Description

Hi, all --

Hmm. There are a few things going on here, and some hair. Let me start with the simplest attention-grabbing bull case: CSIQ is trading at ~25% of readily ascertainable NAV. Note that I'll be footnoting the f*ck out of this claim as we dig in. Alternatively, one could claim on an SOTP basis that you're getting the world's 7th largest photovolataic manufacturer for free. Again, footnotes will be needed!

I'm struggling more than usual with how to organize this. Here's a table of contents, which I might not adhere to:

1: Brief Overview, Mostly Hair

2: Zero-IQ Speculative Case

3: "NAV" And Actually There Is Some Value Here

4: How I'm Thinking About This

5: Other Stuff

(It turns out I did more or less adhere, in a way that structures this as going from risks --> rewards, if you want to skip to the actual long thesis, it's at the end.)

Brief Overview, Mostly Hair

Canadian Solar operates through two main subsidiaries:

1: CSI Solar Co. Inc., a Chinese manucturor of (mostly) PV modules, and,

2: Recurrent Energy, a renewable energy developer, which has meant, "Go out there and build some solar farm and flip it," although they're now moving towards keeping the finished product on the books.

95% of the revenues come from the Chinese manufacturer. Do I need to say, "Chinese manufacturer" again? They're not just a Chinese manufacturer, they make PV modules. The Biden administration has effectively embargoed PV imports from China and is rapidly closing any loopholes like straw-man manufacturing in Southeast Asia. Meanwhile,  an enormous amount of oversupply has developed across the entire supply chain that could take years to absorb. Oh, and people think last night's debate helps Trump's election chances, which would be another kick in the nuts to anything green or Chinese. Every single solar stock is down ~50% TTM and 75% from recent highs. While all this going on, CSI Solar is taking on debt and burning cash to expand it's manufacturing facilities.

(A brief note on that last part -- It's not as crazy and sounds. CSI is big in modules 

(Another parenthetical -- The PV manufacturing process goes like this: Polysilicon, the raw material --> ingot, a big block of polysilicon --> wafer, thin slices of the ingot --> cell, wafers doped so they can "absorb" light --> module, a bunch of wafers jammed & wired together to deliver the electricity. Here's a graphic that should give you some idea:

\

)

sorry, big in modules but not so much in the rest of the chain:

, and think it a strategic necessity to vertically integrate upstream.)

In addition to all the above, one could question whether the industry is investable at all. Here are some 10 year #s for CSIQ to illustrate the point, note that we'll be referring to this table in the next section:

OK, so that's 10 years, they've never paid a dividend, share count has gone up, there are no visible signs of earnings growth, they're trading around 2015 lows, and this is typical for the industry. If there's zero FCF, these businesses are worth -- what, exactly? Why shouldn't we just roll with Falcon44's take from his 2020 JinkoSolar short pitch

( https://valueinvestorsclub.com/idea/JINKOSOLAR_HOLDING_CO/8680562402 ) 

that, "These long-term features of the solar module manufacturing industry–double-digit price deflation, ROIC < cost of capital, and persistent cash burn—are structural [....] It is a tough business where operators must run to stand still"?

As a final bit of hair, Qu, the founder, chairman and CEO, tried to take the company private for $18.49 in 2017 but got shot down by the board.

Zero-IQ Speculative Case

I'm going to duplicate that table and append current metrics:

The "10 Year" table price history shows that this is a volatile stock. This is extremely stupid, but, if it takes three years to get back to the third lowest 52 week high over the last 10 years, ($25.89 from 2019), that's actually pretty good, that's 20%/year! If it takes one year to get back to the prior all time low 52 week high, that's also 20%. Come to think of it, the stock was at $21 on June 3rd of this year. If it takes 5 years to get back to its 30-day high, that's 7.3%, which won't win any stock-picking contests, but beats a kick in the head.

As shallow as this analysis is, it hopefully dispels "dead money" fears; historically, when conditions in the solar industry have improved, the stock has gone up. And even if you think "no FCF means no value", empirically, there's a market for these things.

"NAV" And Actually There Is Some Value Here

NAV is derived from two sources:

1: CSIQ's 62% stake in CSI Solar Co. Ltd, a solar & energy storage manufacturer, which is listed on the Shanghai STAR market, 688472 to be precise. At last quote, CSI Solar has a market cap of >$5B, so this is worth roughly $3B. (Footnotes will be coming!)

2: 100% ownership of Recurrent Energy, a solar developer. A recent investment from BlackRock allows us to conclude that BlackRock, a sophisticated etc., values this at ~$1B on a pre-money basis.

$3B + $1B = $4B, compared to a current market cap of $975MM. So, CSIQ is trading at less than 25% of NAV. 

Here come the footnotes!

1: Using CSI Solar's market cap on the Shanghai market for NAV calculations ... I'm not sure if I want to call that a joke or not. These discounts are a persistent feature of the investing landscape. JinkoSolar (NYSE:JKS) also has its manufacturing facilities listed in China. Too lazy to look up the details right now, but JKS is trading at an enormous discount, something like $5B (allegedly) vs. US market cap of $1B. These apparently unarb-able discounts can be arbitrarily large and persist forever.

I also don't want to do the fundemental work to claim that CSI Solar's Shanghai market cap is justified to begin with. I think it might well be, but I don't want to make it part of the pitch. But the Shanghai mark is relevant for a couple of reasons:

A: If you're religious about CAPM, you should use the Shanghai quote when estimating cost of capital.

B: So ... this is a cyclical. One of the problems with cyclicals is that they can go belly up before the cycle turns. If that happens here, it's self-inflicted, CSI Solar is trading far from distressed levels, I think they could do a 5% follow-on offering over book.

C: It opens up the possibility for management to at some point do something about it. Unlike some, CSIQ's leadership is not completely oblivious to the capital markets:

-- Qu tried to steal the company.

-- They rang the register for $150MM around the top 2021.

-- They IPOd CSI Solar last year.

-- The whole solar developer cycle is nothing but capital, capital, capital.

Going from, say, 62% to 57% would net $250MM, >25% of the CSIQ's market cap. Could that happen? Sure.

I was going to stick this in the "other stuff" section, but, while we're here -- Apropos of CSI Solar valuation, they have a small but very rapidly growing Battery Electric Storage System segment. This is going to be like -- 30% of revenue in the back half of 2024? More than that in 2025? BESS and solar go together like ... maybe not ham and eggs, or peanut butter and jelly, let's say hamburgers and pickles, which I could do without but a lot of people like. The cross selling opporunities are obvious, there's a real reason for this business to exist. Depending on what #s and timeframes you want to pick, this is 20-38% of the size of Tesla's Megapack business, which Adam Jonas thinks is worth >$100B right now, ha ha ha. I think it's too early to put #s on anything in the space, but I could see people (Chinese people, I mean, not American people) arguing that the BESS business alone was worth $2-$3B.

2: For the development business, "readily estimable" might be more accurate than "readily ascertainable". BlackRock put $500MM in as a convertible preferred, converting into 20% of the equity should things work out. Some idiots look at that and say, "Recurrent Energy is worth $2B!" It would be, if you ignore the 9% coupon and seniority.

Fortunately for us, the preferred is PIK-able, so we can get a good idea of BK thinks the equity is worth. The main assumptions are: (1) the scanty disclosure -- 5 years, 9%, convertible into 20% -- isn't some kind of ruse to dupe investors into overvaluing the developer, this is regular deal, there's nothing too funky going on, and (2) Recurrent Energy can force conversion above the strike, and the strike is, as is generally the case, something you'd expect to hit in the ordinary course of business over 5 years if you're retaining earnings, like today + 25%, 30%, even 40%, something like that.

So BK needs to be comfortable with RE PIKing and then getting converted. Depending on how you handle the details of "in kind" and "conversion", they'll wind up with 28%-31% of the equity in 5 years. Because of the seniority, let's round that up to 30%-33%. So, if they're comfortable putting in $500MM now, knowing that their only return might be owning 33% of the venture, the equity is worth $1-1.1B on a pre-money basis. Let me know if I need to step through this, it's pretty boring.

A couple of notes about the RE valuation: The did $100MM operating earnings in 2023. CSIQ bought Recurrent Capital in 2015 for $265MM, folding it into its existing solar development business, which had a similar backlog and thus was worth presumably around the same, so $1B 9 years later doesn't sound out of line. RE wants to own instead of flip when appropriate. They aim to have 4GW of solar and 2GWh of storage by 2026. Given the lead times, this is going to happen, for instance, they're going to be picking up 1.2GWh of storage in Arizona in Q3 2025.  A comp that might ring some bells with VICers is AY, due to it's association with AQN. Lack of details make me reluctant to try adding more precision, but you can see how a portfolio of green infrastructure might be valuable down the line.

A couple of notes about the transaction: This is not cheap capital. BlackRocks's Global Renewable Power Fund IV is looking for 12-13% gross IRRs. This is a 10% position for them, so, pretty high conviction. If the preferred is shooting for 12-13%, you'd think the equity is hoping for a bit more. 

There are a couple of other bits and pieces kicking around -- I think there's excess cash at corporate from the $150MM raise, CSIQ owns 15% of a Japanese infrastructure fund so that's $50MM, etc. The sums might not be large, but neither is the market cap.

How I'm Thinking About This

There's some hair, but 62% ownership of a leading solar manufacturer just isn't a liability, which is what's implied by the current price. What put me over the top on the stock was the reflection that if the company blew out CSI Solar at 20% of book, which they could do overnight, and announced a $100MM buyback, the stock would go up. A lot. Obviously they've not going to do that because it would be crazy.

Here's what it takes to get a quadruple out of this:

1: Blackrock meeting its targets on their Recurrent Energy investment leaves RE worth $2B.

2: You get $2B from the manufacturer if it trades back to 80% of its 52 week high and the US markets value it a massive but not completely insane 60% discount.

Over a 5 year horizon both sound plausible, and maybe something good happens sooner than that.

Business, market, political and geopolitical uncertainty make me reluctant to supersize the position. Maybe I'm just a coward. I don't know if I'd increase exposure on further declines, but the Recurrent Energy backstop would make me comfortable averaging down.

Other Stuff

1: If you're hoping for a nuanced, knowledgable and detailed discussion of possible timings for a turnaround in solar module pricing, boy, are you in the wrong place. CSIQ and peer JKS think, or have said, anyway, that they see better days in H2 2024. Maybe they're right. I've heard other people talk about 2025. The EPS estimates are consensus and assume a better H2.

2: The entire space is completely bombed out. Other fun names to look at are JKS and DQ which are unbelievably cheap. I prefer CSIQ because I derive perhaps misplaced comfort from the non-manufacturing, non-Chinese assets.

3: The balance sheet is misleading. CSI Solar actually had (from memory) $700MM net cash at year end, although I expect that to have come down. The developer obviously deals in a lot of project finance; like, they just signed for a EUR700MM line of credit, expendable to EUR1.4B, and that's just one of many, many facilities. Recurrent Energy could wind up with infinite amounts of non-recourse project funding debt.

4: The company just built a 5GW module plant in Texas and is standing up a 5GW cell factory in Indiana for arrival towards the end of 2025. This is a total  investment of about $1B in US based manufacturing. Look at the market cap again.

5: Common ownership of the developer and manufacturer makes a lot of sense from a business point of view, but (obviously!) not necessarily from a capital markets perspective. I could see a separation down the line.

6: You'll see comments like, "There's enough solar capacity for the next 10 years!" But a good chunk of that capacity is now functionally obsolete ("operators must run to stand still"). 

7: One could argue that this is yet another "catalystless deep value" stock, but it doesn't feel that way to me.

8: Trump. I don't worry too much about this. He might not win, and ... maybe the rhetoric will be more bellicose, but how can he take a tougher line that Biden on the trade issues? Decreased subsidies for solar wouldn't help, but the market isn't going to go to zero, and the US is just one market, albeit a large one.

9: Other stuff that I'm forgetting about, that's what the comment section is for.

To sum up: The stock is really cheap and you need close to war with China and a complete disaster for a global developer for it to be a zero.

Yours,

Bowd

Catalyst

The developer performs according to expectations; a turnaround in solar module pricing and/or excitement about energy storage for the manufacturer.