SolarPVInvestor Forums

I'm modelling TSL as well as other solar 11s. If prices have normalized in 13Q4 to a level where TSL can get 85 cents module ASP, then it won't be a problem for them to post a profit. In such a scenario they would probably also be up from the 65% utilization rate in 12Q4 to 100% and profits would be ample. It doesn't take much for these companies to get bloated bottom lines again. Some have lost a lot and need to repair their balance sheets a lot to get to a healthy debt ratio though, so in those cases it will not enable expansions (thank god), unless the banks suffer from amnesia.

In the more conservative scenario you describe where poly gets a 5 cents price bump and non-Si processing only gets another 5 cents price bump for a total 10 cents price bump from 12Q4 65 cents to 13Q4 75 cents, then that would signal a less radical shift in supply/demand balance, but proabably enough for TSL to get from 65% to 100% utilization rate and that 100% would be required at 75 cents for them to post a profit. In that case the profit would be small, in order of 10 cents EPS for the quarter compared to in the order of $1 for the quarter in the former scenario.

Regarding sensible GM level for an industry, I don't think it can be discussed in isolation. It needs to be related to the opex and interest cost as percentage of revenue. As ASP have gone down 65% past 2 years opex and interest as percentage of revenue has increased a lot. Adding to this that freight has moved from cogs to opex the GM level able to sustain opex, interest and reasonable roi is not the same as it was when 30% GM rendered high NM around 20%. The NM you get from 30% GM today is more like 10% and to get to the 20% NM of 10Q4 a 40% GM would be required today. So the 20% GM today is a breakeven GM, while it was a 10% NM in 10Q4. I guess the rule is simple; add 10% to the GM you thought was reasonable two years ago to get what is reasonable today based on the profitablity level assumed reasonable. Or even simpler select a NM you think is reasonable and add 20% for the GM. In boom times I think NM for integrated wafer to module can reach close to 20%, for integrated poly to module it can reach double that (for imaginary company best in class in every vertical).

My set for Q4, 2013 per watt
0.106 poly @20 at 5.3g
0.45 processing (full utilization on 2400MW)
0.12 opex
0.01 interest
total: 0.68 per watt, modules selling at 0.70,
Poly will settle in vicinity of 20/kg
My opex figures are lower by 2 cents as I think the .14 does not really show much of an improvement from spent currently, only extended on larger volume, say (600MW) is 0.13. So GM at 21% in area of 0.10 to .17 per share net income having tax effects etc.

Quoted from "odyd12"

Poly will settle in vicinity of 20/kg

Based on continued oversupply (implying weak demand growth and no poly tariffs)?

At full utilization they should be able to get opex below 12 cents, possible below 10 cents with their guided opex reductions, but interest would be at least 2 cents. The 45 cents non-Si (including 50% wafer outsourcing) is more aggressive than I've heard Trina guide. Others have guided cost improvements, but my question is how well Trina will keep up.

I hope by 13Q4 things have recovered a bit more than 20 $/kg and 0.70 $/W.

The interest is at 0.02 right now and that is on 60% utilization. I think that some of the loans will be refinanced as well. Loans will be also paid down from inventory reductions. I do see a very tight management in that area, to eliminate exposure to fluctuation and adjustments to lower cost or the market. I am looking forward to inventory levels to improve in Q4, based on reports from IHS, Solarbuzz and zoom
I can easily see 0.45 a year from now. Under 0.50 is already in place for Hanwha and Yingli if we assume it was reached as guided.
Module ASP will go up based on the efficiency factored in and upheld demand. However hanging capacity will hold the lid on it. ASPs variable based on geography will benefit exporters, so unless Trina will lose its global reach they should extend that benefit. Australia and Japan and the US can give them a lot of ASP advantage, Europe not as much and I see them moving away exception Uk. The biggest issue here is lack pronounced consolidation in China, which will hold the could of the cap over the industry.

Quoted from "Klothilde"

Also I see that we have completely different views on where poly prices are headed, which is quite interesting and for sure leaves room for discussion.

I have conservative view on poly increases.
I believe that prices will not go up much more than $20 per kg on average. The industry is not prepared to absorb this increase as higher module pricing is not going to be welcomed to impede demand. There is still hanging capacity aspect remaining dormant now, but with each cent up in ASP threatening to open another production line.

At $20 per kg, most companies in supply chain can produce at cash costs. I also believe there will be exemptions made for long-term contracts with Wacker and OCI. Therefore the move on pricing right now is reactionary to Chinese tariff and while attacking spot, will have limited effect. If the price is kept low (.65-0.75 module) dormant capacity is more so to remain dormant. Therefore there is no need to produce more poly, and imports will increase to tier-1s only but on contractual basis.


Policies in China, imho, do not have intention to revive poly industry on ASP pricing. I think that plants upgrades lowering costs will require money and this how regime can quickly separate its support.

Quoted from "odyd12"

The interest is at 0.02 right now and that is on 60% utilization.

I think it was 3.5 cent on 60-65% utilization in 12Q3. So at 100% it would be 2 cents with rougly same net debt amount and average interest rate. Rates have gone up and debt level are reduced with cash (not affecting net debt) and by reducing inventory (can reduce net debt). Other assets like AR reduction from tighter credit terms and PPE reduction from not net investing (spending less capex than asset depreciation) are possible to reduce net debt. Over all with increasing interest rates and technology upgrade requirements it will be hard to reduce total interest expense a lot before they start to accumulate significant profits again.

Quoted

I think
that some of the loans will be refinanced as well. Loans will be also
paid down from inventory reductions. I do see a very tight management in
that area, to eliminate exposure to fluctuation and adjustments to
lower cost or the market. I am looking forward to inventory levels to
improve in Q4, based on reports from IHS, Solarbuzz and zoom

Agree.

Quoted

I can
easily see 0.45 a year from now. Under 0.50 is already in place for
Hanwha and Yingli if we assume it was reached as guided.

That assumes Trina has made similar progress. For now I see Trina and JKS saturate in progress, while YGE, SOL and HSOL have the momentum.

Quoted

Module ASP
will go up based on the efficiency factored in and upheld demand.
However hanging capacity will hold the lid on it.

Before uncompetitive capacity went dormant or away a lot of panels filled the channel inventories and where dumped on spot markets as small shops were liquidated. Now most of that inventory has finally depleted, thanks to China taking the new production and rest of the world eating from stocks without replenishment. The big question is how much of that uncompetitive part of 60 GW panel capacity have stayed around as dormant and at what ASP they dare light up the shop again? Maybe the brake will not hit as soon as we belive and maybe demand accelaration is at a level where it runs through that cushion? These things are hard to predict. Looking at 2013 geographic landscape, we see a completely new map, China, US and Japan covers 50% of world market, while Germany and Italy together are below 20% and rest of world is above 30%. This will impact global ASP as well as create opportunities for companies to have a geographical mix with ASP deviating quite a lot from the global average.

Quoted from "explo"

I think it was 3.5 cent on 60-65% utilization in 12Q3. So at 100% it would be 2 cents with rougly same net debt amount and average interest rate.

The increase of debt in Q2 by 164M caused increase in payments by around $4M in Q3. The debt was paid down by 94M, so I suspect that this interest dropped. I calculated average rates at 4.6% IR for Q3 and 2.9% in Q2. However I think that rates will return to around 3% in 2013 for the 6 big and 6 small companies, suspect that we will see around $200M reduction bringing overall debt below $1B by Q2. This will be possible by reduction of inventory by as much as $160M (leveled off at $200M) and further cash on hand reduction. I suspect that IP for Q3 be in area of $8M on 600MW or 0.013

Quoted from "explo"

That assumes Trina has made similar progress. For now I see Trina and JKS saturate in progress, while YGE, SOL and HSOL have the momentum.

Trina is a third largest spender on R&D, they will deliver those costs in my opinion

Quoted from "explo"

The big question is how much of that uncompetitive part of 60 GW panel capacity have stayed around as dormant and at what ASP they dare light up the shop again?

That is the question. I think that pricing mechanism which killed Europe are going to be used to kill domestic operators. China is not going to support 400 entities but at least 100 are holding on steady, with finger on the trigger. They are the ones, which will start putting production lines on the moment of notice, pricing above $0.75 cents will probably bring 30% (20GW) of it back and that would be destructive. This is not the question of if. All my Chinese sources describe this to be the behavior.

Quoted from "Klothilde"

What are your current thoughts regarding where the poly price in China is headed this year? Without tariffs I was looking at a range of 25-30 $/kg by year end, but now I think that tariffs can add about 10-15 to this estimate, i.e. 35-45 $/kg by year end. If poly goes up to 45 then an ASP of 85 cents will hardly be enough for TSL to post a profit imo.

I'm trying to stay conservative on poly pricing outlook, since poly capacity is part of my investment case. I model prices that makes sense from a business perspective.

Looking at the international incumbents they typically have the following breakdown:

Cash $12
Depreciation $7-10
Interest $5-7
OPEX $3

With a 12% net margin requirement, which I think is really low for a billion level capital and decade level time-frame investment, considering the risk of disruptive technology destroying your investment, you can add:

Tax $1
Net profit $4

for a $32-37 or mid-point $35 ASP. I cannot see these incumbents wanting their avarage contract price to be below this level. That would just be like saying "we've setup a shop projected to lose money the coming decade, want a piece of that?" to investors.

For GCL and SOL in China it is a bit different, since they are getting capex down to at least half of the internationals without sacrifizing much cash cost. This means their breakdown is more like:

Cash $13
Depreciation $3-5
Interest $2-3.5
OPEX $3
Tax $0.5
Net profit $2.5-3

ASP $25-28

Since I believe the China cost scenario might drive pricing long-term, considering China consumes more than 70% of the stuff, and since I want to be conservative as I'm investing in the production side rather than the consumption side, I currently use $26 as the normalized price level in my models. Short-term in a demand bump or tariff scenario the cost levels of current supply (including tariff cost) will dominate though. $40-45 is not a negligible risk/chance in that scenario and among c-Si the profit will be claimed by those with poly capacity in China, since 10 cents gross profits per watt are moved from purchaser to producer of poly in China (For TSL poly at $45 instead of $25 would be similar to ASP at 75 instead of 85, i.e. quarterly profit at dime level instead of dollar level). TF would also benfit from that scenario.

Note that the 12% net margin that I said I think is a low target for poly business relates to the capex level. Add $2 to the $25-$28 and Chinese tier-1 poly capacity will be happy. For the internationals to be happy you need to add maybe $8 to get a $40-45 range. To me the difference in happy ranges from $40-45 for the traditional incumbents to $27-30 for the Chinese challengers is a major investment case.

All this said the poly expansion plans in 2010 were out of whack and if they had been implemented, then the time until price normalization could have been very long. Luckily much of it was suspended. This general supply situation and tariff impacted China supply situtation will impact China prices more short-term. Again the expansion cyclic and politics disruptive supply/demand balance component is much harder to model and predict than the cost based trend component.

On the dormant 60 GW panel capacity, see my reply to odyd.

Quoted from "odyd12"

The increase of debt in Q2 by 164M caused increase in payments by around $4M in Q3. The debt was paid down by 94M, so I suspect that this interest dropped. I calculated average rates at 4.6% IR for Q3 and 2.9% in Q2. However I think that rates will return to around 3% in 2013 for the 6 big and 6 small companies, suspect that we will see around $200M reduction bringing overall debt below $1B by Q2. This will be possible by reduction of inventory by as much as $160M (leveled off at $200M) and further cash on hand reduction. I suspect that IP for Q3 be in area of $8M on 600MW or 0.013

Thanks for clarifying. It adds up, but 3% seems low. I don't think Chinese banks are in a position to offer those rates anymore..?

Quoted

Trina is a third largest spender on R&D, they will deliver those costs in my opinion

ok, you're probably right, but I'd like to see them guide that there's room to cut cost further, like YGE and HSOL did.

There in lies the volatility of solar stocks and drastic swings in peoples expectations. We are at the point where a penny or 2 swings from a loss to a profit in the near future.

Trina had $15.6M in interest payments in Q3 or $0.04/watt shipped(380MW) After adding in the derivatives and interest incomes they had $0.026 in interest per watt. At full capacity(600MW). this becomes $0.026 and $0.016. If derivative gains stayed constant this would be great, but they swing positive and negative depending on hedging and stock movement. They did lower total debt owed in Q3 over Q2.

Current processing targets are $0.46 for Q4 2012 . 1 company has identified sub $0.40 as target for end of Q4 2013. Others will follow as even SOL at $0.46 Q4 target 2012 still outsources cells. Those making internal cells will have costs lower.

A simple historical 1% increase from 17-18% efficiency gain is a 5.8% saving and cuts costs from $0.57 Q4 target to $0.54 and processing drops to $0.44. Historically efficiency gains and lower grams per watt from gains are around 5-6% annually. This is without any material cost savings or manufacturing efficiency gains. Once can easily see the $0.40 processing targets if looking closer at the numbers.

Personally, my number show at 600MW Trina has an opex at around $0.11-$0.12 and at $0.10+ on 700MW. This will still be above the best of bread in Opex due to Trina footprint in foreign countries and ratios of shipping oversees vs Asia Pac. That same footprint may be why Trina commands a higher ASP as well from those others.

I just to be clear, between Explo and Klothilde, you see poly selling in 2013 at $32 to $45 per kg.
This view would not have anything to do with Renesola having own poly production, so they would get massive benefits if price got to this level?

Quoted from "Klothilde"

Thanks for your thoughts on poly, Explo. Your cash costs seem a little low to me for Siemens, so I'll go and revisit. Thought GCL was hovering around $15/kg cash.

The 13 $/kg was based on SOL's target, which I'm quite confident with, since they've confirmed a $5 depreciation component in their $18 target. GCL depreciation distribution among their 65 kT poly and 8 GW wafer is less clear, but I think the poly depreciation might be close to $4 and they've guided below $18 cost, so that would make their cash cost around $14. But as you saw in my example the cash cost effect on price is straight forward, while the $1 added depreciation cost implies added interest and roi coverage requirement too. So GCL having more cash componenent in its $18 would just lower their happy price range.

Quoted from "Klothilde"

Regarding TSL vs. YGE the funny fact is that as per Q3 TSL was one cent lower than YGE in production cost, namely 52 cent vs. 53, if you take out the underutilization penalty. TSL is aiming at < 50 cents at YE12 and at 45 cents at YE13. Historically TSL and YGE have been surprisinlgy, or better fishily synchronized when it comes to cost per watt, mostly running at parity or at one cent difference. So I don't expect a big difference in production cost for Q4 or year end either. I speculate that YGE's 45 cents is just a stretch goal that got out by mistake.

I checked the Q2 CC for TSL now and that's correct, they have guided with uncertainty for 45 cents non-Si in 2013. I've updated my model. In that case 75 cents ASP should be enough for profitability even without full utilization if poly stays in a $25-30 range. Note that the 150 MW of projects in 2013 that they've guided for is what takes them above breakeven.

Quoted from "Klothilde"

As to the dormant component capacity: We are not talking about dairy products that go bad here. These are machines that are ready to fire up at the push of a button and crank out product at cash cost. Given the low degree of automation / high degree of manual labor in China it's pretty easy to reconfigure and ramp up a cell or module line that has been idle for some time.

We have to remember that these modules kept selling from both factory and retail inventory long after plants where shut down. What cost level do these small tier 3 and 4 plants have? They need a margin of profitability or at least positive cash flow to bother with recruiting labor, sales and admin staff and get organized and put in cash for working capital to acquire raw material and build finished goods inventory and give credit on sales. These hassles delays shut downs and will delay/prevent start ups. Entrepreneurs might have gotten tired of spending their time, energy and money on this sector. Haven't usually 90% of the guys from the religious boom thrown in the towel permanently when the secular boom happens? Looking at other industries like automotive as an example..

Quoted from "odyd12"

I just to be clear, between Explo and Klothilde, you see poly selling in 2013 at $32 to $45 per kg.
This view would not have anything to do with Renesola having own poly production, so they would get massive benefits if price got to this level?

Odyd, I use 26 $/kg in my models. This reflects a low end of my estimate of what can be expected from cost of production in China. $45 is the high end of what can be expected from cost of production outside China.

As I wrote, it is because I have invested in a company that has poly production that I want to be conservative with the setting the price level expectation of polysilicon too high. If I had been invested in buyer of polysilicon then using $45 in my models would instead be the conservative approach.

For 2013 we have supply/demand playing tricks on us and the cost motivated price might not matter as much. Prediction for poly price considering tariffs for 2013 could be $20-25 range if oversupply pressure continues, but if demand normalizes poly price in 2014 and tariffs remain then we can see much higher prices than that in 2014.

In all the cost structures that people are posting, nowhere do I see the cost of the 36% tariffs that companies that produce cells in china, TSL, YGE, CSIQ etc... have for sales to the U.S. (and soon possibly in Europe), So the 52c all in costs becomes 71c for those U.S. sales (and soon possibly Europe).

And even if these companies decide to take the cell production elsewhere, you are still talking about increased costs versus production in china.

Quoted from "explo"

For 2013 we have supply/demand playing tricks on us and the cost motivated price might not matter as much. Prediction for poly price considering tariffs for 2013 could be $20-25 range if oversupply pressure continues, but if demand normalizes poly price in 2014 and tariffs remain then we can see much higher prices than that in 2014.

ASP for polysilicon is also a deterrent for high production firms creating overcapacity, when price would go up. No different than any other value chain component would. I think that prices will stabilize in area of the 20-22 for this year. OCI is not even fully utilized today, having higher prices than spot. That (20-22) price level will ensure nobody starts poly production. I do not see module go beyond 0.72 to 0.75 range. There reason is simple. Utilization levels are 70% for most module makers, but for the average year were at 60%. There is 25% of shipments form China still actively delivering with "no name" companies. Tier-1 ones will hold the price down, and improve their production costs further, forcing them to shut down.

Trina looks to be stating similar future costs the same as most companies. The issue with Trina from the comments in Q3 is primarily utilization rates and existing inventory levels. Once inventory is cleared and utilization rates increase, their costs should be like all other peers only delayed by a quarter . This parity should be reached around Q2.

Here are some excerpts from seeking alpha transcripts from Q3. They all are stating similar numbers if you factory factory utilization. JKS near 100% has a $0.01 lower cost where Trina has a $0.05-$0.06 lower at 100%.

YGE:
As a result of all the aforementioned efforts, we expect to bring our non-silicon costs down to below $0.50 per watt and poly silicon costs close to industry average level by the end of this year.

we are target to reduce to roughly $0.45 by the end of this year. And for next year we see there is another list of $0.05 to $0.08 for the cost reduction for the non-poly silicon processing

our expectation is to reduce the costs and down below $0.50 for the non-poly silicon part for Q4. That is the average cost. So the supplementary comment I made is when we exit this year and I am expecting the non-poly silicon costs will be somewhere close to $0.45. But that is an exit point.

JKS:
In Q3, the nonsilicon costs, $0.47, is also including the utilization. So for Q3 it is around 75%-80%, roughly. So basically, the cost is already including that.

In the total cost, $0.47, $0.05 is depreciation cost right now

Trina:
Due to lower capacity utilization, our non-silicon manufacturing costs increased from $0.52 to $0.54 in the third quarter, primarily due to the higher unit depreciation in energy costs resulting from reduced output.

If we come back and so we can ship more and fully utilize, we can come back to the depreciation by approximately $0.05.

SOL:
on the module side, $0.52 to $0.53. We think there's additional room, so we think it can go down to that low.

CSIQ
We remained on track to achieve our goal of all-in module manufacturing costs in a range of $0.55 to $0.60 per watt by the end of 2012.

the cell conversion cost is about $0.13 to $0.16 and the cell to module conversion cost is the below $0.20.

HSOL:
We maintain our earlier projections at non-poly processing cost should fall to $0.50 per watt or below by year-end

The $0.50 is including $0.06 so as for devoting $0.06 lower utilization rate the loss, and so our processing cost will be around $0.46 or $0.47.

CSUN:
I want to reaffirm our conversion cost reduction road map which is targeted to achieve cell and module conversion cost of $0.15 and $0.21 per watt at the end of this year

we will achieve this through better supply chain management, better production efficiency and

Quoted from "larryvand"

In all the cost structures that people are posting, nowhere do I see the cost of the 36% tariffs that companies that produce cells in china, TSL, YGE, CSIQ etc... have for sales to the U.S. (and soon possibly in Europe), So the 52c all in costs becomes 71c for those U.S. sales (and soon possibly Europe).

The tariff is in the cell not all components. All of them buy cells elsewhere to ship to the US. That cost is increases their COGS, but it must be less than tariff, otherwise just pay the tariff. The offset is in the US ASP.