Canadian Solar (CSIQ) once again delivered another mixed bag with their first quarter 2011 earnings released last week. Unlike Trina Solar's (TSL) and Yingli Green Energy's (YGE) pre-announcement on lower than expected shipments last week, CSIQ much like Jinko Solar (JKS) actually reported relatively strong shipment volumes. This was surprising since Canadian Solar was generally not recognized as a first tier Chinese module supplier like TSL or YGE. At least for the first quarter, secondary tier names apparently performed better than their larger peers in terms of maintaining shipment volumes.
For the first quarter of 2011, Canadian Solar's shipments expanded to 244mw, up from 237mw in the prior quarter and up from 185mw reported a year earlier. Again, CSIQ's sequential shipment growth was impressive as it came in a seasonally weak quarter for the industry and in light of relative weakness among larger peers. Revenues however declined sequentially from $452.7m to $443.4m due to lower average selling prices ("asp"). Gross margin also contracted to 14.7% from 17% on a sequential basis as unit costs remained under pressure due to high polysilicon and wafer procurement costs.
Net income fell to $5.9m mostly due to a large net foreign exchange loss as well as higher interest expenses. On an earnings per share basis ("EPS"), Canadian Solar posted a .13 EPS in the first quarter which was well short of Wall Street consensus of .41 in EPS. The $17.9m in net foreign exchange loss CSIQ recorded in Q1 accounted for .41 in EPS loss. It's unclear what magnitude of foreign exchange translations were reflected in Wall Street estimates. Factoring out all foreign exchange translations, CSIQ earned $23.9m which fell short of my $27m ex-forex net income estimate. Despite posting gross profit inline with my estimates, lower net income resulted from higher than expected operating and interest expenses.
Gross profit relative to shipment volume is the best gauge of a company's earnings power. However despite how close my gross profit estimate was relative to CSIQ's shipments ($65.3m on 244mw vs. my $65m on 245mw estimate), the cost metrics I used did not parallel statements made by the company. As I noted in my prior article, statements made by Canadian Solar's management did not coincide with the company's actual results. In short as previously stated, cost metrics mentioned by the company should have yielded much higher levels of gross margin than the company reported.
Part of the difference may be the result of a miscommunication. It's possible the company's stated procurement costs are accurate, but the actual costs realized in the quarter were higher on a blended average basis. This difference however cannot fully explain the gross margin gap. In other words, Canadian Solar continues to suffer from cost issues not entirely disclosed by the company, intentionally or otherwise. Management did previously mention near term material processing cost increases such as silver paste as one of the reasons for higher non-silicon costs. Other processing cost inefficiencies may also be present. In CSIQ's Q1 2011 conference call, management did guide total cell and module processing costs to trend towards .50/watt in Q3 2011 which would be more inline with peers.
In addition to the strong first quarter shipments Canadian Solar reported, the company also expects second quarter shipments to be even stronger. For Q2 2011, CSIQ expects recognized shipments to range between 245-255mw. An additional 100mw was also shipped under "delivery duty unpaid" or essentially cash on delivery terms. As a result, these shipments will be recognized in the following quarter when it reaches its destination and after payment is received. In essence, Canadian Solar sees demand strong enough in the current quarter to effectively ship approximately 100mw more on a sequential basis, representing over 43% quarter over quarter shipment growth.
Although 100mw of revenues won't be recognized until the third quarter, second quarter blended unit costs should reflect the cost of this volume. This will in effect blend unit costs higher since CSIQ has to procure cells in excess of their cell capacity to fulfill the 350mw of second quarter demand volume. Luckily for Canadian Solar, wafer and cell asps have fallen significantly since the start of the second quarter. The bad news is that spot market module asps have also declined equally in magnitude. The combined effect should lower CSIQ's gross margin, per watt gross margin, and thus gross profit in the second quarter relative to the prior quarter.
Canadian Solar's official Q2 2011 guidance is for shipments and gross margin to range between 245-255mw and 13-15%, respectively. Much like direct peers recently, CSIQ also reiterated their 2011 shipment guidance of 1.2-1.3gw. Based on this shipment guidance as well as cost metrics detailed in their Q1 2011 conference call, the following estimates have been compiled for Canadian Solar's second quarter.
CSIQ Q2 2011 Earnings Estimate:
Shipments: 250mw module, 10mw system kits
Asps: 1.60/watt module
Unit Costs: 50 x 1.35 = $67.5m, 170 x 1.36 = $231.2m, 130 x 1.42 = $184.6m
Per Watt Unit Cost: $486.3m / 350mw = 1.38/watt
Gross Profit: 250mw x .22/watt = $55m module + $1.5m kit = $56.5m
Gross Margin: $56.5m / $415m = 13.6%
Operating Expenses: $30m
Net Interest Expense: $8.5m
Net Income: $14m
Diluted Share Count: 43.7m
The estimates above only represent operational earnings. Non-operational items such as net foreign exchange translations as well as unannounced one time gains or charges are not included. In terms of net foreign exchange effects, it is impossible to estimate with half a quarter of currency trading left. As of today's date, overall rmb vs. usd and euro vs. usd exchange rates have been stable relative to the end of the prior quarter. Thus as of today, Canadian Solar should experience very minimal net foreign exchange gains or losses. The actual degree can vary dramatically since the company could change exposure or hedging ratios at any time even if key exchange rates do not materially change for the rest of the second quarter.
CSIQ's $17.9m of net foreign exchange loss in Q1 was a bit surprising since the company realized losses in both currency translation as well as from hedging. As the term implies, hedging should balance risks. Currency losses might be offset by hedging gains or vice versa. For CSIQ to report losses in both metrics implies the company made more directional bet against the appreciation of the euro well in excess of its euro denominated asset risk. In fact, given CSIQ had 75.7% European exposure in the first quarter but only reported a small $0.7m currency loss suggests the company was almost fully naturally hedged, ie European revenues were collected in usd or other currencies less volatile than the euro.
In addition Canadian Solar noted in their earnings conference call some EPC projects may be completed at least partially during the second quarter and thus some revenues may be recognized. Without proper guidance on the degree CSIQ's systems business may impact current quarter's revenues and gross profit, it would be impossible to even attempt an estimate. Based on the statements made by the company, it appears revenues from their systems business may range in the low tens of millions for Q2 2011.
Lastly it also should be noted while Canadian Solar's first quarter sales as well as second quarter shipment guidance was exceptionally strong, the company's accounts receivables also increased by a significant amount on a sequential basis. Accounts receivables ending Q1 2011 were $306.7m, or up 80% from Q4 2010's level of $169.7m. Management did explain the increase was mostly attributed to higher volumes shipped late in the quarter. Given the company's past history of questionable business practices which led to write offs of doubtful accounts, such significant sequential increases in accounts receivables during a quarter some industry peers cited weakness should at least be recognized.