|Shares Out. (in M):||131||P/E||0||0|
|Market Cap (in $M):||4,133||P/FCF||0||0|
|Net Debt (in $M):||670||EBIT||0||0|
Sign up for free guest access to view investment idea with a 45 days delay.
While we regard the solar sector as broadly undervalued, we believe TerraForm Power (TERP-$31.64) provides an attractive risk-reward for those that don't want to own an upstream panel maker or downstream project developer, and prefer a company with stable, high margin, contracted cash flows. For those concerned about the risk of a meaningful step-up in interest rates near term, this idea may not be suitable for you. However, this risk can be hedged out using an MLP ETF. We see 15%+ near-term upside and +30-50% upside potential within the year
The current share price reflects recent equity issuances required to purchase assets from First Wind but not the associated cash flows.
We believe the upcoming earnings announcement this Wednesday, February, 18th could serve as a near-term catalyst and expect the dividend to be raised 11% from $1.08 to an annualized run-rate of $1.20. TERP’s current indicated yield of 3.4% bumps to 3.8% after this expected dividend increase. This compares favorably to comp NRG Yield (NYLD, $51.72), which currently yields 3.0%.
TerraForm has averaged a 3.17% annualized declared dividend yield since it went public and has only once ever briefly traded over a 3.8% yield. If TERP shares were to remain at current levels after the dividend increase, the dividend yield (of 3.8%) would represent a +2 standard deviation divergence from the long-term average. The historical peer average dividend yield is close to this average at 3.25%.
At a projected run-rate indicated dividend of $1.20 and a 3.25% yield, we think the stock could be valued at close to $37 today, or 17% higher.
Based on management’s conservative guidance, which doesn’t include any incremental M&A or the full $75m of cash available for distribution (CAFD) required to be dropped down by SunEdison as part of its IPO Support Agreement, TERP’s run-rate dividend by the end of 2015 would be at least $1.33.
Moreover, our base case forecast, which we detail below, calls for an annualized dividend run-rate of $1.57 by the end of 2015. Applying today’s indicated yield of 3.4% to this projected run-rate would yield a price target of $46 by year-end.
TerraForm Power (ticker: TERP) owns and operates solar and wind electricity generating power plants in OECD countries (primarily the US, Canada, UK, and Chile). Pro-forma for the projects acquired in the recent First Wind acquisition, TerraForm owns 409 projects totaling 1.5GW of capacity with a 16-year weighted average contract life. The split between solar and wind is about 67%/33% and the utility/distributed generation split is 73/27%, respectively.
Solar and wind plants require very little OpEx or Capex, have long-term contractual off-take agreements with high quality credits, and thus produce very stable, predictable, high-margin cash flows. Gross and EBITDA margins for the nine months ending September 2014 were 88% and 85%, respectively.
TerraForm was formed as a "YieldCo" and IPO'd in July 2014 at $25 (reaching over $34 on its first day of trading) by its sponsor and 50% majority owner (65% at the time of IPO) SunEdison, Inc. (SUNE, $21.45). SUNE formed TERP for the purpose of being its captive off-take vehicle for renewable power plants in OECD countries (a soon-to-be announced EM YieldCo will be the off-take vehicle for emerging markets projects). As an aside, we regard the term "YieldCo" as an inappropriate moniker for the company at this stage of its lifecycle and believe that TerraForm can be more viewed as a "Total ReturnCo" or, as referenced in TERP's S-1 and recent corporate presentation, a "dividend-oriented growth company". Over the next few years, we expect the vast majority of the stock's returns to come in the form of capital appreciation rather than yield.
TERP does not develop the projects it owns. It buys completed, inter-connected, and operational projects from SunEdison (drop-downs) or other third-party developers using a cost of capital arbitrage afforded to it by being a dividend paying publicly traded vehicle in a yield starved world. We estimate its current cost of capital arbitrage to be over 10 multiple points of CAFD, and we believe that its cost of capital arbitrage will stay sufficiently high for the foreseeable future to allow for meaningful dividend growth and shareholder value creation.
The SunEdison Relationship
SunEdison is one of the largest renewable energy developers in the world (solar, and more recently wind, after its acquisition of First Wind). It is responsible for all aspects of developing power plants - from sourcing to permitting, financing, and construction. It is also responsible for the operations and maintenance (O&M) once the plants are up and running (for a small fee from TERP and other project owners). In order to scale its development business quickly, SunEdison uses a capital-light model. It primarily outsources the construction of the plants, while using its sourcing, permitting, and financing expertise to structure and win deals. Because of its capital-light approach and use of third-parties in various parts of its development process, SunEdison has a large network of resources from which to source new projects. SUNE’s solar project development business has grown organically at a 78%+ CAGR from 2010 to 2014E and is expected to compound at 44% from 2014 to 2017; this excludes additional growth from First Wind.
As part of the IPO Support Agreement, SunEdison is required to offer TERP drop-down projects with $175m in annualized CAFD by the end of 2016 ($75m in 2015 and $100m in 2016). Additionally, TERP has a six-year Right of First Offer (ROFO) on all eligible projects developed by SunEdison. As the sponsor, SUNE retained Incentive Distribution Rights (IDRs), permitting the company to receive an incrementally greater percentage of cash flows above certain levels, relative to common shareholders. SunEdison is not permitted to sell its IDR's until it has fulfilled its $175m CAFD commitment and must hold 25% of its Class B common shares until either three years after the IPO or distributions are above the top IDR threshold. The IDR splits are as follows:
100% of dividends up to $1.35 per share go to common shareholders.
85/15% split to common shareholders above $1.35 until common shareholders receive $1.58.
75/25% split to common shareholders from $1.58 until common shareholders receive $1.8056.
50/50% split for all incremental dividends above $1.8056 per share.
The actual math associated with the IDR waterfall implies that total dividends must reach $1.89 per share in order for common shareholders to receive $1.8056 per share and SUNE to hit the 50/50 high-split.
We believe SunEdison's creation of TerraForm Power was a key inflection point for the company. As a captive off-take vehicle, TerraForm has provided two key benefits to SUNE; lowering SUNE’s cost of capital and allowing for the retention of incremental value in each project it develops (SunEdison was previously written up by aviclara181 in Dec 2013 and cfavenger in Nov 2014). The success of SUNE is almost directly tied to the success of TERP (and SUNE’s future Emerging Market YieldCo). SunEdison will be supporting TERP and doing everything it can to help TERP trade well in order to maximize value creation for both companies. This includes dropping down projects at opportune times (via the use of its newly formed $1.5 billion intermediary warehouse facility) and at valuations beneficial to both entities. SUNE is also incented to assist TERP in getting to the IDR high-splits as soon as possible, which means helping TERP grow as rapidly as possible, including accretive third-party M&A. We don't believe this aspect of the story is being fully appreciated by the market today.
Having a reliable, low cost of capital off-take vehicle for the projects it develops creates a potential virtuous cycle for SUNE. It enables SUNE to sell its projects for a predictable price, which in turn provides more capital to develop more projects, which in turn creates scale advantages that lower the cost of solar systems, and ultimately results in a greater addressable market.
In summary, to be long SUNE, one must implicitly be long TERP. We believe the investment thesis for TERP may actually be somewhat easier than SUNE simply due to the difference in valuation complexity. It is much easier to put a value on a stock based on dividend yield than to deconsolidate financial statements and perform a complex sum of the parts, as required for SUNE. Having said that, we are constructive on both equities.
2015 Outlook; Expect Significant Increase in Dividend per Share:
TERP's current indicated and declared dividend represents an annualized $1.08 or 3.41% yield. On February 18th, we estimate TERP will raise the dividend by ~$0.03 quarterly or $0.12 annually. Notably, the current share count and capital structure reflect the financing related to First Wind (deal closed on Jan 29th), but the resultant cash flows are not yet captured in the indicated yield. Based on management guidance, First Wind is expected to add an incremental $0.22 of CAFD annually. Adding $0.03 ($0.12 annually) to the dividend this quarter should be very manageable. We think TERP could probably raise the dividend by $0.055, or the full $0.22 run-rate, but management would prefer to gradually raise the dividend to show continual growth.
As elementary as it seems, we think once the next dividend is declared (in conjunction with earnings on Feb 18th) and the indicated yield moves closer to 3.8%, the gap to peers and TERPs' historical indicated yield will be too large (+2 standard deviations), equity screens will pick this up, and the stock will begin to appreciate. This appreciation will likely occur gradually.
Given the growth profile of the business (19% dividend growth forecasted from 2015 to 2019), we think a 3.8% indicated yield is probably too high, as it implies a ~23% required annualized return. TerraForm comps favorably on this metric against its YieldCo peers. The chart on page 14 of the January 14th, 2015 presentation sums up the gap nicely.
Looking at the company's brief history, TERP has on only one occasion traded above a 3.8% indicated yield. After the dividend is increased, the stock is expected to gradually revert to its average indicated yield and something more closely approximating that of its closest peers.
Raising the dividend to $1.20 would suggest a stock price close to $37, based on a 3.25% targeted yield. Management has guided to $1.30 in total dividends for 2015 (declared in 2015). If the first quarter dividend is $0.30, the last 3 quarters will need to average $0.333, or an annualized $1.33. At a 3.25% yield and indicated dividend of $1.33 at the end of 4Q14, the stock would be worth $41.
We think management's 2015 guidance for dividend per share is quite conservative and see upside to $1.40 on a reported basis. The company's 2015 dividend guidance does not include the full $75m of CAFD commitment from SunEdison (only about $45m is included). Under our base case forecast (detailed in the Bottom's Up Analysis section below), we think the year-end 4Q15 run-rate dividend could end up being $1.57 ($0.39 quarterly). At a 3.25% to 3.50% targeted yield range, the stock would be valued at $45 to $48 per share (representing 42 - 52% upside).
Comparable valuations suggest material upside potential for TerraForm as dividend ramps
The table and chart below highlight the current and historical indicated yields for the comp group of YieldCo's and MLPs. Among the YieldCo's (excluding RNW.TO which is not forecast to grow), only PEGI has a higher indicated yield right now. If TERP's dividend is raised to $1.20 as we expect, at the current $31.64, the stock would be trading at a 3.8% yield, well above the YieldCo group average. If our forecasts are correct, and TERP achieves a dividend run-rate of $1.57 (45% above current levels) per share exiting 2015, the share price should respond accordingly and relative valuation would suggest attractive upside from current levels.
Valuation discussion and long-term growth
Over time, we believe TerraForm will trade primarily on its dividend yield (a combination of both the indicated and the forward run-rate yield). The crux of the circular valuation question for the company is what is the appropriate yield/cap rate for the stock because this valuation not only determines the current share price, but also impacts how much value TERP can create over time.
The current dividend yield and overall cost of equity is dependent on expected future growth. Unfortunately, since TERP plans to dividend out 85% of its cash available for distribution (CAFD), dividend growth needs to be financed by the capital markets, which creates this circular story. Therefore, to wag the dog and for the TERP model to work (or any yield vehicle - MLP's, REITs, etc), clear and transparent forward guidance and continued execution is key, since the sensitivity to the current and future value of the equity relative to a small change in the yield is large.
We think TerraForm and SunEdison have done a good job guiding the market to the underlying growth inherent in the business model and the massive opportunity to come (solar is still less than 1% of growing global electricity capacity). As noted previously, we think the solar industry is on the cusp of significant growth, as the cost of solar is now lower or on par with traditional energy sources in many markets. In fact, Lazard recently estimated that by 2017, utility scale solar would be competitive with both gas and coal in the US, and Deutsche Bank said that up to 80% of the world would be at grid parity by the end of 2017.
Because of the misguided fear that lower oil prices will crimp renewable energy development and a lack of history in executing on promised growth, the stock, in our opinion, has not yet traded in line with its underlying growth prospects (19% dividend CAGR from 2015 to 2019). Other comparable YieldCo’s with lower forecasted long-term dividend growth targets command a better valuation because of the perceived strength of their sponsor.
The large acquisition of First Wind, subsequent equity raise, and evolving capital structure has also contributed to an overhang on the equity. The current capital structure reflects recent equity and debt issuances for the First Wind projects, but the indicated dividend yield does not. We expect this to begin to change with TERP’s earnings report this Wednesday, February 18th.
Oil accounts for less than 1% of current electricity generation in the US, and 0% of incremental new capacity. The only markets where oil is close to relevant in determining the cost of electricity is in Hawaii and Chile. Even if lower oil prices contribute to lower natural gas prices, a large percentage of the retail cost of electricity in the US is from transmission/distribution and grid costs. The fundamental correlation between oil and solar demand is misguided, and we don't expect oil to have any material impact on solar and wind development. Upcoming earnings and conference calls will likely provide further clarity in this regard. Moreover, SunEdison and TerraForm will be holding an analyst day in New York on February 24th, which should serve as an excellent opportunity for investors to get further educated on both stories.
Visibility in the business is excellent. If global economic growth were to come under pressure, we believe that TERP’s growth would continue unabated, based on SUNE’s sizable contracted backlog of projects.
A lack of history shouldn't be confused with poor execution. We believe TerraForm and SunEdison have executed very well in the short time that TERP has been a public company. At the time of the IPO, the initial dividend was set at $0.90, and the company targeted a 15% CAGR over the first three years. Since then, it has acquired over 100MW of DG plants at attractive returns, diversified into wind and expanded its addressable market with the purchase of First Wind projects, accelerated the timing of some SunEdison drop-downs, raised the actual dividend by 20% to $1.08, and raised its forecasted 5-year dividend CAGR from the time of the IPO to 24% (19% CAGR from 2015-2019). Management's current dividend guidance is as follows (actual amount paid, not the run-rate):
What is the right yield?
The short answer is we aren't sure, but given the discount to comps and the expected growth in the coming years (even raising capital at current or higher cap rates), it probably shouldn't be where it is in the near term, given the current juncture of the business (i.e. robust expected growth). Looking at it another way, management is forecasting dividend growth of 19% annually from 2015 to 2019. Adding in the $1.30 in actual dividends (4.0% yield) expected to be paid in 2015 implies a 23%+ required rate of return. At that cost of equity, the market is implying that these assets are either very risky or that the growth profile of the business is much lower than the company is projecting. And it means if the company can hit its growth targets, the minimum return with no yield compression would be 23%. That is a pretty compelling risk-reward profile.
We believe the assets are not much more risky than traditional power assets. Both the sun and the wind have infinite lives, and the economics of solar and wind plants have been pretty well established and tested to this point. The technology is only getting better which means efficiencies on future projects will improve, degradation rates will improve, costs will be lower, etc. Counterparty risk is mitigated by an average S&P credit rating of A- for its off-take partners, and all contracts have long lives (average across the portfolio is 16 years). As the portfolio adds more DG projects, concentration risk, which is low to begin with, should be lowered further.
One of the biggest questions is what the long-term dividend growth rate past the 2015 to 2019 hyper-growth phase will be. If we assume management's dividend guidance through 2019 is achievable, and discount those dividend payments back at the required rate of return implied by the equity price today, we can back into the various computations for the long-term dividend yield and growth rate that the market is implying today.
Our takeaway is that even at a normalized long-term dividend yield of 5.0%, the market is requiring 13.0% long-term dividend growth. That means a 18.0% cost of equity for a $20bn+ market cap company generating 75%+ Ebitda margins at that point in time. We think that is too high.
One of the bear cases for the long-term growth profile of the dividend is the ability to grow once the IDR’s hit the high-splits. What most people forget is that SunEdison has the ability to reset the IDR’s in the future if they felt the dividend could not grow fast enough, or if the market was pricing in lower growth.
Back of the envelope analysis:
We estimate current run-rate CAFD pro-forma for First Wind and recent drop-downs is about $195m. SUNE is required to drop down a total of $175m of run-rate CAFD by the end of 2016 ($75m in 2015 and $100m in 2016). That means that without any M&A or additional drop-downs beyond that required, total run-rate CAFD by the end of 2016 will be $370m.
Depending on the type of project (utility scale versus distributed generation) and the leverage assumed, the CAFD per Watt conservatively ranges from $0.10 to $0.12. To put that in perspective, the existing portfolio has an overall CAFD per Watt of $0.14 (CAFD of $214m for 1,507MW, see slide 8 in the January 14th, 2015 presentation). Third-party acquisitions, including First Wind, averaged $0.105 of levered CAFD per Watt, and TERP management has said that unlevered CAFD per watt is around $0.17. At a 65%/35% targeted debt to equity split, the levered CAFD at a 6% cost of debt comes out to about $0.13/W.
Using our conservative $0.10 to $0.12/W range, SunEdison will need to develop and drop down anywhere from 1.46 to 1.75GW of eligible (OECD countries) projects to fulfill the $175m CAFD commitment. SUNE has already identified and added over 3.3GW of projects to the Call Right List.
Taking a look at SunEdison's backlog and pipeline as of 9/30/14 (pro-forma First Wind), and it is pretty clear that SunEdison has more than enough projects to fulfill the $175m CAFD commitment. In fact, just the backlogged projects will fulfill more than the $175m drop-down commitment, even if only 50% of EMEA and Latam were eligible to be dropped into TERP. SunEdison defines backlog as those projects that will be completed within two years AND have signed PPA's or feed-in-tariffs. Basically, these are a done deal with little risk of being cancelled. TERP and SUNE assume a 90% conversion ratio for backlog projects.
The probability weighted pipeline (60% conversion ratio) could add another $115m to $170m+ of CAFD at our conservative CAFD per Watt estimates. Qualified leads (40% conversion rate) and leads (10%) are gravy on top. Additionally, DG projects are not included in backlog and pipeline. By their nature, DG (commercial and residential) projects have shorter sales cycles and build times, so often do not even flow through the disclosed backlog and pipeline. As the fastest growing segment of the market, DG provides an added layer of comfort over the viability of the SunEdison drop-down commitment and the future growth of the business in general. TERP has noted in recent presentations that they have 10.7GW of expected power plant drop-downs from SUNE, based on 70% of the above backlog, pipeline, and leads located in TERP markets. In addition, they have noted that there is an estimated $221m of unlevered CAFD from the First Wind pipeline alone, and in their November presentation (slide 52) they show $804m of total backlog and pipeline CAFD.
TERP's stated required cash on cash equity rate of return on projects it purchases is 8-10%. At the mid-point, TERP is buying the projects for 11.1x CAFD and would thus need $1.94bn of equity capital. At an implied dividend yield of 4.0% (conservative) and 85% payout ratio, it can raise that capital in the public markets at a valuation of 21x CAFD, resulting in a 10.0 multiple point cost of capital arbitrage. At the current share price of $31.64, TERP would need to issue 26.3m shares for the 2015 CAFD commitment of $75m. Because of the accretion, the stock would then be worth $36.56 at the same 4.0% dividend yield. Again assuming a 4.0% yield, the next $100m of CAFD would be financed at $36.56, adding another 30.4m shares, bringing the year-end 2016 base of shares outstanding to ~187.3m. Total CAFD of $370m at an 85% payout ratio would mean total dividends per share of $1.68. Running that through the IDR waterfall and the common dividend per share run-rate would be $1.66, which is in between the actual 2016 dividend guidance of $1.53 and 2017 guidance of $1.90. At a 4.0% dividend yield, the stock would trade at $41.50, or 31% upside, plus dividends along the way.
Keep in mind that was using a 4.0% dividend yield, issuing equity at the share price prior to the accretive CAFD being purchased (equity capital should be raised based on expected accretion). At a 3.5% assumed dividend yield on the $1.30 in actual dividends paid in 2015 (not the run-rate), and using a starting share price of $37, the dividend per share run-rate would move to $1.72 at the end of 2016 with an ending share price of $49. As projects are gradually dropped down over time, accreting value to the company, the share price should gradually move higher, so the shares issued should be less than that calculated. In addition, that assumes all equity. TERP currently has a $550m corporate revolver that is undrawn, and $130m of unrestricted cash. It can lever up above its targeted 3.0 to 3.5x net corporate debt to CFADS target (5.0 to 5.5x total consolidated debt) to opportunistically price its equity raises.
Additionally, none of the above includes third-party M&A, or incremental drop-downs beyond the required $175m of CAFD. The only aggressive part of this analysis is not taking into effect the IDR impact on the cost of equity raised. At the high split (50/50), incremental equity is theoretically being raised at 2x the implied dividend yield. Dilution from the equity raised makes it less than 2x, and even so, the dividend guidance and growth rate assumed by TERP, at least through the end of 2016, seem pretty safe to us.
The back-of the envelope analysis above provides a high level framework for thinking about the business. We have a built a detailed iterative model driven by four primary variables 1) the rate of project drop-downs and acquisitions over time as measured in Megawatts, 2) the levered CAFD per Watt generated from those projects, 3) the purchase price of those projects, driven by the levered cash on cash equity return, and 4) the implied cap rate (dividend yield) at which the projects were financed at using an 85% payout ratio.
End of 2015 = $44.00
End of 2016 = $54.00
End of 2017 = $58.00
In our base case scenario we begin with a conservative starting share price of $35 under the assumption the stock should trade between a 3.25% and 3.50% when the dividend is raised to an annualized $1.20. The starting price plays a huge role in future value due to the iterative nature of the financing model.
Our cap rate used to finance incremental CAFD is based on a 3.50% dividend yield in 2015, with the rate increasing 25bps each year through 2019.
We have built in the impact of the IDR's taking a greater percentage of the cash flows, ie, the effective cap rate will be higher as the common shareholders will need to pay the same price for less cash flow to them.
We have built into the model a step-down in the share price each time the required dividend yield is increased from one period to the next. That step-down impacts the price at which capital is raised from that point on.
We don’t expect any new equity issuances in 2015 prior to 3Q once the S-3 is effective.
Total drop-downs of 2.0 to 2.5GW from 2016 onwards represents about 60-70% of SunEdison’s recent 2016 guidance of 2.8 to 3.0GW developed, and 2017 guidance of 3.8 to 4.0GW.
Not included in this analysis is large scale M&A. We assume a few hundred megawatts of individual projects each year, but expect TERP and SUNE to do at least one large scale acquisition that includes both projects and incremental pipeline, similar to First Wind.
End of 2015 = $46.75
End of 2016 = $58.00
End of 2017 = $62.75
In our bull case, we assume a starting share price of $37, under the assumption the stock trades at a 3.25% dividend yield on our $1.20 run-rate.
Our cap rate in 2015 is premised on a 3.25% dividend yield, and we have rates rising 25bps per year thereafter.
Note that SunEdison's forecasts used an implied 3.0% yield. Our bull case yields a far more conservative valuation.
End of 2015 = $39.00
End of 2016 = $47.00
End of 2017 = $50.00
In our bear case, we assume a $31.64 starting price (current share price), which would mean either TERP doesn't raise the dividend to $1.20 as we expect, or the stock trades at a 4.1% yield after they do - well above its historical average.
Dragging that implied valuation forward, we use a cap rate premised on a 4.0% dividend yield. We then increase rates 25bps a year through 2019.
Even in the bear case, the cost of capital arbitrage and opportunity set is so large that TERP can still create significant value and generate attractive common shareholder returns.
Even at a 5.0% yield in the high-split IDR scenario, TERP can still create value.
TerraForm's management is technically independent of SunEdison's, but the Board remains well under the influence of SunEdison (3 of 9 seats, plus TERP's CEO and COO, both whom came from SUNE). Decisions over the purchase price of individual SUNE assets are done to be mutually beneficial for both entities. SUNE has the option to be paid in either cash or equity for the projects it drops down, but currently expects to receive 100% cash. CEO Carlos Domenech was the former CEO and CFO of SunEdison prior to its acquisition by MEMC, and was the President of the Solar Business thereafter. We believe he is a very capable CEO.
Wednesday, February 18th - We expect the dividend to be increased 11% to an annualized $1.20 (3.8% implied yield).
Tuesday, February 24th - SunEdison Capital Markets Day.
Ongoing - We expect the dividend to be raised each quarter, and we see a run-rate dividend of $1.57 per share by the end of 2015.
Are you sure you want to close this position TERRAFORM POWER INC?
By closing position, I’m notifying VIC Members that at today’s market price, I no longer am recommending this position.
Are you sure you want to Flag this idea TERRAFORM POWER INC for removal?
Flagging an idea indicates that the idea does not meet the standards of the club and you believe it should be removed from the site. Once a threshold has been reached the idea will be removed.
You currently do not have message posting privilages, there are 1 way you can get the privilage.
Apply for or reactivate your full membership
You can apply for full membership by submitting an investment idea of your own. Or if you are in reactivation status, you need to reactivate your full membership.
What is wrong with message, "".