Here is what Odyd wrote:
As you are aware, public companies are responsible for reporting material events within ten business days. The sale of the Chinese plants was announced on April 12 certainly referring to a transaction which took place in March or within Q1 during the release. The conspiracy theory which is the core of your article is that CSIQ pulled sale of those plants to mask losses or hide the lack of efficiency.
The crux of your observation is that recognition of the revenue should have taken place at the time of the announcement, and since was not stated it had to be a collusion.
I am sure that you are aware of accounting cycles and processing which requires step by step analysis of every transaction.
In Q4 cc on March Qu made clear reference to projects to be closed in Q1. He addressed them as profitable, which did not escape your analysis. I quote: “Now, the project we’re going to close in China, a good project; it should have something like high teens to 20% gross margin. Those are very profitable projects.”
All facts lead to a conclusion that transaction took place after March 21st and before March 31st. The company complied with regulatory obligation via news release but did not complete accounting cycle on the transaction. They would not, as the period was not over. When I read the news, I was hopeful they would recognize it in Q1 not that it would be recognized in Q2.
My comments on the loss. I agree that not having this transaction would impact the result, but this falls into an alternative outcome scenario no different than stating if Jinko did not have 25% OEM modules sold it would have a loss despite improved margin. Facts should remain facts and be reviewed on the ground what happened and not what would happen “if.”
My comments to alleged lack of competitive edge. I have taken 176MW of total solution where plants sales are recognized and applied 18% average to the gross margin. Upon extracting $5M revenue for solar systems selling electricity and extracting 50% gross profit from it, I have arrived at modules sales with ASP over $0.41 per watt and the cost of $0.36 blend or 12.1% GM. Those results indicate a couple of things, Canadian continues to receive higher ASP than Chinese peers, and their cost is coming down. The cost which has been still impacted by tariff payments as per Qu.
I quote from Q4 cc on March 21st: “Canadian Solar has ramped up its production facilities in southeastern Asia in the course of 2016 and has not imported solar products from China to the U.S., or using Taiwanese solar cells in its solar products shipped into the U.S. since February of 2017.”
This means the cost of modules sold in Q1 to the US had tariffs until February or half way the period. If we say this is 1 cent, the cost is relatively the same and in line with Jinko. Let’s further use Jinko as a point of relationship here. Based on capacity ending on Q2, the cell to module penetration for JKS is 53% versus CSIQ being 64%. Wafer penetration to the module for JKS is certainly beating the CSIQ, which will find this closer to cell capacity only by Q4. One item needs to be pointed out that wafer expansion for Jinko is in mono, the cost of mono polysilicon is greater than that of poly. While both use DWS, a black silicon wafer for CSIQ is going to cost a lot less. I expect silicon cost for Jinko to be slightly higher than Canadian for this reason and therefore overall costs. Why is this important to margin? If you look at the efficiency of Ku modules offered by Canadian, the mono-PERC modules produce at the end of the scale 10W more than JKS, and the black silicon modules meet a rating of mon-PERC made by JKS. If Canadian continues to get higher ASP, at least at a minimum, both companies have equally good chance to have similar margin, where I believe Canadian will have lower costs by the end of Q4.
Lastly plants. Based on operational plants in three categories on balance sheet, the value is currently set at $1.4B while the cost of $1.2B is reduced by electricity generation. The company insists on having $167M of gross profit or a gross margin of 10.5% quoting $1.6B. As you may know, Brazilian plans which were sold before COD will have to be revenue recognized based on completion, so they are not included. Using about 60% drop of GM into net profit, Canadian can probably collect $100M or $1.73 per share from it. I did agree with you that margins would become compressed, but as soon as Trump puts Suniva MIP in place, plants will gain in value, and I can see this already play a role in negotiations. You also learn about margin compression from FSLR and SPWR which had written their plants, and particular SPWR which cost is beyond anything we could consider competitive.
While many of my points as well as yours remain to be seen, I consider Jinko weaker financially. Particularly with the CapEx needs and the liquidity curtailed by the immense amount of revenue in account receivables which pose a huge risk at a minimum to the execution of business goals, and at worse with a loss if not paid."