Unlike the majority of the photovoltaic (PV) companies, Canadian Solar (CSIQ) once again performed exceptionally well in the second quarter as evident in its earnings report. Unlike more integrated large scale Chinese peers such as Trina Solar (TSL) and Yingli Green Energy (YGE), Canadian Solar’s cost structure is not the lowest among its peer group, but it is low enough to generate positive net income. During a period where so many smaller participants are struggling for mere survival, financial strength is a key variable for one of the most important criteria in determining demand for a company’s products - bankability.
While CSIQ was considered more as a second tier solar module provider in the past, its cost structure which was less evident during periods of higher average selling prices (asp) has moved the company ahead of entrenched brands in terms of product bankability as module asps declined below many peer’s ability to manufacture. PV modules used in multi-million dollar solar projects must be guaranteed to operate at a specific level for upwards of 20 years. As price declines have put the survivability of many less competitive producers into question, lower cost providers such as Canadian Solar have been able to gain market share simply because many peers have become relatively less bankable.
Canadian Solar’s second quarter module shipments grew to 287MW, up from 244MW recorded in the first quarter. The company’s first half 2011 shipments of 531MW surpassed its second half 2010 volumes of 437MW. CSIQ’s continued growth in shipments is incredibly impressive since the first half has typically been much weaker than the second half. In comparison, the industry’s installation rates in the first half of this year were significantly lower than the second half boom of 2010. Similar to other large US listed Chinese peers, Canadian Solar has been gaining significant market share at the expense of higher cost peers even ones with significant branding advantages in the past.
Market share gains have come at a cost however. As noted in my recent Canadian Solar update, module asps along with pricing across all verticals of the value chain have fallen dramatically since early May. The pricing declines which have made CSIQ relatively more competitive simply because higher cost structure peers became priced out of the market has hit its own margins. Much like direct China based peers such as Trina Solar and Yingli Green Energy, Canadian Solar still has room to reduce its unit production costs to help offset end product pricing declines. Although cost reductions may not be realized immediately due to inventory blending effects as described in prior articles, CSIQ’s module unit cost is still low enough to generate a reasonably positive gross margin.
During the second quarter, the company’s gross margin declined from 14.7% in the first quarter to 13.2% as module pricing dropped at a faster rate than the company could blend down its inventory costs. Part of the gross margin contraction was also due to higher outsourcing ratios as more cells were procured to satisfy demand above the company’s internal upstream capacities. While the simplistic generalized view on declining gross margin is negative, this is not necessarily true for solar companies with varying layers of integration. Companies such as Canadian Solar which use a more flexible integrated structure may choose to procure more silicon wafers or solar cells to supplement internal capacity during periods of high demand. Lower ratios of in house production would generally translate to lower blended gross margin. However, products using outsourced components in most cases still generate positive margin otherwise the company would not do the additional business. As a result, overall gross margin would decline but at the benefit of incremental gross profits.
Thus while Canadian Solar’s gross margin did decline marginally, overall gross profits stayed relatively flat on a sequential basis at $63.7m vs. $65.3m in the first quarter. Operational expenses related to higher shipment volumes, increased research spending, and higher interest expenses caused a sequential decline in operational earnings which fell from $23.9m to $13.9m vs. my $12m estimate. CSIQ was still able to surpass my operational net income estimate despite higher than expected operating costs due to incremental revenue and profits from its systems EPC (engineering, procurement, construction) business segment. I noted in my CSIQ Q1 review, additional revenue from this new segment could be realized but since the company gave very little detail it was impossible to estimate accurately and thus was excluded from my estimates.
U.S. GAAP net income for the second quarter came in at only $7.1m due to non-operational foreign exchange and hedging losses as Canadian Solar remained over hedged against the euro. Total losses relating to currency was $6.9m in the quarter or roughly half of the company’s operating profits. With the Euro currently down vs. the USD by slightly over 6% in the third quarter, much of the hedging losses CSIQ incurred in the first half of 2011 should reverse assuming the company kept its hedging activities constant. U.S. GAAP earnings per share (EPS) was 0.16 in Q2 which missed Wall Street consensus of 0.28 per share. As usual, it is unlikely analysts factored non-operational items such as foreign exchange effects into their estimates because CSIQ’s prior guidance which the company did not miss implied higher operational results. Excluding the $6.9m net foreign exchange loss, Canadian Solar’s Q2 operational EPS was 0.32 per share.
For the third quarter, Canadian Solar expects shipments to increase to 350-360MW in part due to a delayed recognition of a large volume shipped but not recorded as revenue in the second quarter. Module selling prices are expected to decline to around 1.30/watt from the 1.55-1.60/watt range realized in Q2. As with other peers such as TSL and YGE, module asp declines should exceed the rate CSIQ will be able to blend down costs. Therefore gross margin should also decline on a sequential basis. The company’s official guidance calls for gross margin to range between 9-12% in the third quarter.
From these indications, an estimate for CSIQ’s third quarter earnings has been compiled. As usual with all my estimates, only metrics stated directly or implied through company comments in its earnings report and conference call were used without speculation beyond these parameters. These estimates only represent operational earnings and excludes any unannounced non-operational gains or losses. With only a few trading days left in the third quarter for currency trading, a separate estimate for CSIQ’s net foreign exchange translation has also been calculated based on key currency ratios (EURO = 1.36 USD) as of today’s date.
CSIQ Q3 Earnings Estimate:
- Revenues: $547m
- Shipments: 355MW module, 20MW system kits, 20MW EPC
- Asps: 1.30/watt module
- Blended Unit Cost: 1.17/watt
- Gross Profit: 355MW x .13/watt = $46m module + $12m system = $58m
- Gross Margin: $58m / $547m = 10.6%
- Operating Expenses: $39m
- Net Interest Expense: $10m
- Tax: $2m
- Net Income: $7m
- Diluted Share Count: 43.5m
- EPS: 0.16
Potential Net Foreign Exchange Gain: 15m
Potential Net Foreign EPS Gain: 0.35
Unlike my original Q2 estimate which excluded Canadian Solar’s new systems business, my Q3 estimate includes 20MW of EPC business the company stated in its conference call is likely to be recognized. Revenue and margin profiles for this new segment is based on information disclosed in the second quarter and assume these metrics remain constant in the third quarter. Much like the industry’s extremely dynamic variables regarding the solar module business, exact figures for CSIQ’s new systems kit and EPC businesses may also change meaningfully from quarter to quarter.
By my estimation, these segments contributed almost 8% of the company’s revenues in the second quarter and is likely to increase to over 10% of overall revenues in the third quarter. Gross margin for both these segments were also disclosed to be higher than its tradition solar module business. While still a small part of Canadian Solar’s overall revenues, the non-module aspect of its business should have a meaningful impact in the third quarter. CSIQ expects contribution to further increase with 2012 systems segments contribution around 25% of overall sales. In short, the company’s system businesses should significantly impact forward earnings in a positive sense while differentiating the company from what many regard as a commodity solar module production business.
It is not a huge surprise Canadian Solar is moving towards this strategy of further downstream integration since many U.S. listed direct Chinese peers are also doing the same. In fact, major U.S. based solar companies such as Sunpower (SPWRA), MEMC (WFR), and First Solar (FSLR) started concentrating on downstream systems integration in the past couple of years in order to carve a niche in the industry instead of trying to compete with low cost Chinese competition on component production alone. Without strong systems integration divisions, Sunpower and MEMC might find it extremely challenging in placing solar products at a higher cost structure vs. lower cost large scale Chinese peers such as Trina Solar, Yingli Green Energy, and Suntech Power (STP). Even low cost First Solar may find it more difficult in placing its lower efficiency products vs. higher efficiency crystalline counterparts when costs are very rapidly declining to FSLR’s level on a balance of systems aspect.
While Canadian Solar has had a spotty record of execution in recent years, it has given a general longer time horizon strategy and cost outline which should make it a survivor during a period where so many smaller scale peers are being forced out of the market. While the company still has no plans to fully integrate, through strategic partnerships namely with what will be its main silicon wafer supplier GCL-Poly, CSIQ’s “virtual” integrated model should lead to a very competitive cost structure. With indications of the ability to procure silicon wafers at or below 0.50/watt and the ability to process it into solar modules at or below 0.50/watt, Canadian Solar is implying it should be able to reach 1.00/watt module unit costs or lower by the end of 2011 assuming inventory levels have been fully normalized to current procurement rates.
This cost structure if realized in the coming quarters should make CSIQ extremely competitive within the solar industry. It is quite possible that the per watt gross margin level I estimate Canadian Solar to record in the third quarter may be lower than under normalized conditions. While similar scales of gross margin for the entire industry will be much lower than 2010 record levels, it is still high enough for companies such as Canadian Solar to net positive earnings. With its recent share activity along with its U.S. listed peer group so dramatically discounted, the markets appear to be pricing the sector for failure. That may be a premature and flawed conclusion especially when the industry leaders are still making money while the competition is being squeezed out.