In this article we will go through the cashflow of Renewable Energy Corporation (REC) that is listed on the Oslo stock exchange. We will look at what REC sells, what I am basing my prognosis on, factors in the future, historical growth numbers, and several graphs comparing REC with Peers and showing costs of production.
But first here is the overall estimates I have for REC going forward. This is a green field analysis, meaning if all things go REC's way and the market swings upwards.
Overview of rec cashflow during quarter 1 2013 until quarter 2 2014:
REC ASP per quarter for modules:
q12013: 0,73 $ per watt
q22013: 0,77 $ per watt
q32013: 0,77 $ per watt
q42013: 0,77 $ per watt
q12014: 0,80 $ per watt
q22014: 0,80 $ per watt
Rec EBITDA cost per quarter (module cash cost + capex per watt + sg&a & r&d per watt):
q12013: 0,71 $ per watt (0,61 + 0,02 + 0,08) @ profit 2 cents per watt
q22013: 0,69 $ per watt (0,58 + 0,03 + 0,08) @ profit 8 cents per watt
q32013: 0,66 $ per watt (0,55 + 0,03 + 0,08) @ profit 11 cents per watt
q42013: 0,60 $ per watt (0,52 + 0,00 + 0,08) @ profit 16 cents per watt
q12014: 0,59 $ per watt (0,50 + 0,01 + 0,08) @ profit 21 cents per watt
q22014: 0,58 $ per watt (0,47 + 0,03 + 0,08) @ profit 22 cents per watt
REC MW shipments per quarter and EBITDA cashflow
q12013: 190 MW shipped @ 4,24 million $ profit
q22013: 213 MW shipped @ 16,72 million $ profit
q32013: 213 MW shipped @ 22,90 million $ profit
q42013: 213 MW shipped @ 34,34 million $ profit
q12014: 235 MW shipped @ 49,08 million $ profit
q22014: 235 MW shipped @ 51,48 million $ profit
Total 178,75 million $ ~ 1024 million NOK
REC ASP per quarter for FBR polysilicon:
q12013: 17,29 $ per kg
q22013: 23,75 $ per kg
q32013: 25,00 $ per kg
q42013: 25,00 $ per kg
q12014: 25,00 $ per kg
q22014: 25,00 $ per kg
REC FBR EBITDA cost per quarter (polysilicon cash cost + capex per kg+ sg&a & r&d per kg)
q12013: 17,14 $ per kg (12,0 + 2,14 + 3) @ profit 0,15 $ per kg
q22013: 18,95 $ per watt (11,5 + 4,45 + 3) @ profit 4,80 $ per kg
q32013: 18,45 $ per watt (11,5 + 4,45 + 2,5) @ profit 6,55 $ per kg
q42013: 14,71 $ per watt (11,5 + 0,71 + 2,5) @ profit 10,29 $ per kg
q12014: 16,14 $ per watt (11,5 + 2,14 + 2,5) @ profit 8,86 $ per kg
q22014: 18,45 $ per watt (11,5 + 4,45 + 2,5) @ profit 6,55 $ per kg
REC FBR MT shipments per quarter and EBITDA cashflow
q12013: 4125 MT shipped @ 0,63 million $ profit
q22013: 4125 MT shipped @ 19,78 million $ profit
q32013: 4125 MT shipped @ 27,00 million $ profit
q42013: 4125 MT shipped @ 42,44 million $ profit
q12014: 4125 MT shipped @ 36,56 million $ profit
q22014: 4125 MT shipped @ 27,00 million $ profit
Total 153,40 million $ ~ 879 million NOK
Float zone, Electronic grade, Siemens sg, off spec and silane gas sales:
Total cash flow profit from these volumes has varied between 80-150 million NOK in the last quarters. In q1 rec received 27 million $ (155 mil NOK) compensation for a silane gas contract. I expect the cashflow from these products to be around 55,72 million NOK (10 million $) giving a total of 27 + 60 = 87 million $ during the next 6 quarters. ( 498 million NOK positive cashflow.) In additional there could be a potential cash flow of 16 million $ from a tax settlement (not added in the calculations here) source: www.columbiabasinherald.com/politics/art...?mode=jqm
Total all cashflow: 419,15 million $ ~ 2401 million NOK.
What does renewable energy corporation (REC) sell?
Let us first look at REC SILICON. They annually produce 20000 Metric Tonn (1 MT = 1000KG.) of polysilicon. This happens at the factory in Butte in Montana and Moses Lake in Washington. Of the total Fluidized Bed Reactor (NYSE:FBR) silicon is produced at around 16500 MT annually. The rest is float zone (FZ), electronic grade (EG) and Siemens based solar grade (Siemens SG) and the production is around 3500 MT annually. The Siemens process is traditional in the industry, the difference is that REC uses silane gas and this causes more pure polysilicon.
What are these products?
FBR: Silicon of a purity where any contaminating substance is so small that silicon is at 99.9999% purity (6N purity) and up to 99.999999% (8N) purity. This is often called solar grade (SG) silicon. Another advantage to the FBR is that it is made of small bullets (2-3 mm size) and this gives advantage during production of ingots for making modules. The traditional Siemens chunk method most polysilicon is produced is in chunks crushed into smaller pieces ( 20 - 30 CM ) and so the smaller FBR bullets or granules can flow more easily and fill spaces in a melting oven much better than chunks of bigger sizes. Also the flowing form of the granular gives advantage when putting the polysilicon into an ingot melting crucible, and it also gives advantages during the melting processes where you can recharge the crucible and also top it during melting.
Why is this important? Because every ingot oven has components that are used in the melting process. Rec has a patent on a crucible coating that makes the crucible last 3 meltings but for the most the industry has to replace a crucible for each melting. In photon international number 6 2012 page 160 the cost for ingot production is detailed. The metal used for melting (the crucible) is 14,8% of the cost for a multi crystal module (c-si) and 20.8% for a mono crystal module (m-si.) Also electricity is a major cost component during heating, 15,5% for m-si and 36,4% for m-si. Since the granules allow less use of electricity and crucibles they are an advantageous product.
Floatzone: Extremely pure silicon. The process is base don making a silicon rod that is not in contact with anything else than silicon, Thus it is almost without any contamination of the silicon at all. For those who want to read more:www.pveducation.org/pvcdrom/manufacturin...-silicon
Since the purity is very high it can be used as a component in measuring devices and also used as a component in hybrid cars. More details can be found in this presentation:www.recgroup.com/PageFiles/2060/March_16....pdf
Electronic grade: Purity of 9N to 11N between 99.9999999% to 99.999999999% purity of silicon. One of the customers rec have for this polysilicon is Intel.
Off spec: Polysilicon at a lower purity than wanted. Rec usually sells this at a lower ASP.
Silicon Powder: Because of the production method of FBR there are also sometimes formed even smaller granules that are dust, this dust is sold at low ASP. One such customer is Eging in china. (www.egingpv.com/ )
Silane gas: Produced by converting 98% pure silicon (Metallurgical grade silicon - MGS) by a series of distillation processes. Among other things it is used in thin film modules and flat screens.
The Butte factory in Montana produce the EG and FZ. At Moses lake the FBR is produced.
Goods used by rec silicon: Mostly MGS and other gas/liquids used in the distillation process of MGS to silane. MGS costs REC around 37.5% of total costs of FBR granules. This is around 4.7$ per kg in costs. One major contribution to this is the tariff US have on Chinese MGS at 139.49% (news.xinhuanet.com/english/world/2012-03... ) But even with the tariff REC has a very expensive MGS contract, the prices has been around 3,1$/kg now in US and around 1.7$/kg in china recently.
For more details about REC silicon costs and also costs if the produced in china the following presentation is recommended:
How about REC solar and the production of modules? How big is the capacity now and what kind of product do they sell?
Rec solar has about 850MW of both wafer, cell and module capacity. They do some contract manufacturing, one of these sources is Jiawei (www.solarchina.com.hk/ , source of contract manufacture: www.recgroup.com/PageFiles/2422/T%C3%9CV... )
They have two major products currently, peak energy with backside passivation making it perform better during morning / evenings than standard modules (better red wavelength sensitivity also giving it less temperature coefficient and NOCT operating temperature.)
The other product is lead free and more environmental friendly modules.
What do I base my prognosis on?
First of all I have to stress it is very hard to scientifically predict 1,5 year ahead (6 quarters) ahead in time but these are educated guesses based on historical performances.
Historically price reductions have been at around 10% per quarter. Recently however the ASP has actually increased and it is speculated that the demand has finally catches up with supply. Also there has been a recent trend of tariff differentiated market and different ASP in different world regions. Japan currently has some 30% higher ASP than Europe and even more so than China. Canada is also a high ASP region. The American market is already blocked slightly for Chinese production because of tariffs and they have to add cell costs from outsourcing cells to circumvent tariffs. Many European importers have worried about buying Chinese modules now because of pre registration in case of European tariffs, this has also put pressure on ASP in Europe.
If one looks at the ASP reductions year over year one sees the trend has been lower ASP but less so in high growth periods than in overcapacity markets. Recently however ASP for modules has reached a price level competive with other energy sources and modules are now a minor part of the total energy cost of the solar systems (about 1/3 of the cost.) This gives a compelling argument for price stabilization of modules and perhaps even a small increase in ASP.
2009: -36% ASP (Challanging market)
2010: -17% ASP (Growth market)
2011: -25% ASP (Start of overcapacity)
2012: - 36,4% ASP (Overcapacity and bankrupcies)
2013: +10% ? (grid parity and tariff differensiated markets causes ASP to increase)
2014: + 5% Explosive growth causing further ASP increase while systems cost go down due to less soft cost and balance of system costs like inverters
Good sites to follow the spot prices are photon international newsletter on module spot prices ( for example in march newsletter: www.photon.de/newsletter/document/74961.pdf )
Pvinsights on wednesdays (pvinsights.com/ ), energytrend pv.energytrend.com/ and solarzoom www.solarzoom.com/
How about FBR in rec silicon?
In 2010 the prices was down -13%. (growth market)
I 2011 the prices was down -12% (start of overcapacity)
I 2012 the prices was down - 52,65% (overcapacity and bankruptcies)
I 2013 I believe the ASP will increase because of the major market in china putting a tariff on imports, this will cause ASP to increase to allow polysilicon to be imported still while having a profit for the selling company. The argument is simple, there is not enough high quality polysilicon available in china at the price levels in the market currently. The tier 2 Chinese companies have price levels above 30$/kg and so prices can go between 25-28$/kg. I believe a price around 28$/kg for 2013 is possible.
2014 with a growth market should see prices even further up. But there are a lot of capacity coming online in start of 2014 so the first two quarters I think price will fall back down to 25$/kg.
Good sites for spot prices are sunsirs trade prices (before purchases are made so a very early indication of future price trends) www.sunsirs.com/uk/prodetail-463.html
Also 100ppi.com has more detail but requires a Chinese translation. www.100ppi.com/
Finaly you have pvinsights on Wednesday pvinsights.com/ and solarzoom www.solarzoom.com . Energytrend pv.energytrend.com/pricequotes.html and new energy finance about.bnef.com/ (payed service) also track polysilicon prices. Photon international also does this as a paid service. www.photon.info/photon_site_consulting_e...?ActiveID=1116
How about factors in the future?
Tariff differensiated market:
As I have calculated (seekingalpha.com/instablog/901705-eystei... ) there will not be a big enough market for chinese modules in 2013 (with 15-20 GW of real capacity) so chinese modules will cause price pressure. If Europe, india and America all have tariffs these markets will have higher ASP as the Chinese will not be able to compete in these markets. The high Japan ASP will continue but have more pressure, the reason is that the Japanese only want high effiency modules and only a small portion of the Chinese capacity is at 250 WATT or more per module. The big question is how much the chinese themself take of the market in 2013 ? With subsidies up to 4$/watt the 10 GW goal of 2013 seems possible (source www.pv-tech.org/news/china_calls_for_dom... )
When it comes to the Chinese tariff on polysilicon the rumors have it this is only for polysilicon used domestically and not for modules shipped internationally. Also there are not enough capacity of low cost polysilicon in china so the net effect will simply be that ASP will increase for polysilicon in China and most western companies will still sell to China without a tariff penalty.
For rec the tariff market will have benefit since they are producing in Singapore. They will be able to enter the Japanese market (already a major contract now in q4 2012) and still enjoy the fruits of higher ASP in Europe, India (also having an anti dumping investigation so if tariffs possible higher ASP) and USA. Since REC are heading for over 95% modules over 250 watt (45% in q4 2012) they will also have an easier time selling in the high effiency market of Japan. That statistics historically says that Europe will enter tariffs with china (2011 eu tarrifs out of 45 started investigation 11 was terminated without effect and 75,5% was put into effect. source: trade.ec.europa.eu/doclib/docs/2012/june... .) Also there has been calls in the EU parlament for faster implementation of tarrifs: www.pv-magazine.com/news/details/beitrag...)
Even without the tariffs in place the pre registration has allready caused havoc on european module prices:
I also belive REC will have a price premium because of increased module watt performance.
(www.recgroup.com/PageFiles/2422/Real%20W... and www.recgroup.com/PageFiles/2422/REC_best... ) Also since rec has a very low defect ratio the insurance premium should be low. A study showed that between 5 - 22% of modules in the study was defect with an average of 8.8%. (Source www.renewableenergyworld.com/rea/news/ar...)
I also belive 2013 with be the start of the big bankrupcy wave. Suntech have already defaulted on bonds (source: www.pv-tech.org/news/cross_defaults_trig... ) and it is speculated some capacity will be gone from the market and some will be taken over by the government. Many minor producers have already gone bankruptcy. I tried to contact a list of some 600 tier 2 & 3 producers and less than 10 had still working emails.
The production capacity of the world is probably shrinking because of all the old equipment and factories shutting down. I belive the market currently are overcalculating capacity in the world and the current upswing in ASP is due to capacity more aligned to demand in the market. Pvsolarinvestor.com has shown that from 2011 capacity of 38.7 GW the capacity in 2012 shrunk to 32,5 GW. (Source: solarpvinvestor.com/database/capacity/wa... . ) My own data shows shipments in 2012 from tier 1 producers shrunk 11% compared with 2011. I think also for 2013 some capacity will be gone, perhaps from a major player like Suntech. At the same time 2013 shows signs of a growth year and I think the market will be around 34-35 GW for 2013. Allready now 11 countries in EU have reached grid parity. (Source: www.pvparity.eu/news-on-homepage/news-si.../ .)
Historical growth numbers:
If you look at historical growth numbers from EPIA ( www.epia.org/uploads/tx_epiapublications... page 12 ) you can tell the growht has been huge from year to year.
2003 578 MW
2004 1114 MW (92,7% growth)
2005 1429 MW (28,3% growth)
2006 1575 MW (10,2% growth)
2007 2529 MW (60,5% growth)
2008 6330 (150% growth) MW
2009 7436 MW (17.4% growth)
2010 16817 (127 % growth) MW
2011 29665 MW. (77% growth)
2003 201 MW
2004 708 MW (252%)
2005 1002 MW (41,5%)
2006 987 MW (-1.5%)
2007 1972 MW (99,8%)
2008 5297 MW (168,6%)
2009 5803 MW (9,6%)
2010 13367 MW (130,3%)
2011 21939 MW (64,1%)
Average 95,55% growth.
Estimated 2012 numbers: 21634 MW (- 1,4%) (policy driven)
Estimated 2013 numbers: 16591 MW (- 23.3%)
Estimated 2014 numbers: 19041 MW (14,8%)
All in all the world market will probably increase in 2013 while the european market will shrink. Pure logic implies that if the market grows and the production shrinks the market will return to more normal prices. Also there is a trend that after a low growth year usually the market explodes again. The expetion is 2005-2006.
One last factor that should be mentioned is Balance of system (BoS) costs. They now are a much higher part of the overall system costs. (from half to one third of a solar system cost.) Many have said that BoS costs can be reduced further and all signs show that this is happening now. The net effect is that prices are reduced for systems, while allowing the modules some less price pressure. For interested readers they can read more about BoS costs in the following articles:
www.greentechmedia.com/articles/read/Sol... og www.greentechmedia.com/research/report/s...-2013
A short summary is that the wave of bankruptcies will have hit during 2013, butt his will not affect core markets very much due to tariff barriers that will not allow tier 2 & 3 Chinese manufactures to dump products in the countries with tariffs. At the same time we are moving from an overcapacity market to a growth market in the new grid parity age we have entered so this will ease the pressure further on module prices. Additionally Balance of System costs will ease the pressure on module prices.
If we look at cost reductions for REC modules and fbr historically it has been the following:
2009: -13% for modules (big increase in capacity)
2010: -36% for modules (fbr beginning to be used for modules)
2011: -30% for modules(q32010 to q32011) Fbr -40% (q3 2010 to q32011) (the effect of cheap fbr and fully ramped up Singapore)
2012: -42,56 % for modules, -17.3% for fbr (Effect of lower prices in the whole chain together with price reductions from Singapore and Moses Lake.)
2013: -20.% ? for modules and -7% for fbr. (effect of new raw goods contracts on MGS and continual improvements + upgrade of equipment in Singapore.)
2014: -20% ? and -10% for fbr?
Most of the cost reduction in FBR will come from cheaper raw goods. For Singapore the raw goods are already at rock bottom prices and it is more about upgrading equipment like the centrotherm purchase of new diffusion equipment. (www.centrotherm-pv.com/en/press/news/new... )
Exchange rates used in report
Taiwan dollar/USD 0,0337
Not only the costs of goods make up the costs of a module. For every shipment of a module you have to add costs to research the technology behind the next generation of modules, you have to add the costs of your loan interests and you have to add the costs of selling, freighting and administration of your module sales. The following tables looks at costs when you also add interest costs divided by number of shipments in mw and sales general & administration + Research & development divided by shipments.
The q4 report specifies on page 37 what segment has what debt.
Rec silicon has 7469 mil nok liabilities while rec solar has 2196 mil nok liabilities total between the two divisions is 9665 mil nok. The ratio is then 22.7% interest costs to rec solar and 77,3 % to rec silicon. For interest rate this ratio will be used to spread out costs of financials between rec solar and rec silicon.
For capex the asset ratio will be used. Rec silicon (10307 mil nok) and rec solar (3728 mil nok) has a total assets of 14035 million nok. Ratio for capex will be 26.5% to rec solar and 73,5 % to rec silicon.
Since sg&a and r&d is known the ratio used will be the one presented in the quarterly material from rec itself.
Float zone, Electronic grade, simens sg, off spec and silane gass sales:
Total cash flow profit from these volumes has varied between 80-150 million nok in the last quarters. In q1 rec got 27 million $ (155 mil nok) compensation for a silane gas contract. I expect the cashflow from these products to be around 55,72 million nok (10 million $) giving a total of 27 + 60 = 87 million $ during the next 6 quarters. ( 498 million nok positive cashflow.)
Note the high capex during 2010. At this time REC was still purchasing new equipment for Singapore and Moses Lake. The 100-60 mil $ per year is more the run rate for smaller maintenance and small technological upgrades.
2014 bond. www.recgroup.com/en/ir/Debt-and-credit-f.../ Maturity date 16/09/14. (Q3 2014) 11% interest rate. Of the 2014 bond only 650 mil nok is left. (113.5 million $.)
320 million euro 2014 convertible euro bond. www.recgroup.com/en/ir/Debt-and-credit-f.../ Maturity date 04/06 /2014 (q2 2014) 6.5% interest rate. The 2014 euro convertible bond have a left over of 2358 mil nok. (411.7 million $.)
500 million nok 2016 bond www.recgroup.com/en/ir/Debt-and-credit-f.../ Maturity date 03/05/2016 (Q2 2016) Nibor 3 months + 4.35% interest rate. The 2016 bond have 500 mil nok left. (80.3 million $.)
700 million nok 2018 bond www.recgroup.com/en/ir/Debt-and-credit-f.../ Maturity date 03/05/2018 (q2 2018) 9.35% interest rate. The 2018 bond have 700 mil nok left. (122.2 million $.)
246 million nok indeminifcation loan. 2014 46 million nok (8 million $) maturing and 2016 200 million nok maturing. (34,92 million $.)
Details of owed amount is found in note 6 in the q4 report:
Total debt maturing in 2014: 533,2 million $ / 3054 million nok.
Financial expenses - comparison with Peers.
Research and development
A few details of modules currently in production and how research budget has affected temperature coefficient and other technology aspects:
Renewable energy corporation:
Backside passivation modules. Eco friendly modules with less lead. Fluidized bed reactor silicon produced modules that gives lower energy payback time. Automated production gives quality advantage. Future patent of easy to lock modules with no effect on temperature coefficient. Temperature coefficient pMax 0.40 on peak energy. www.recgroup.com/PageFiles/2422/REC%20PE....pdf
Multi: Watt 235-260 watt c-si. Eco (less lead) pMax 0.46 235-255 watt c-si. 1665x991x38mm 18kg. Noct 45.7 +-2.
N-type wafers. "Panda" technology. Mono technology. Low temperature coefficent (pMAX 0.42% per celcius above 25) Low initial light induced degradation good performance at low radiation (94% at 200 Irradiance w/M2) www.yinglisolar.com/assets/uploads/downl....pdf
Mono: 250-280 Watt. 1650x990x40 mm weight 19.1 kg. Temperature coefficent pMax 0.42%. Noct 46 +-2.
230-260 Watt. 1650x990x40mm. 19.1 kg. pMax 0.45%. Noct 46 +- 2.
Honey cell technology. Trinamount. (Mounting technology for lower BoS)
Multi: Noct 45 +- 2. pMax 0.43. 18.6 kg. 1650x992x35 mm.
N-type wafers. (From nexolon) 20 % cell efficiency. pMAX 0.38%. Noct 45 +-2. 18.6kg. 1559 x 1046 x 46. 327 watt.
Virtus I & II technology.
virtus i: pMax 0.38 Noct 45 +- 2. 19 kg 1640x 992 x 40 mm. 250-260 watt.
Virtus 2: pMax 0.30 Noct 45 +- 2. 19kg 1640x 992 x 40 mm 250-260 watt.
Multi: 1650x 992 x 45 mm. 19kg. pMax 0.43%. Noct 45 +- 2. 240-260watt.
Mono: 1650 x 992 x 45 mm. 19kg. pMax 0.43%. Noct 45 +-2.
Multi: 18.5kg. pMax 0.45%. Noct 47 +- 2. 235-260 watt. 1650x 991 x 40 mm.
Mono: 15.5 kg. pMax 0.43%. Noct 45 +-2. 245-270 watt. 1646x987x23mm.
multi: 72 cell. 270-295 watt. Noct 45 +-2. pMax 0.45%. 26kg. 1966x1000x50 mm.
Mono: 60 cell. 240-265 watt. Noct 45 +-3. pMax 0.47%. 20kg. 1652x 1000x45 mm.
1200 x 600 x 6.8 mm. 12kg. 154 cell. 82.5 - 92.5 watt. pMax 0.25%. Noct temperature not specified.
Multi:235 - 255 watt. 0.43% pMax. Noct 45+-2. 1638x 982 x 40 mm. 19kg.
Mono: 240-260 watt. 1638x 982 x 40 mm. 19kg. pMax 0.45%. Noct 45 +-2.
255-270 watt. Pmax 0.43%. 19.1kg. 1640x 990 x 40mm. Noct not specified.
pMax 0.5%. 19.1kg. 1640x990x 40mm.Noct not specified.
Multi 230-260 watt.
Noct 45 +-2. pMax 0.43%. 1640x992x35mm.
240-260watt. Noct 45+-2. pMax 0.44%. 1640x992x35mm. 18.2kg.
MK 2 and MK 3 wafer technology.
Noct 45+-2. pMax 0.45%. 1642x 994 x 40 mm. 19kg.
Noct 45+-2. pMax 0.47%. 1642x994x40mm. 19kg.
Currency gain / loss and fair value hedging
During the last 3 years REC has come out ahead when you compare the loss from currency with the hedging and deratives done. (310 million $ ahead.) Since these losses/gains are real in a cashflow sense that means they should be added to any cashflow estimates. I have conservatively estimate 0 income / loss on these hedges and currency gain / losses , but historically obviously REC has done better than this.
Depreciation & Amortization
These are not cashflow costs but obviously are of interest for operation profit margins. (Numbers in million $)
Interestingly out of some 600 respondents of Chinese tier 2 & 3 module manufactures only a handful (under 10) had working emails and responded to my enquiries about capacity. Since the total tier 2&3 manufactures was around 1100 companies and a majority of the Chinese companies had no longer working emails it is possible a lot of this capacity is now offline or dormant waiting for ASP to approach 2011 levels above 1$/watt.