Further Musings on Solar Stocks

by: Jack Yetiv

Yesterday, my third article on the solar space appeared on these virtual pages. That article built upon my two previous articles, which offered some forward-looking views on the solar space in general and concluded that Trina Solar (NYSE:TSL) offered the best risk-reward ratio in this space.

I submitted a table I had put together, which calculated forward PEs for 11 stocks in this space. The table is set out again below (with JA Solar's (NASDAQ:JASO) price "% change" corrected after some sharp-eyed contributors found my math error in failing to account for JASO's 3:1 stock split).

The table shows that as of the April 22nd close, three of the solars (LDK Solar (NYSE:LDK), Solarfun (SOLF) and Akeena Solar (AKNS)) are lower now than they were three months ago, and five have gone up more than 10%. Two are about even money - Suntech Power (NYSE:STP) and Yingli Green Energy (NYSE:YGE). [Note: Caution, the solar stocks on the morning of April 23 (the time of writing) were moving around a lot, so some of this may not be accurate by the time you read it today]. The leader has been First Solar (NASDAQ:FSLR), which has gone up 74% (from $171 to $298). In second place is Canadian Solar (NASDAQ:CSIQ), which increased 41% from $18.56, where I recommended it to $26.14, where it closed yesterday. Third was JASO, with a 28% increase, SunPower (NASDAQ:SPWR) went up 23%, and TSL increased 18%.

Because CSIQ has gone up so much, it is no longer as much of a bargain as it was when I recommended it—its 2008 PE has increased from 11 to 15. Because TSL's price hasn't moved as much—despite the fact that earnings estimates for 2008 have increased substantially (from $2.83 to $3.39)—TSL is now the best bargain in this group, sporting a forward (2008) PE of 12.

Other metrics on TSL are also quite compelling—TSL is projected to increase revenues from about $300 million in 2007 to $800 million in 2008, and to more-than-double its earnings from $1.54 in 2007 to $3.39 in 2008. One analyst believes that TSL will make $4.19/share in 2008. Another positive for TSL is that it has secured much of its polysilicon requirements from 2008 til 2015 (TSL has secured 95% of its 2008 requirements). Finally, TSL still has a long way to go to revisit its 52-week high of $73.

One of the reasons that TSL dropped (during the January solar bloodbath) more than some of its peers is because Trina was planning to build its own poly factory at a cost of $1 billion. This was viewed by me (and obviously the market) as a major negative, for three reasons. First, it wasn't clear where TSL was going to get the money to build this poly fab. Therefore, there was the fear of stock dilution or of punishing interest rates in order to get enough cash to build the poly plant. This was one of the reasons I did not like TSL as much as CSIQ the last time around even though the PE for TSL was pretty close to that of CSIQ. My second reason for disliking TSL's foray into poly manufacturing was that it would de-focus management from the main task at hand—making and selling topnotch solar panels at the best price possible. Finally, I was very concerned that by the time that TSL made any poly in that plant (several years from now), poly would be so cheap and so available that TSL's effort would have been for nought—if not worse.

In any case, finally coming to its senses, TSL management announced about two weeks ago that it had canceled the plans to build the poly factory, causing the stock to go up several dollars that day, and over the past two weeks, TSL has established a new base in the low 40s. It should be noted (although I am not positive about this) that the average earnings projection of $3.39/share may be based on estimates that have not been updated since the cancellation of the poly fab. If cancellation of the fab decreases expenditures going forward or enables greater concentration on TSL's main business, there may well be upside to the average earnings projections of $3.39 for 2008.

I still like CSIQ, although it does not offer as good a value as TSL, in my opinion. Its PE is higher than TSL's, while the expected revenue growth of these two companies is almost identical ($300 million in 2007 to $750 million in 2008 for CSIQ). What differentiates these two companies is their gross margins, with TSL's gross margins substantially better than CSIQ's. Although this would normally be viewed as an advantage to TSL, if CSIQ improves its gross margins—which should not be that hard to do—there may be some decent upside to earnings estimates, thus lowering its forward PE closer to TSL's.

Also, CSIQ has been doing some work with UMG silicon, and according to the last conference call, there is about $100 million of upside revenue potential in 2008 if the UMG production pans out as expected. If the UMG silicon production works out, it is likely to push 2008 earnings over $2/share (note that the highest analyst estimate for CSIQ in 2008 is $2.02), putting CSIQ's PE also at 13 (like TSL's). I also like the fact that the market cap of CSIQ, at $700 million is a bit smaller than TSL's, and that only five analysts follow CSIQ versus eight for TSL, giving CSIQ the potential for greater recognition in the marketplace and therefore, greater stock appreciation.

But taking everything into account, I like TSL a little better at a PE of 13 than CSIQ at a PE of 15, but might consider picking up a decent CSIQ position if it became available around $24.

A few other solars merit some discussion. ReneSola (NYSE:SOL) is a new entrant that just recently IPO'ed, and offers a 2008 PE of 17. But the earnings ramp on SOL (from 86 cents in 2007 to $1.08 in 2008) is not as impressive as TSL's double or CSIQ's 6-fold increase, plus, we know less about this company because it has only been public a very short time. Thus, at a PE of 17, it does not get ahead of TSL or CSIQ, in my view.

SOLF, YGE and JASO trade at PEs of 20-30, and I do not believe offer more compelling values compared to TSL and CSIQ that would justify those higher PEs. If a reader knows of a significant advantage one of these three has over TSL or CSIQ, please comment about it.

In my view, although SPWR is clearly the technology leader in the group today (they are cranking out 22% efficiencies in the lab, and are selling 19.3% efficient panels commercially), its growth rate has definitely decelerated. Indeed, on the conference call last week, SPWR indicated that revenues in Q3 of this year will be flat with Q2. In my view, even given its technological prowess, SPWR is overpriced at a forward PE of 43. In addition, I believe other companies will substantially narrow the efficiency gap in the next year or two (note that STP's Pluto technology is reputed to be achieving production efficiencies of 18-19%).

Suntech Power (STP) is a closer call for me. It has actually done worse since my last article, having dropped from $51.70 on 1-25-08 to $49.18 yesterday. I panned STP in January because I thought it was overpriced at its PE of 31, but I picked up some shares last week at about $45, at a PE of about 28. The reasons for this are as follows: STP's price got nailed recently because of lower than expected revenues, lower margin due to higher silicon cost, some foreign-exchange losses and unexciting 1Q08 guidance that was below 4Q07 actual revenues.

These negatives have been baked into the current price, and to the extent that STP will report better than expected, it should have room to run up given that it is still trading at ½ of its 52-week high. The question is, what is the likelihood that STP will report better than expected?

I think there is a decent chance of that. So far, companies that have reported have reported higher ASP's than most analysts have modeled. Since every penny of additional ASP goes to the bottom line, the small 20-cent-per-watt ASP increase can have a significant impact on the bottom line. Second, recent market events seem to suggest that the poly situation (which really hurt STP's margins in 4Q07) has improved, which should aid STP's margins. Third, foreign-exchange losses of $6 million should be easily reduced by simple hedging. Next, STP is the biggest solar PV producer in the world, which ought to give it a decent leg up on its competition in this business where scale is a significant advantage.

Next, STP has a couple of products that may give it a competitive advantage—Andalay, which it has licensed from AKNS, and which reportedly is a slick and inexpensive way to install solar panels, and building-integrated PV [BIPV], where actual building structures (actual roof, walls, etc) make electricity. Therefore, I think STP is a buy in the low-to-mid 40s.

Finally, STP announced Tuesday that it has opened up offices in Australia, a market that I believe will generate significant solar PV demand for STP. STP's CEO attended university in Australia, and of course, Australia and China do a lot of business already.

I will close by discussing FSLR, whose prospects I didn't like in January, and which I like even less now. I say that despite acknowledging that with a 74% increase, FSLR outperformed all the other solars, by far, including CSIQ, in the last 3 months (but then again, the million-dollar houses in California also went up every month until they started crashing in 2007). Although the consensus earnings for FSLR in 2008 is $2.49, even at $3, FSLR is trading at a forward PE of about 100. For any company to justify that sort of PE, it has to be able to more than double its earnings on an annual basis for several years into the future, usually because it has a proprietary product or service for which there is no meaningful competition (think Microsoft (NASDAQ:MSFT) OS in the distant past, although even in those days, I don't think MSFT garnered forward PEs of 100)).

One stumble and all of a sudden, FSLR no longer merits a PE of 100, and contraction of the multiple kills the stock price. One need not look far to find examples of this—look at STP and SPWR, both of which are trading at about half of their 52-week highs (Apple (NASDAQ:AAPL) and Google (NASDAQ:GOOG) are two more examples of fabulous companies who got knocked down because they failed to meet continuing outrageous expectations—and they "only" trade at forward PEs of 30 or 40).

Stocks like FSLR remind me of a Ponzi scheme or of our housing bubble—the only reason you are willing to pay "X" amount for something is because you believe someone else will pay YOU more than "X." But there is always a peak and a crash, whether it's a ridiculously-priced house or a ridiculously-priced stock.

Frankly, I would not even pay a forward PE of 30 or 40 for FSLR, and here's the reason why not:

The KEY reason FSLR is worth more than the other solars today is because it is the price leader in terms of cost per watt. But there is only one reason for that price leadership—which is that polysilicon is priced in the hundreds of dollars per kilo, whereas within 2 to 3 years, it will be priced in the dozens of dollars (and not that many dozens at that). Since poly is the #1 manufacturing cost for FSLR's competitors, an 80% decrease in the cost of poly (from, say, $200 to $40)—as is likely to occur over the next several years—will eliminate FSLR's cost advantage. Keep in mind also that FSLR also has one cost disadvantage that its poly brethren do not have—FSLR has to build two panels in order to generate the same amount of electricity that a single poly-based panel will make. That will double FSLR's cost for aluminum, glass and other costs to manufacture a second panel.

When you take this into account, I would not be surprised if in three years, a single poly-based 300-watt panel costs no more to make than (2) 150-watt FSLR thin-film panels. And all things (including price) being equal, who wouldn't prefer ONE 300-watt panel versus two FSLR 150s? In other words, FSLR's panels will actually have to be at least 10-20% cheaper to induce purchasers to buy them because the amount of room they require and the extra installation costs will have to be compensated for.

Finally, how much upside is there for a company—any company—with a market cap of $23 billion? I simply don't see 10-bagger potential for FSLR, whereas CSIQ at a market cap of $700 million or TSL at a market cap of $1 billion could go up 10-fold in a few years and still have a market cap less than HALF that of FSLR today. Another way to see this is to look at the price targets for FSLR, which range up to maybe $350 or $400 (I don't follow this closely, so maybe I am wrong here). That gives an upside of maybe 35%. In contrast, if later in 2008, it looks like TSL will make $5/share in 2009 (consensus today is $4.49, but it was $4.10 just a week or two ago), and by the end of 2008, TSL's PE expands modestly to 16 against 2009 earnings, the resulting price of $80 represents a 2-bagger.

In conclusion, I believe that the solar space will grow vigorously over the next decade, and I believe that at the present time, TSL offers the best risk-reward ratio in this space.

To those not familiar with the solar space, I must close with a few cautions. First, solar stocks are extremely volatile, with 20% daily moves not unusual. So if you do not have the stomach for seeing your position lose 20% of its value in one day, solar may not be the place for you.

Second, the solars have gone up a lot in the past few weeks, and there is a good argument to be made that they need to slide back before they go back up.

Third, there is reputed to be a psychological connection between oil prices and the solar stocks. To me, this was more of a factor in the past than it is now, but it remains a factor. Thus, if oil slides back from $119, where it closed yesterday (as I expect it to), it may take the solars with it.

Fourth, some people believe that the stock market in general is overbought right now and due for a correction back to 12,000, and some think even less than that. That correction may well take the solars with it, although the solar often don't follow the broader market.

On the positive side, we are in solar earnings season, and I think the solars will report well, and several may well get upgraded after announcing earnings. In addition, there is a fair possibility that in the next month or two, Congress will pass the energy bill that will extend the old incentives and create some new ones. Although a fair bit of that is already baked in, I believe the passage of this bill may move the solars up another 5-10%.

Finally, over the next few months, I believe we'll hear other utilities announce their decision to implement solar in a fashion similar to that of Southern Calif Edison. This will spur the notion that we are close to achieving grid parity, and if that pans out, increasing demand will translate to significant earnings growth for the solars.

Whether the "bullish" forces I have discussed above beat the "bearish" ones, I cannot say. But if you are comfortable with the volatility and other risks discussed above, I think that TSL at $43 is a compelling buy, as is CSIQ at around $24.

Disclosure: I own a large position in TSL, and a decent-sized position in STP. My CSIQ shares were called away from me this past weekend since I sold $25 calls against my position about 3 weeks ago, but I may buy more CSIQ on a dip. I am not short any other stock, but also hold a large position in Penn West (NYSE:PWE), which will be the subject of a future article.