2011 was one of the most challenging years for the photovoltaic ("PV") solar industry, and this was no better represented than by Trina Solar's (NYSE:TSL) performance. TSL had been one of the most consistently growing and profitable companies within the PV sector, but 2011 may be the first year in the company's history as a public company where it may post an annual U.S. GAAP net loss.
The recent and ongoing consolidation cycle in the industry has caused a chain reaction of inventory liquidations by defunct manufacturers, which has led to pricing erosion across the entire solar value chain by over 50% in some cases. As with past cycles of average selling price ("ASP") deterioration, all manufacturers had to struggle with managing inventory levels and blending average costs down towards lower market prices. In most cases, this led to margin compression that would not correct until blended inventory costs matched real time market levels. As with other peers, Trina Solar's third quarter 2011 earnings could not escape the industry's grueling conditions.
TSL's earnings were not a major surprise given that the company had already warned for the third straight quarter. Similar to U.S. listed peer LDK Solar (NYSE:LDK), Trina's quarter was far below original expectations. Shipments declined by 6.6% sequentially to 370MW, well below original guidance of 480-520MW made at the time of its second quarter earnings report. Unlike most large U.S. listed solar peers such as Yingli Green Energy (NYSE:YGE), Suntech Power (NYSE:STP), Canadian Solar (NASDAQ:CSIQ), Jinko Solar (NYSE:JKS), and LDK Solar-- which posted flat to higher sequential module shipments, Trina's volume decline was even more surprising. Perhaps for the first time, TSL had lost relative market share to direct peers as the company struggled to realign its sales away from more risky channels to more sustainable ones.
As a result of not only lower shipments, but dramatically lower ASPs, Trina Solar's revenues declined by 16.8% sequentially to $481.9m. Quarterly declines would have been greater had the company not posted revenues from solar projects sold, which amounted to $22.4m. Gross margin collapsed from 17% in Q2 to 10.8% in Q3, although part of the contraction was due to a $19.1m inventory provision. Excluding the write down, adjusted gross margin would have been 14.8% compared to my revised margin estimates of 10.7% GAAP and 14.8% adjusted, respectively. As noted above, lag time in blending inventory costs lower did not match real time ASP erosion, which resulted in margin compression.
While Trina Solar managed to keep most of its operating costs flat, an additional $10.3m accounts receivable provision further deepened operating losses. U.S. GAAP loss for the third quarter came to $31.5m, or -0.45 in earnings per share ("EPS"). Excluding the receivables charge and foreign exchange translations, TSL's operational loss for Q3 would have totaled $21.6m, which compares to my $25m adjusted figure. Similar to CSIQ, had Trina not altered its currency hedging strategy during the period, it could have posted significant net foreign exchange gains which may have reversed much of its first half foreign exchange losses totaling $34.9m. The company's large cumulative net foreign exchange losses in the first three quarters of 2011 was as unusual and uncharacteristic of Trina Solar as its third consecutive earnings warning.
In perhaps a perfect storm situation for Trina Solar, as well as for the solar industry, the company did mark significant milestones. Although, to a high degree, related to lower procurement costs across the entire value chain than the company's own initiatives, module processing costs declined to 0.65/watt from 0.73/watt on a sequential basis. Coupled with much lower polysilicon costs, Trina Solar's module unit cost would be among the lowest in the industry as it targets a sub-0.80/watt goal in the coming quarters once blended inventory costs normalize to real time procurement levels. Similar to LDK Solar's cost achievements, industry leading cost metrics should result in margin expansion after the industry's consolidation matures and inventory liquidations abate.
For the fourth quarter of 2011, margins at TSL should contract further. Due to a high concentration of long term contracted polysilicon agreements, silicon procurement costs may lag real time spot market pricing. To an extent, this was evident in the company's blended silicon costs of approximately $62/kg during the third quarter, when most major peers quoted ranges of $50-55/kg. For the fourth quarter, and as noted in its earnings conference call, Trina indicated a blended cost range of $48-50/kg-- which was well above spot market levels of around $30/kg. Perhaps management's comments portrayed more conservatism given its prior earnings warnings, but at face value TSL's costs will not decline to equal magnitudes to ASP erosion noted across the industry. As a result, the company guided Q4 margins to be approximately 10%.
Trina Solar's fourth quarter earnings should also be heavily impacted by another sequential unit shipment decline forecast which calls for a range of 320-350MW vs. 370MW reported in the third quarter. Combined with an ASP forecast decline above 10%, TSL should post another U.S. GAAP loss for the fourth quarter. As usual, an estimate has been compiled solely on information provided by the company, and does not use metrics outside ranges stated directly or implied by management. The estimates below mostly represent operational figures which excluded unannounced gains or charges.
TSL Q4 Earnings Estimate:
- Revenues: $380m
- Shipments: 335MW
- ASPs: $1.10/watt
- Blended Unit Costs: $0.98/watt
- Gross Profit: 335 x 0.12 = $40m + $2m system = $42m
- Gross Margin: $42m / $380 = 11.1%
- Operating Costs: $64m
- Net Interest Expense: $9m
- Tax: $0m
- Net Loss: -$31m
- Diluted Share Count: 70.5m
- EPS: -0.44
Because TSL may increasingly report non-module revenues such as system project sales, actual revenues may differ. The company booked $22.4m or almost 5% of its third quarter revenues in unannounced solar project sales. In its Q3 conference call, management noted further systems revenue was expected in the fourth quarter but without proper guidance, only rough estimates can be made. The revenue figure above assumes moderate systems revenues half in magnitude compared to the prior quarter with gross margins inline with management's comments.
Additionally, foreign exchange translations were assumed to be marginal for the fourth quarter since TSL reported very minor net currency exchange translations in Q3, when large currency moves-- namely between the euro and the usd-- were recorded. With key currency moves much smaller in magnitude in the fourth quarter, and assuming the company kept a constant hedging practice, net currency effects should be even smaller and thus excluded from the estimates above.
The fourth quarter and last quarter of the fiscal year often marks house cleaning for a company's books. As a result, additional charges may occur. Although management did not expect another inventory provision, another albeit smaller inventory charge could still result in the fourth quarter. In addition, overdue accounts may also be written off in further receivables charges similar to the $10.3m provision taken in the third quarter. The dramatic collapse in polysilicon pricing at the start of the quarter may also cause TSL to write down prepayments made for some long term contracts at unfavorable pricing. Although management stated most long term contracts had successfully been renegotiated lower, the industry did witness prepayment forfeiture to large incumbent polysilicon producers during the past couple of quarters.
There may be some potential positives however. As noted in my previous TSL review, a tax reversal for overpaid taxes could occur if Trina was granted renewal of its high technology tax status. Additionally, no taxes were assumed in the estimates above, but the company may record a tax benefit due to operating losses similar to the third quarter. Other reversals such as the $10.3m accounts receivables provision taken in the third quarter, which exclude insurance reimbursement potential. In general, although Trina Solar may book some non-operating gains in the fourth quarter, it would still be unlikely such items would be large enough to reverse an expected U.S. GAAP loss for the quarter.
Despite likely fourth quarter losses, Trina Solar's continued improvements in cost structure place the company in a very favorable strategic position within the industry. Once industry conditions normalize, the lowest cost producers should experience the highest margins, as TSL has shown in the past. The 0.12/watt module gross margin I estimate for TSL in the fourth quarter should be its trough low and if the company can achieve its target of below 0.80/watt module unit cost in the coming quarters, per watt gross margin could potentially double from trough lows. While far from the 0.50-0.60/watt margins experienced in past booms, even a per watt gross margin of 0.25/watt for solar modules would be more than enough to return Trina to healthy profitability.
As noted in my recent CSIQ and LDK reviews, the reason why per watt gross margin may bottom at this normalized level is because it would likely be the minimal level for less integrated or cost effective peers which make up the bulk of the industry simply to break even at an operational level. Perhaps eventually, as Trina's CEO predicted early last year, the industry may continue to consolidate until only a dozen extremely large scale highly integrated companies remained. At such a point, margins would be less pressured by peers but more reliant on a company's cost advantages relative to demand reflective of price elasticity in the industry. Given how Wall Street has lowered valuations for the entire industry to levels as low as 1-2x earnings of past cycles, perhaps any reversal to a more normalized and sustainable environment for the sector has already been fully discounted.
Additional disclosure: No position in STP, CSIQ.