Another one from Credit Suisse this morning:
First Solar (FSLR)
Systems profitability less than expected
Bottom line – systems profitability less than we expected. FSLR hosted a special guidance call for 2012 before market hours today. Company guided 2012 GAAP rev/EPS to $3.70-$4.0bb and $3.75-$4.25/sh, below consensus at $4.09bb/$7.42. Key negative was that system margins are lower – lower pricing, higher costs than expected – as a result our pipeline valuation declines from $28/sh to $19/sh. We are revising our PT to $30 from SOTP of a $19 DCF pipeline valuation and the remainder from the module business and net cash. There were some positives – operating cash flow will improve, below the line items like opex/start up costs/taxes were better than our model, and company further lowered its long term cost metrics. We are revising our 2011 EPS from $6.51 to $4.87; 2012 from $8.13 to $4.08; and 2013 from $7.77 to $5.82.
System price variances vs our model. There had been visible positive data points such as the up to $1.36bb sale of 190MW (AC) AVSR to Exelon and $967mm DOE guarantee for 230MW (AC) Agua Caliente projects that led us to assume ~$3.66/watt system price (DC) in our models for CY12. FSLR has also been vocal in saying its panel and BOS costs are declining – combined, these two will only be ~$1.60/watt including under absorption charges in 2012. FSLR guided to $1.7bb in system revenue and 1.2GW in system completions, well above our model at $1.25bb/866MW. On the cost side, this suggests a “system ASP” (same as cost as system margins are 0) (ex-panel) of $1.65/watt, which was well above our estimate of $1.45/watt (which we thought was reasonably conservative – assuming 87c/watt standard BOS + 25-50c/watt of non-standard costs). FSLR is certainly not assuming a third party channel price less than its cash cost which is probably ~65c/watt, as the company noted a 10% volume reduction in its channel sales would impact EPS by 30c – if panels were sold below cash cost, fewer panel sales would increase profits.
Assuming FSLR sets channel prices for the 740MW at 0% GM (~75c/watt) – the implied system price including panels to get to revenue guidance (excluding Desert Sunlight) was only ~$3.25/watt, well below our $3.66/watt estimate. The margin difference of 60c/watt on systems, combined with slightly lower channel panel price assumptions (we were previously at 81c/watt for channel modules) and combined with the previously disclosed revenue push out on Desert Sunlight (we were modeling 100MW of non-GAAP Desert Sunlight revenues, company guided to 175MW of completions) led to the variance in guided EPS vs our estimate.
Is there potential for FSLR? Company improved its 2015 panel manufacturing cost targets to 50-54c/watt, and increased panel efficiency targets to 14.5-15.0% - both were better than 52-63c/watt and 13.5-14.5% before. Also, 2015 BOS cost targets were lowered to 70-75c/watt vs 91-98c/watt before. Company argued that assuming a 10c/watt penalty for lower CdTe efficiency, and adjusting 5c/watt for freight/warranty/insurance/recycling that FSLR includes in COGS which perhaps some Chinese c-Si companies do not include, c-Si companies need to reach ~57c/watt in total cost/watt to be competitive with FSLR. Right now c-Si companies are guiding to ~60-65c/watt in non-poly cost, and assuming 12c/watt in silicon costs, that would still be 15-20c/watt higher than FSLR’s suggested cost levels. FSLR also believes that it can get its system price (ex-development costs likely ~25c/watt) to ~$1.40-$1.60/watt - these levels are much more competitive for solar PV in general, and would accelerate aspects like converting existing thermal pipelines to PV. Utilities have begun to de-emphasize their focus on solar thermal due to the better competitiveness of PV.
Derivative implications: Negative for poly. While there are aspects to debate on FSLR’s own system profitability, the rest of FSLR’s guidance call and takeaways were negative for the overall solar sector. Company presented a rather sobering analysis that concluded that European subsidized markets are shrinking (cited Spain, Italy, Germany, France, Czech boom/bust bubbles, and also the rapid decline in new solicitations in California, and suggested emerging new markets have more complex tendering processes and local content requirements). Company juxtaposed this shrinking solar demand with significantly increasing supply in the industry. We cannot begin to fathom what this means for the 100s of thousands of tons of poly capacity that is already online and the 10s of thousands of tons of planned additional poly production through 2012. We reiterate our view that poly prices will drop < $20/kg soon, and that Tier 1 poly makers will cut production substantially to bring a semblance of balance to industry. We think it is unlikely that any poly maker will have positive EBIT margins starting in 1Q12. Poly assets may need to correct to a 30-40% discount to book value to reflect excess capacity. FSLR’s c-Si competitors have a more variable cost structure (the average capex for wafer to module conversion is < 25c/watt, vs >$1/watt for FSLR, and ~$1/watt for poly + wafer).
Positive: Several improvements below the line (opex, start up expenses, tax rates).
Guidance for opex/sales of 12% in 2012 and 7-8% in 2014 is below ~17% in 2011. Also the opex in CY12 was guided to only $480mm – below our CY12 pre-call expectations of $543mm. Also, start up expenses were guided to $10-$20mm – this was much lower than the $75mm we were expecting pre-call. Tax rates were guided to 15% also below 20% we were modeling pre-call.
Neutral: Focus on utility scale systems in sustainable markets.
FSLR is expected to announce a new 3 year plan during its next earnings call to detail its plans to compete in “sustainable” markets. The company will emphasize utility scale systems with a levelized cost of energy (LCOE) of 10-14c/kWh and $1.40-1.60/W system prices (excluding 20c of site specific costs). Rooftop and off grid systems will be de-emphasized. FSLR has had early success in India with ~190MW in 2011 and ~120MW in 2012 out of total market of 600MW (CS est), and in China with a 2GW memorandum of understanding for Ordos.
Neutral: 280MW Mesa capacity expansion plans put on hold.
The company had previously announced the postponement of Vietnam capacity and has now postponed the Mesa, AZ facility as well. New 2012 year end capacity of 2.5GW will still only be utilized at ~80% for 2GW of production. Before 3Q11, FSLR had run at full utilizations consistently. Production and capacity plan cuts will help FSLR and other manufacturers to save cash and mitigate ASP declines.
Negative #1. 2012 guidance below expectations.
New revenue guidance of $3700- $4000mm, below CS at $4223.5mm, and cons at $4089.5mm. New EPS guidance of $3.75-$4.25, below CS at $8.13, and cons at $7.42. 2012 guidance for 1.2GW in systems (including 180MW that will not have revenue recognition for Desert Sunlight) was significantly higher than ~400MW in 2011, and our model at 866MW – however, as we note above, due to revenue recognition issues, GAAP EPS guidance was much lower than expected.
Negative #2: Pipeline cannot be replenished at same rate.
Management emphasized that subsidy cuts in Europe and the fulfillment of RPS requirements in California and elsewhere will prevent replenishment of the project pipeline in these markets. California has enough contracted renewable projects to fulfill its 33% RPS by 2020, but someprojects will be cancelled leading to some more project development. FSLR cites CPUC data to estimate 600MW in California RPS solar PPAs (including CSP) in 2011, down 87% from the peak of 4.6GW in 2008. FSLR will focus on sustainable markets with less dependency on subsidies.
Negative #3: Cost reductions stagnant in 2012 due to under absorption.
While FSLR lowered its longer term cost reduction target, its 2012 average module cost of 72c (including 5c of under absorption cost) is stagnant with 2011 costs of 74c. Lower utilizations will lead to higher module costs per Watt for nearly all manufacturers. FSLR is running below full utilizations in 2H11 and 2012 for the first time. The company has guided 80% utilization in 2012.
Negative #3: Company lowered its long term gross margin target to 15-20%, and operating margins to 8-12%.
We were expecting higher longer term operating profit margins closer to 20%. |
