Enphase: Anemic Revenue Growth with Alarming Warranty Obligation Growth
I believe that Enphase (ENPH) stock is currently overvalued due to a number of poorly-understood issues. First, the company is generating little to no revenue growth despite rapid growth in its primary end market. Second, the high failure rate of its second-generation inverters appears to be causing the company to lose market share as well as incur significantly higher warranty expenses than expected when it originally sold the products. Third, the company's reported gross margins and operating income are benefiting significantly from a curious reduction in warranty accruals for its third-generation product. I struggle to understand how Enphase can reduce estimated warranty expense for the third-generation product-which carries a 25-year warranty vs. the second generation's 15-year warranty-after its experience of significantly underestimating the warranty expense for its second-generation products. With a warranty expense accrual similar to its second-generation product, Enphase would be generating lower GMs and higher operating losses, calling its valuation at a premium to the ABB purchase of Power-One into question given its anemic growth rate and quality issues.
Enphase is the largest provider of microinverters in the world. Microinverters convert DC power into usable AC power. Micros are primarily used in residential solar installations and compete with string inverters. Micros are significantly more expensive to install, but claim to provide better yield vs. string inverters. I could spend pages laying out the argument for micros vs. string vs. newer technologies like DC optimizers, but that is for another day. Instead, I will focus on a few alarming trends emerging in ENPH's financial statements after examining their latest 10-Q.
First, ENPH is struggling to grow revenue despite rapid growth in its largest market this year. At the recent solar industry conference in Chicago, the President of SIA stated that U.S. residential solar installations are expected to grow 59% in 2013. ENPH claims to be the leader in this market and it comprises the majority of the company's revenue; however, its revenue is expected to grow only 6% this year. The company did see a benefit from increased buying in Q1 and Q2 of 2012 due to the expiration of the 1603 federal tax credit making those quarters a tougher comparison, but that doesn't explain why actual unit sales were down Y-o-Y in Q3. From the 10-Q:
We sold 426,000 microinverter units in the three months ended September 30, 2013, compared to 431,000 units in the same period in 2012. The slight increase in net revenues was attributed to a change in revenue mix between microinverter revenue and microinverter accessories revenue. The increase in revenue from a larger mix of microinverter accessories sales was partially offset by a decline in the average selling price of our microinverters.
The company was able to grow revenue slightly in Q3 due to higher accessory sales offsetting a decline of inverter sales unit volume and a lower average selling price. Lower unit sales in the face of a rapidly growing market led me to believe the company must be losing market share either to the flood of other microinverter products that have come to market or to string inverters.
Second, the company's warranty obligation has nearly doubled in the last year and the failure rate of the second-generation product has been significantly higher than expected. The following table was provided in the 10-Q:
WARRANTY OBLIGATIONS (in thousands)
Three Months Ended
Nine Months Ended
Balance at beginning of period
Accruals for warranties issued during the period
Changes in estimates
Balance, at end of period
From the table we can see that in the last 12 months the total warranty obligation has increased from $15.5 million to $29.4 million. A few items of note from the table: accruals for warranties issued during the quarter were down 45% from a year ago on a 1% decrease in unit sales. This must mean that the company expects much lower failure rates of third- and fourth-generation products versus second-generation inverters, which is interesting given that the third- and fourth-generation inverters have a 25 year warranty versus a 15 year warranty for second-generation inverters. We can also see that settlements have increased 97% versus last year, so the company is definitely seeing more failures.
The commentary in the Warranty Obligations section provides even more insight into what is happening. From the 10-Q:
During the three and nine months ended September 30, 2013, the company experienced actual failures of its second-generation microinverters that exceeded its then current failure rate estimate. Based on continuing analysis of field performance data and diagnostic root-cause failure analysis performed on returned units, the company concluded that it was necessary to increase the estimated failure rates for its second-generation product and recorded additional warranty expense of $7.3 million and $15.9 million for the three and nine months ended September 30, 2013, respectively. In addition, the company recorded a reduction to warranty expense of $3.1 million for the three and nine months ended September 30, 2013 related to a decrease in the expected failure rate of the company's third-generation product. During the third quarter of 2013, the company concluded that there was sufficient historical data of actual field performance of previously sold third-generation products to support a lower estimated failure rate.
ENPH is clearly having an issue with its second-generation inverter and has been continuously increasing its obligation estimate since last year, but it appears only to be getting worse. Commentary about 2012 from the 10-Q:
During the three and nine months ended September 30, 2012, the company experienced actual failures of its second-generation microinverters that exceeded its then current failure rate estimate. Based on continuing analysis of field performance data and diagnostic root-cause failure analysis performed on returned units, the company concluded that it was necessary to increase the estimated failure rates for its second-generation product and recorded additional warranty expense of $4.8 million and $4.4 million for the three and nine months ended September 30, 2012, respectively.
One of the most surprising numbers was the benefit of $3.1 million ENPH recognized during the most recent quarter by lowering the expected obligation for third-generation inverters. The oldest third-generation products have been in the market about two years and have a 25 year warranty versus a 15 year warranty on second-generation inverters. I understand it must be extremely difficult to forecast an obligation for a 25-year warranty, but I would expect the company to be ultra conservative given that they have horribly underestimated the obligation on the second-generation inverter. This just doesn't appear to be the case with the lower warranties-issued-during-the-period expense in Q3 as well as the benefit taken during the quarter. ENPH's gross margins would have been 23% for the third quarter instead of 28% if the company did not recognize the benefit. Because the gross margin is heavily influenced by estimated future failure rates, those failures and reserves need to be very carefully scrutinized.
The increasing failure rate of second-generation inverters may also explain some of the market share loss in the U.S. I have had conversations with two large residential solar installers in the U.S. who are moving away from the ENPH products due to higher maintenance costs over the warranty period. Installers typically warranty a new system for at least 5 years. During this time they are responsible for repairs. When an ENPH inverter fails, ENPH mails a new inverter free of charge, but the installer incurs the cost of sending a technician out to actually replace the inverter on the customer's rooftop. This cost is several times the cost of the inverter, making failures a significant expense to the installer despite the warranty. These installers are also the primary interface with homeowners. If they de-emphasize the ENPH microinverter it will eventually cause ENPH to lose market share
Lastly, looking at ENPH's balance sheet, the overall warranty obligation has become significant. Recall the current portion of the obligation, which is $8.4 million, is in current accrued liabilities. Long-term warranty obligation is $20.9 million for a total of $29.3 million. Total liabilities are nearly $80 million, making warranty obligation almost 37% of the liabilities. If we treat the obligation as debt, the company is now in a net debt position of $6.9 million. The growing liability is not detrimental to cash flow now, but it will be at some point. The company has yet to turn a profit and is expected to be slightly profitable next year if they grow revenue 25% after only 6% growth this year. Assuming an average selling price decline of 10%, the company will have to grow unit sales by 39% after almost no growth this year. I believe this is highly unlikely in an increasingly competitive market.
The trend of lower sales growth and increasing warranty obligation growth at ENPH is troubling to say the least. The company is clearly having higher-than-expected failure rates with older models in the field. This issue becomes especially troubling considering they increased the warranty by 10 years on newer models. It also appears to me that the third-generation model had been introduced before the second generation failure rates really increased, so the issue, or issues, may not have been apparent and addressed in the newer model. I am happy to see the significant increase in disclosure about the warranty obligation in the most recent 10-Q, which will help us monitor the quarterly impact on gross margin. This is important given the larger role warranty appears to be playing in more recent quarters.
The warranty issue is likely to give pause to potential acquirers as well, but looking at the valuation of the most recent deal in the space this does not appear to be an attractive option anyway. Power-One (the second largest inverter company in the world) was acquired by ABB on July 25th of this year. Giving ENPH the same trailing-twelve-months revenue multiple as Power-One's takeout we get to a stock price of $3.78 (also treating the warranty obligation as debt which has a $0.70 impact on the valuation). While using revenue multiples is not ideal, ENPH's lack of profitability make it necessary. I would note that Power-One generated over $100M in operating income in 2012. I would argue that with the potential problems a 15-25 year warranty could entail, ENPH should trade at a discount to the Power-One takeout multiple. A multiple of 0.5x EV/sales yields a price of $2.48, which I believe is more than fair for a low growth, low margin, money losing business. If the trend of increasing failure rates continues it could ultimately prove fatal for the company as a loss of customer confidence could lead to stalled revenue growth (like we have seen this year). Without significant revenue growth the company will struggle to ever reach profitability and eventually customers will have to wonder how valuable is a 25-year warranty issued by a company that never makes money.
Disclosure: The fund I manage has a short position in ENPH.