On September 16, 2013, JinkoSolar Holding Co., Ltd. (NYSE: JKS) issued a press release announcing the commencement of a 3.5 million ADS offering. This news came only two trading days after announcing a new strategic direction described in JinkoSolar’s New Power Strategy.
In the light of this equity raising news, this article will follow up where the aforementioned article left off, using the described Jinko case and cost estimates there to compare the profit profiles of two different business models for project development – plant sales only vs. power sales only. The growth assumption will be that Jinko develop 200 MW in 2013 and 300 MW in 2014, and then can increase development growth in MW by 10% annually for 20 years; i.e., in 2015 they develop 110 MW more than in 2014, when they developed 100 MW more than in 2013. The time frame for the comparison will be the 20-year lifetime of a plant. The net profit assumptions for the two models will be based on the previous article’s estimates of $0.34 for plants sales and $0.09 for power sales.
Prices in the PV industry have been on a consistent decline due to cost reductions. Five percent annual price declines will be assumed in the comparison. Matching cost declines for constant profit margins, and thus 5% annual profit per watt declines, will be assumed. Note that both the install growth and price decline assumptions are lower than the observed trend the past 30 years. The reason for this is that the high rates of the past might change to rates that seem more sustainable for the long term, as we approach grid parity in market-by-market now and gradually move from excessive to moderate subsidy levels in most markets, as install volumes are gaining mass. China can be seen as a market with moderate subsidy level allowing long-term sustainable mass installs. This is in contrast to some markets that had excessive subsidy levels causing boom/bust types of markets. Spain is one such example.
The 5% decline will apply to prices of sold plants and to FiT rates for new plants, while FiT rates for existing plants remain fixed. This is an important difference between the two models, allowing slower decline of average profit per watt for the power sales model. A simple timing assumption where revenues for plant and power sales are based on the previous year’s built and owned plants will be assumed. For both models, a $0.05 expense per developed watt to cover operating and interest expenses will be added to the respective segment’s cost. These are indirect expenses that can be considered related to project development, but are not capitalized into the project assets due to their indirect nature. Given all these assumptions, here is how the two models would perform during the 20-year time frame (profits in million US$)(Jinko's Business Segment Profiles) Jinko's Business Segment Profiles.
Note how the table compares two alternative business models’ return on project development efforts, one where all developed plants are sold and one where all are kept. The growth of power generation capacity for the latter model is in Jinko's Power Generation Capacity Table. Now to the interesting part – the profit generation profiles of the two models: 10 years and 20 years
The 10 years profit generation table has zoomed in on the first half of the 20-year time frame. Zooming out to look at the full-time frame reveals in 20 years profit generation table.
In the beginning, plant sales would result in higher annual profits from the developed projects, while in the long run, keeping plants for power sales will generate much more profit. Some underlying factors can be seen in the table, such as generation capacity growing much faster than plant builds, especially in the early years, while plant sales grow at the same rate as plant builds. Another factor is the slower decline in power sales profit per watt due to FiT rate, and thus revenue per watt only declining for new builds and remaining fixed for existing plants.
One should acknowledge the political risk of national governments breaching FiT contracts during their 20-year term, while those selling plants only have business risk during the shorter (maybe 1-year) time frame from start of development until completion of a sale. The same thing applies to other ownership risks like natural disasters. So the normal risk/reward relation applies here. Geographical and legislative diversification can effectively reduce such risks. Jinko builds a lot of small- to medium-size plants across China, so the main exposure is to China’s national government policy, which to some extent applies to other operations in the country, too.
A couple of other things should also be noted. First, the above profit profiles are not actual profit projections, since coarse models using assumed parameters do not qualify for that. The purpose of the analysis is to identify the difference in profit growth profiles for the two models under identical conditions of project development and price declines. The second thing to note is that the very different profiles of the two models results from not assuming a high exponential growth in installs, offsetting the price declines with a wide margin. In such a scenario the plant sales model would come closer to the power sales profit growth profile. Alternatively, the same effect (to a different extent) is achieved by assuming no price declines. So one could say that the power sales model is the more conservative choice for the long-term investor wanting high profit growth with high certainty that does not rely on very high install growth or slow price declines.
One way to think of the two models is from a profit retention perspective. For the power sales model, profits of existing plants remain year after year for 20 years and any new project development adds directly to those profits, making profit growth very easy. For plants sales, each year you first need to develop the same amount of projects as last year plus an additional 5% to cover price declines just to retain last year’s profits; then any additional project development will add to the profits from last year. The benefit with the plants sales model is the quick return, which can be handy in a scenario with a lot of near-term investment opportunities and no other (attractive) funding source. In the example above it takes five years for the power sales segment to pass the plant sales segment on annual profits and eight years to pass it on cumulative profits.
A mix of the models in the early days of project development could make sense to fund a larger project development pipeline, but in the Jinko case – as opposed to many peers – they can be quite profitable on panel sales alone; thus, the project development for Jinko can be funded by panel sales instead of plant sales until it can be self-funded from power sales. The equity-funding requirement for project development is roughly $0.31 per watt, assuming the $1.25 CapEx and 75% debt leverage. For Jinko, this roughly means that selling one plant would fund the development of two new plants (one to replace the sold one and one for growth), or selling four panels to fund one plant panel, assuming their average selling price (ASP) trends toward US$0.70 with the large ASP increase in the EU and China happening now. For many peers, selling panels at the same $0.70 ASP (though right now many of them sell at higher ASP than Jinko) would only generate gross profits and not much net profits, so for them, unlike Jinko, it makes more sense to sell plants in the early days of project development to fund their pipeline unless panel ASP moves higher. Maybe this prerequisite is what is reflected in Jinko’s announcement of power sales focus compared to peers’ communicated intention to sell plants.
The equity raise just announced by Jinko could strengthen the perception that Jinko are aiming to fund project development investments, without use of the quick returns from plant sales, in order to quickly ramp up the power sales model. With a stretched balance sheet, this equity boost might be needed for the initial project development funding until panel profits grow from rising ASPs and power sales profits start to tick in. The equity issue has become very trendy among Jinko’s peers lately, so they are far from alone in exploiting recent stock price appreciations in the sector to strengthen balance sheets after a two-year ordeal of industry depression.
Comparing the Chinese panel-producing peer with the largest project pipeline, we see this opposite model where plants are built for plant sales. Canadian Solar (Nasdaq: CSIQ) is doing global project development focusing on Canada, the US, China and Japan, and in some of those markets can reach more than twice the revenue and profit per watt on plant sales that Jinko can achieve in China. SPVI CEO Robert Dydo describes in New Solar Investment Template Includes Canadian Solar a more detailed picture of this near-term profit-focused model.
Disclosure: Long JKS, CSIQ
