The stock performed well with a 42% return YTD.
Q3 CAFD has been aided by a volatile item.
While the 2019 outlook is at risk, the stock trades close to my DCF FV.
It may well be the right time to moderate exposure to AY.
Just two weeks ago, Atlantica Yield (AY), one of my favorite stocks, touched a 52-week high of $26.15 and has stayed close to that level since. This 52-week high also happens to be the highest stock price since the summer of 2015, when the YieldCo sub-sector crashed. Adjusted for last week’s $0.41 ex-dividend, the stock price would have hit a fresh 52-week high last Friday. We are at the pinnacle of a 42% return rally over the past year (both 12 months and year to date). Either way, the point is clear, Atlantica has performed well.
Atlantica Yield 5-year stock price chart. Source: Seeking Alpha, Dec. 1st, 2019.
The question is whether the stock is still an attractive buy or not. A good starting point is reviewing recent events and future catalysts.
Q3 results update
In November, AY published Q3 results that looked unexciting with CAFD up and EBITDA down YoY. The positive takeaway for investors was the 2.5% quarterly dividend increase from $0.40 to $0.41. This further increased Atlantica’s dividend track record and has given the market more comfort with the stock. In my view, however, the dividend level or yield is not as important as the quality of the assets, current results, or future prospects.
EBITDA was a bit below par in the quarter, mostly caused by poor performance of the North American solar farms. The overall picture was salvaged by strong CAFD generation. For a more complete overview, I added Atlantica’s Q3 CAFD reconciliation below.
Source: Atlantica Yield Q3 release.
There are many figures in the table, but the most important ones are the adjusted EBITDA in bold, and everything from there to CAFD. Specifically, I like to filter for one-offs. For example, it is no problem when CAFD is impacted by temporary working capital (aka seasonal) items like deposits into restricted accounts, or restricted cash at project level, but non-monetary items and dividends to noncontrolling interests are likely recurring.
Atlantica’s CAFD in the quarter was propped-up by a $26.9m dividend payment from its unconsolidated subsidiaries, on which neither the press release nor the full Q3 report provide any clarification. Because the pro-rata share of EBITDA is only $7m, it is not reasonable to assume that the $27m distribution of Q3 will be recurring.
There are multiple seasonal working-capital-like items that tend to reverse in Q4. Last year, these items amounted to a negative $195m for the first nine months, and these seasonal items were $184m positive ('reversed') in Q4 2018. This seasonal cash flow glut is sorely needed to fund principal amortization of project debt as Q4 isn’t a great quarter, EBITDA-wise.
I created the CAFD waterfall chart below to show YTD CAFD excluding these volatile/seasonal items.
Source: author’s own calculations. Cash EBITDA excludes minority adjustments and non-monetary items. Dividend adj. is an adjustment factor to account for the volatile nature of dividends received from unconsolidated subsidiaries in 2019.
Of course, to get a sense of proportion it is good to compare this to last year. A similar waterfall chart for 9M 2018 is shown below. Because project debt service for a given project remains stable over the years, repayments are expected to go up, while interest expense should decline. Save for small acquisitions and an occasional refinancing, this explains a large part of the shift of interest to debt repayments in 2019 versus 2018. Therefore, this is not a very interesting or concerning item.
What should be concerning is that the seasonal items in 2019 amounted to just -$182m (vs -$195m in 2018). This is an issue because debt service will be higher in Q4 2019 than it was last year while the seasonal 'wintercoat' consisting of reserves at project level is thinner than it was last year.
The company achieved a CAFD of $39m in Q4 last year and though the seasonal/volatile items are hard to forecast, it doesn’t look too good, given that there is less potential for restricted cash reversal, while debt repayments are up.
Source: author’s own calculations. Cash EBITDA excludes minority adjustments and non-monetary items.
At the start of the year, Atlantica issued a FY 2019 CAFD guidance range of $180m to $200m. To reach midpoint in this guidance, the company has to generate CAFD of $50m in Q4. As suggested, I would already find it quite an achievement if the company can beat last year’s Q4 CAFD of $39m. So far, the company hasn’t issued an update on its 2019 CAFD guidance since the publication of the 2018 annual results in February.
The bottom-line is that CAFD could disappoint in Q4. However, management has some room to prop-up CAFD. They can potentially manage working capital and restricted cash to offset the EBITDA headwinds experienced this year, but I would be surprised if CAFD came in at the top end of the annual guidance.
Though YieldCo investors often focus too much on dividend levels, short term results that are influenced by weather and volatile items like working capital, it is the assets that really matter. So while I think that CAFD generation in Q4 will move the stock, it is really not that important for its fair value.
When we zoom out and look at multiple in the sub-sector, Atlantica doesn’t look that expensive versus its peers, even when we take the low end of 2019 CAFD guidance.
*For this calculation, net project debt was excluded from EV, and interest on corporate debt was added back to CAFD. **Pro forma payout uses run rate CAFD. Source: author’s own calculations and estimates. Figures for Transalta Renewables (OTC:TRSWF), or RNW are in C$. For NEP, the proportionate share of debt, EBITDA, and CAFD to public unitholders was used after also accounting for convertible preferred shares and financing deals.
One of my favorite metrics, the EV/EBITDA of AY is still comfortably below the lowest figures seen at some of its peers. A metric that the market seems to care about is the stock CAFD yield (which proxies dividend capacity), and on this, Atlantica is also among the cheapest of its peers. What makes this better is that corporate debt at AY is relatively modest.
I have written about the valuation of AY before, and it is ‘cheap for a reason’ as one calls it. Though Atlantica has a nice and long contracted revenue base compared to peers, its outlook after the power purchase agreements expire is bleak. This is due to its CSP solar assets, which in my opinion, have a future that isn’t as bright as that for PV solar assets. That said, I do find most of AY’s peers either overvalued or fairly valued as all of these companies own depreciating assets for which no reasonable provisions are made by CAFD. Therefore, I believe that a DCF model is the best way to value YieldCos.
Atlantica’s Solana Solar Farm in Arizona. Source: Abengoa.
Late September/early October, I had a DCF price target of $27.25 for AY. After the $0.41 dividend, and a cut in 2019 CAFD expectations, I estimate fair value to be $26.7. The stock trades at about $26, and I think that the 3% upside to my fair value estimate doesn’t make the stock a particular attractive buy at this point.
The stock of Atlantica Yield has performed much better than the business this year. Though the short-term performance of the assets is not that important, investors do tend to respond to CAFD disappointments and guidance misses. Aside from that, the stock has almost caught up with my estimate of fair value. Therefore, I think it is the right time to moderate stock positions in Atlantica Yield.
Disclosure: I am/we are long CWEN.A. I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it (other than from Seeking Alpha). I have no business relationship with any company whose stock is mentioned in this article.