36 replies to this topic

[odyd]

I got few words on deceleration of shipment volume for December. Interesting numbers and locations. For those who purchased data, those who have been chugging along, things seem to be going well into a direction, with some companies planning to sell a lot to exotic places. South Africa has showed up as the destination as well. Also some details are showing for OEM work from CSIQ and cell shipments to Canada have some visibility. In all China has to be a big play for some of the players otherwise they will not meet it their guidance. Once again for those who are purchasing data this is a lot clearer.

[explo]

Interesting, when do you think the report will be available for purchase?

[odyd]

I am thinking in the area of the 25th of Jan. Now it could be later as Jason does a lot of sorting out and that takes couple of days.

I will have notes added to the next issue of it, with some of the observations made by Jason based on declaration. Looks interesting.

[odyd]

According to IHS the Q4 combined shipments are 11GW. We have captured 2GW in two months and do not include China.
http://www.pv-tech.o...e_in_2q_in_2013

I got to be honest I do not know where is this number coming from. It would be interesting, as they are speaking about "leading Chinese companies" to understand what they mean. In short I do not see this.

[explo]

Let's see, we have 2 GW in two months exported from China. Naively linear extrapolate to 3GW in three months. China market is big for the Chinese companies in Q4, so assume only 50% goes to exports in Q4. This means the Chinese produced 6 GW in Q4. Assume that on modules Chinese players have 60% of market (its higher on wafers), then we coluld have 10 GW globally shipped in Q4 as an estimate extrapolated from the export data with crude assumptions. So the export data does not have to contradict the 11 GW global shipments and Chinese increasing their share in Q4, but I agree the number seems big. If true and combined with the low utilization in 2H12 the health of inventory levels should have improved.

Sidenote, I think the finer detail patterns rather than the macro patterns is where the China export data contribute the most, especially when looking at trends over many months. I look forward to see the December numbers.

[odyd]

Well I did somewhat take this approach but in the statement they have said couple of things which alarmed me. They have said leading Chinese manufacturers shipped more than was expected. Unless they were expecting less than guidelines, this would make sense, but as you are aware there are certain parts of the picture, which are not going to happen. I think that 50% for China assumption is good, but this means that they shipped 3GW of modules into domestic market, this is more than they shipped in 2011 and first three quarters combined. I think Solarzoom would noticed. Lastly going with this pace, who would ship the 4GW? This is a massive figure, I am not seeing this to be honest. They also talk about 10GW of installs I believe. I approached them to give me the idea of the breakdown. I am rooting they are right, but their calculation is a combination of many sources and estimates. Sounds like we have more concrete data, but less attractive one.

[odyd]

Dr de Haan replied to me and called own report a forecast and being a model. He referred to using qualitative confidential data from Chinese suppliers. Well, we will see.

[odyd]

Based on our numbers export fallen around 41%, according to this article it was 33%
http://www.solarserv...ember-2012.html
Close enough considering that maybe some ASP adjustments took place or product mix was different. If we stick to dollars and watts using our numbers for November in US, we are looking at average of 0.69 per watt.

[explo]

That's interesting having both the MW and USD we can calculate the ASP for the market. It's not possible to get the USD directly in the export data I guess?

By the way, when do you think the December data will be available for purchase (I think you projected today previously)?

[odyd]

I was told data arrives on 25th and it is processed for couple of days. I would imagine we should see it next week. On dollars and cents we had this talk before. It is considered too sensitive for markets and companies, and not cost effective for SPVI to have it. It is apparently available on Solarzoom database with subscription. I am hoping that we can get more buyers of it so they do not pull the plug on us.

[explo]

May USD total per market, but not for individual companies is less sensitive? Anyway the report will be intersting reading.

[odyd]

Apparently there is a limit to what they want to sell at SPVI's price. And it is about markets as a whole. They do not want average pricing for a country market to be seen by anyone. It is obvious to see what companies sell in a space, to create underselling condition for them. I suppose if we had 1000 reports sold a month, I could insist on adding something like that for large following. Selling fewer than 10 to 15 a month is not going to pay for someone’s time to crunch numbers, I was told. Once again I dropped the talk.
I hope we get enough buyers that we can create sort of “a la carte” reporting, however they have 120,000 subscribers. I am disappointed with very limited interest in those reports. I thought the Western investor would be more keen to learn more on this.

[odyd]

Hi All, December data is here now.
Enjoy, please download a Q4 Export Data Presentation in the pdf format
http://solarpvinvest...om/paid-reports

[explo]

Thanks. Another piece of interesting data. New markets emerging and for the Chinese the US market seems to be rather submerging. Some solar 11s have significant exports to multiple markets outside Europe and North America.

[odyd]

Have you looked at the presentation?
It offers a look without numbers.
http://solarpvinvest...esentation1.pdf

[explo]

Have you looked at the presentation?
It offers a look without numbers.
http://solarpvinvest...esentation1.pdf

I missed that in the rush. Very nice presentation. It shows just some examples of the kind of information you can filter out from this data. With the raw data you can get monthly resolution, actual numbers and look at all top 30 companies and the residual group. I think your deck is a good illustrator of the data value.

[explo]

odyd, what will be really interesting now that we have the export details is to look at the Q4 ASPs. We know that some have huge had domestic shipments and that ASP is very low for that (lower shipping costs compensates a bit for this), so this could give an indication about the geography component influence on the ASP. I've been looking for a model for ASP for quite some time and I think some answer will come here. For now I've assumed that we have the following influence in order:

1. Made-in (Germany and Japan high, China very low)
2. Sold-in (Japan very high, US good, Germany low, China very low)
3. Conversion efficiency (each 5w power or 0.3% efficiency in a 60x6" panel can add two cents value)
4. Brand value. First the bankability lists and now the independent warranty insurances have neutralized much of the economic value arguments of brands (do I want to sacrifice some of my IRR to sponsor World Cup or Formula 1?). Previously banks, now insurance companies have done the DD for us.

Point 1 seems to be written in stone, though I find it a bit hard to understand the huge differences. How are Japanese designed modules manufactured in China priced by the way? Point 2 makes sense, different markets have different price levels. Point 3 should be a given, you want to maximize your IRR, conversion efficiency contributes to this. Point 4 is debatable, that's why the Q4 ASPs will be so interesting. If YGE and TSL (tier 1-1.5) are below CSIQ, HSOL (tier 1.5-2) and SOL (tier 2-3), then point 2 (or 2&3 in case of SOL) outweighs point 4.

[explo]

Have you looked at the presentation?
It offers a look without numbers.
http://solarpvinvest...esentation1.pdf

The nice and free presentation is worth a direct teaser here:

[explo]

Going through the data and my expectations from guidance, some warning flags are seen and some bright lights are seen.

Warning flags: JKS, TSL. Will miss unless they more than double already guided significant amount of domestic shipments.

Bright lights: SOL, HSOL. Will significantly beat if shipping expected domestic amount. Might beat even if shipping much less than expected domestic amount.

CSIQ and JA are aligned with my expectations. CSUN slightly above. YGE are below export expectations, but have already raised guidance, which suggests that they shipped maybe 100 MW more domestically than guided.

Above is based on exports data, which don't have to align timing wise with shipment recognition, but comparing quarterly exports with guided quarterly run-rate should still make sense.

The other 3 I'm not bothering about. STP hasn't reported 2012Q2 yet and their website have been down for more than a week now. LDK is brutally dismantling (cell capacity gone, module sales gone, wafer contracts cancelled) probably for going into hibernation until the market is good for using their poly and wafer assets. DQ is doing impairment testing of PPE (which is 80% of their assets) and module business is gone. LDK and DQ had basically no module exports in Q4. STP had 40% less exports than SOL in Q4, to give a picture of how things are going for the old no 1.

Note: China was so huge in Q4 that anyone can compensate weak exports with big a domestic shipment surprise. So although JKS, TSL, STP and LDK has worrying low exports, it does not have to indicate that they'll miss guidance. YGE proved that.