Terraform Power’s Recent Moves Support Dividend Growth Of At Least 15% In 2016
Summary Terraform Power closed the first part of their planned transaction with Invenergy, 832MW of net wind power plants, increasing their current portfolio from 1.9GW to 2.8GW. I found it interesting that the company ended up changing the financing package for this transaction, which is expected to deliver unlevered CAFD of $139mil in 2016. Terraform also recently updated their purchase agreement with SunEdison for the producing assets of SunEdison’s deal to buy Vivint Solar. The updated deal projects that Terraform will pay $799mil to purchase 470MW of producing assets. These assets should produce annual CAFD of around $73-75mil. Both transactions support an increased dividend in 2016. The actual amount will depend on the timing of the transactions, but I expect 2016 exit rate of at least $1.60/share. Terraform Power (NASDAQ: TERP ) has had a pretty busy last couple of weeks. The company completed a large part of their planned transaction with Invenergy on Wednesday, buying 832MW of wind assets. This comes on the heels of the announcement last week of a renegotiation of the terms of SunEdison’s (NYSE: SUNE ) purchase of Vivint Solar (NYSE: VSLR ). The updates have removed many of the concerns investors have had with TERP, and the stock has responded in kind, as it has almost doubled from its late November lows. As I explain below, these transaction also support a continued dividend increase in 2016. TERP data by YCharts TERP came out with solid 3rd Quarter results in early November, but management’s unwillingness to confirm their previously estimated 2016 dividend increase to $1.75, and liquidity concerns at sponsor SUNE, led to the stock plunging over the next two weeks $6.73. At that point, the current yield was 20%. The market finally came to its senses when David Tepper announced a large stake and sent an open letter to management. Invenergy Transaction: Sources/Uses of Fund Changed One of the interesting takeaways from TERP’s announcement of the Invenergy transaction is the fact that they changed the way they ended up financing it. Back in July when the deal was made, the plan (see page 17) was to place half the MW into TERP immediately, and hold the other half in a SUNE warehouse. From August thru the 3Q earnings call in November, the plan (see page 18) was to only drop down 265MW and place the rest in a structured warehouse. Now it seems that management has decided to place all of the projects directly into TERP immediately. Sources of Cash Uses of Cash Non-Recourse Project Debt assumed or incurred with respect to transaction $801m 832MW of Wind Assets located in US and Canada $1,962m Pro-rata portion of $500mil new non-recourse term loan $417m Cash on Hand (incld proceeds from TERP $300mil senior note offering in July 2015) $744m $1,962m $1,962m The press release notes that, “once all projects are operational, the first year adjusted EBITDA (before minority ownership) is expected to be $147 million, and unlevered CAFD (before all project and HoldCo debt payments) is expected to be $139 million.” Unfortunately, this doesn’t clarify how much we should expect the project and HoldCo debt payments to be, so it’s tough to predict how much of the unlevered CAFD will actually be available for dividend payments. My best estimate is to use the initially projected cash on cash yield of 8.4%, which equates to about $62mil (8.4% x $744mil cash on hand). We’ll have to wait for TERPs 4Q results for more details. The new $500mil term loan charges LIBOR + 5.5%, with a 1% LIBOR floor, meaning that TERP is currently paying 6.5%. It matures in 2019 and can be prepaid anytime, so it’s very likely that TERP will refinance this as soon as they can organize something with better terms. Vivint Transaction Terms Improve Last week TERP and SUNE announced that they had improved the terms of the Vivint transaction. TERP was able to reduce their initial purchase commitment from $922mil down to $799mil, by only paying for completed installations, and paying a reduced fee of $1.70/MW. The final total will depend on the actual number of producing MW transferred when the Vivint deal closes sometime during Q1 2016. In their Q2 earnings presentation, management noted that they expected the 523MW to generate average unlevered CAFD of $81 annually, so I project that the 470MW delivered at close will generate $70-75mil annualized unlevered CAFD. Considering the fact that TERP used substantially all of their cash on hand at the end of Q3 to fund the Invenergy transaction, it’s likely that they will be drawing on their revolver for much of the $799mil. The revolver’s rates are currently under 3%, so the annual interest would be about $24mil. Thus, I expect final CAFD to be about $50mil. 2016 Updated CAFD Projection Based on their 3Q presentation, TERP expected to generate CAFD of about $208mil before taking into account the Invenergy and Vivint transactions. If we assume that the Vivint transaction closes by the end of Q1, we should see annualized CAFD at the following levels next year: Quarter End Q1 Q2 Q3 Q4 Annualized CAFD $270m $320m $320m $320m CAFD distributed (85%) $230m $272m $272m $272m Per Share Quarterly Distribution $.35 $.40 $.40 $.40 The quarterly distribution estimates include the effect of SUNEs IDRs. I don’t expect TERP to actually increase the dividend to $.40 for Q2, even though it seems that operations would allow this. Rather, it’s likely that they will prefer to show steady quarterly increases up to $.40/share in Q4. This confirms my view that these transactions will cause TERP to increase their dividend by at least 15% over the next year.