117. Even allowing for the original article's author erring in using a single qtr of EPS to calculate the p/e ratio, please tell me how you get $.20 per share EPS for SPWR when no quarter they have had is at $.20/share. Frankly, I'm puzzled. It is just factually false however you look at it. SPWR's earnings are estimated to nearly double next year from this years so the p/e is about half of todays on 2008 earnings. Nowhere near 639.67 p/e ratio however you slice it by one quarters of earnings, or however. Unless you have some magic ratio formula you apply to all estimates that allow you to come up with estimates that just don't exist elsewhere. And if we ARE using one quarters estimate to determine p/e lets examine last qtr's reported number for LDK, or $.29/share. That gives LDK a 138 p/e at just over $40/share (today's price).

If I didn't know any better, and I don't, I'd guess that this post is just the original author signing in under a different name. Please post the actual link for these "earnings estimates" you say are correct and match the original authors posted "spreadsheet".

I used Yahoo Finance pages and the earnings estimates aggregated and the listed average earnings estaimte. Checking on Google finance as well as Marketwatch the estimates provided were the same ones I originally used in my post.

Again, people, please do a little homework after reading a post like the original authors to do a quick check on the estiamtes they are using. Please check the numbers I am using. I have absolutely no problem as they will prove out. The estimate numbers I have posted are the ones being touted as consensus estimates on Yahoo Finance, Google Finance, and Marketwatch.

2007 Nov 02 03:02 PM

[Special Sit...:]

The solar industry has serious polysilicon supply constraints, as their is a steep learning curve in polysilicon production, and high fixed investments in plant and equipment, creating high barriers to entry. Thus, the power in the industry is currently concentrated in the hands of polysilicon suppliers, including MEMC (WFR), and Wacker Chemie (WCH.F). Downstream names that are worth looking at are SPWR, STP, ESLR, FSLR, and ENER. My best advice for fast growing, emerging industries is to buy the suppliers. Solar machinery is made by Roth & Rau (R8R.F), while the afformentioned MEMC and Wacker Chemie are excellent plays on polysilicon and are making money hand over fist. For more information, go to solarbuzz.com, an excellent, free industry website.

2007 Nov 03 02:31 PM

[Vince Sanchez:]

You can check the "Solar Stocks Comparison Table (11/2/2007)" at: www.cnanalyst.com/sola....
Also, I checked Google finance and article's numbers are ok

I assume differences depends on sources and dates.
Besides numbers, a qualitative analysis is quite important to get overall sense of the sector and individual performances.
We'll get results on NOv 15th for LDK, then see who is correct.

2007 Nov 03 04:20 PM

[User 119825:]

Where did you find the date Nov 15 as the day they are reporting earnings? I've looked everywhere and can't find it. The last time they reported was Aug 1st and I assumed they were reporting on Nov 1st..Some even said it was Nov 7 or Nov 8. I'm getting too many different days. can you show me a link that says it is Nov 15? Thankyou

2007 Nov 04 01:11 AM

[Vince Sanchez:]

I got the date from a broker, who told me he found that date at bloomberg .... but not sure, since I haven't find the original link.
In google finance there are some more "good" discussions about LDK, see: finance.google.com/gro...

.. by the way they posted:

From Dow Jones News....


LDK Solar Co.Ltd. (LDK) Corporate Event Announcement Notice

11/01/2007 17:04


Nov 01, 2007 (Wall Street Horizon via COMTEX) --
LDK Solar Co.Ltd. (LDK)
Expected next earnings release:
Announcement date: 11/8/2007 - After Market
Earnings Quarter: Q3

2007 Nov 04 06:05 PM

[solar jim:]

I'll try this one more time. Further checking and trying to find the EPS numbers in the article I think I have discovered the discrepancy. Yes, Google Finance does list SPWR's EPS as $.20/share. I believe this is fully diluted earnings per share and on a trailing 12 month basis. Yahoo Finance, Marketwatch (the link provided on Google Finance for earnings forecasts), ScottTrade (my broker), and most of the other sites that list earnings forecasts do so on a pro forma basis for the forward fiscal year or forward 12 months.

Using diluted earnings on a trailing 12 month basis is not the best way to view, or value, a young company that is in power growth mode. This is why the majority of analysts, traders, hedgefunds, etc..., use pro forma earnings. Pro forma earnings allow an investor to see what earnings would be without extraordinary items.

As an example, say you are a real estate developer and you want to develop a property into an office tower. At the end of construction and in your first year, even if fully leased, the costs for materials and labor as well as tenant improvements will drive you to a negative or very small first year NOI, or income/profit after all costs and expenses. Buildings are sold and bought based on capitalizing the NOI. If you sold your building the first year based on including these extraordinary costs in your finance model you'd lose your shirt, pants, hat, and every other item you own. So what actually happens is you either take out the extraordinary items and capitalize the income that you would get in the second year without those expenses, or you take a multi year income stream and discount that to todays dollars (Discounted Cash Flow) as the out years of your income stream do not have those current year extraordinary expenses. The same idea works for companies that build plants that last for decades.

Either way, you get a value that is not based on diluted earnings where the expenses are higher due to new capacity expansion. Once that plant is built it is good for decades. You don't have to spend that money on that plant every year the company is in existence. That is why it is better to look at pro forma earnings as opposed to diluted trailing 12 months earnings. While diluted earnings are a nice historical account, what does it mean, at all, going forward?

2007 Nov 07 10:16 PM

[solar jim:]

Enter your comment hereLet's try another way at looking at this chart above in the original article. FSLR has diluted earnings reported of $1.35/share by GOOG finance on trailing twelve months. This is after the 3rd qtr of reported earnings. But again, do you buy a stock based on what they've done, or what they will do in the future. Anyone who has watched the market for a long enough time to see a cycle or two, knows that the market is a forward looking mechanism and the past is not at all as important as what you say you will do going forward. It's a game of "what have you done for me lately"?

Consider the above table in the orginal article. Now that we are one more qtrs data forward from when this tables data was collected (I assume it was based on data through the 2nd qtr of 2007) it is interesting to see the updated results for FSLR at the very least. FSLR had diluted earnings of $.49/share for the last qtr. That put's them at a diluted earnings per share of $1.35 for a trailing 12 month p/e of 141, which is much much less than the 185.59 p/e indicated in the above chart. Considering the price of FSLR on the day the article was writtten, or $155/share, that indicates a p/e of 115. This indicates a one qtr decrease differential of 36% based on one qtrs of earnings. So after you do the math on the other companies after releasing preetty much in line numbers it is easy to see that while FSLR or SPWR may still be more expensive, the differential between a FSLR/SPWR and the also rans is getting much closer due to the fact that the FSLR/SPWR combo is increasing earnings, whether pro forma or diluted, faster than the rest of the bunch. Which leads me into the very reason that they are valued higher than the other, they are better and make more money. It's as simple as that.

2007 Nov 15 06:29 PM