Weaker ethanol and sugar prices as well as the core operating structure had a negative impact on Cosan (CSAN)

Volatility and complexity have long been aspects of Cosan (NYSE: CSAN) story that investors simply have to accept if they want to own the stock. None of this has gotten easier since I last wrote However, there is no clear opinion about this Brazilian raw materials and logistics group, and shares have been hit by weaker commodity prices, unfavorable currency fluctuations and a lack of enthusiasm about the company's complicated and often opaque structure and management's longer-term capital allocation plans.

I can certainly argue that the valuation is too low today, but this argument must always be accompanied by the caveat that a large part of this thesis is based on unpredictable commodity price fluctuations. I like Cosan's core businesses, especially the Raizen Ethanol, sugar and fuel trading as well as the Rumo railway activities, but I like the Complexity and I don't think management communicates particularly well with investors, although I do believe they have a good eye for long-term value creation.

I wouldn't be surprised if Cosan shares are trading significantly higher in a year's time, but given the sheer number of companies and stocks available for investment, I find it increasingly difficult to argue that it's worth it for an investor to grapple with the complexities of owning Cosan shares.

Okay Q2 results despite many moving parts

Analyzing Cosan's quarterly results is never particularly easy, and the company makes it worse by seemingly changing the way it presents information (the earnings releases) every quarter. On top of that, there are a lot of accounting adjustments and consolidated/unconsolidated details you have to wade through.

Overall, second quarter results were OK, despite weaker commodity prices for ethanol and sugar, little progress on earnings and some sluggishness in volumes for lubricants and natural gas. Rumo was an outstanding performer and I have little to criticize here.

Revenue increased 15% year-on-year and approximately 8% quarter-on-quarter. Raizen, the company's joint venture with Shell plc (SHEL) grew 18% year-on-year and about 8% quarter-on-quarter, with ethanol and sugar both up strongly (19% and 21%, respectively) on higher volumes (up 19% and 26%, respectively). Rumo saw revenues rise 29% year-on-year on modest traffic growth (up 3%) and a 32% improvement in tariffs. Other contributors were weaker: Compass lost 5% on weaker residential gas volumes and still-sluggish industrial demand, Moove gained 3% as prices offset some volume weakness, and Radar fell 6%.

EBITDA is where things get complicated. Raizen's EBITDA fell 29% as reported, but when you adjust the numbers for inventory, IFRS 16 and other factors, underlying “adjusted adjusted” EBITDA rose 13%. The rest of the business is thankfully more straightforward, with Rumo up about 49% year-over-year and Moove up about 16%; Compass saw similar discrepancies between reported adjusted and “adjusted adjusted” numbers, with EBITDA up either 42% or 18% depending on how you calculate it. Ultimately, EBITDA rose 15% using Cosan's adjusted method, or as much as 26% using another method.

Prices for the core countries of Southeast Europe have become more demanding

Since Raizen accounts for a large portion of revenue (almost three-quarters) and EBITDA (almost 30%), market dynamics for sugar, ethanol and energy (collectively referred to as “SEE”) play a major role.

When I last wrote about Cosan, I liked the prospect of near-term leverage on healthy sugar prices and market momentum, but I was concerned about the longer-term outlook. At the time of writing, prices were at $0.19/lb, and although they rose to $0.28/lb in late 2023 (when Cosan shares were significantly higher), they have since retreated to around $0.177/lb for spot prices, and futures suggest further decline in the low-to-mid $0.17/lb range.

The same is true for ethanol prices in Brazil. In São Paulo, prices for anhydrous ethanol have increased from R$3.08/litre to around R$3/litre, although admittedly this is not a particularly dramatic drop.

With the fall in sugar prices, ethanol production has become relatively more attractive, but the market is struggling with oversupply as companies are limited in their ability to shift production from one to another. Efforts are underway in the Brazilian parliament to increase the ethanol share from 27.5% to 30%, which should lead to a consumption increase of about 4%, but the process to pass the “Combustivel do Futuro” law is slow and a decision has been postponed until September (at the earliest).

Meanwhile, it is difficult to make more than educated guesses about the sugar market. I do not mean that lightly, there are just too many variables related to weather and other governments' export policies to make any definitive statements about pricing. I would say that the market currently looks relatively balanced given expected crops and exports over the next two years, and bad weather or export restrictions could lead to another round of higher prices.

Numerous moving parts, and not all of them are easily visible

One of the ongoing issues with Cosan is that due to the nature of the company structure, there are a lot of moving parts and accounting complications (like the previous EBITDA discussion). I don't think Cosan is intentionally trying to be obtuse, but I also think they could do more to provide clean, consistent reports that give investors easy insight into the health of the company – institutional investors can spend a lot of time digging into the accounting, but the average retail investor can't.

Speaking of moving parts, Cosan has completed its collar financing for its stake in the Brazilian iron ore company. valley (VALE) and converted 66 million shares into a direct stake. The company now owns a direct stake of 4.2% and a call spread of another 1.4%. This investment in Vale has always been at least somewhat controversial, as there are hardly any obvious synergies with the other parts of Cosan and management has not made a really convincing case for it from a strategic point of view.

This is part of an ongoing problem at Cosan. While I believe management has a good eye for value and I understand why they don't want to publicly disclose their investment plans, it creates the impression of a hodgepodge of assets with little coherence or synergy. Holding companies have largely fallen out of favor and those that remain typically trade at double-digit discounts to the sum of their parts due to the complexity and inefficiency of these structures.

The outlook

Cosan is not the most efficient REE producer in Brazil and probably never will be (Adecoagro (AGRO) is doing better there), but earnings have improved and I would say that overall the SEE business is still doing well. If Raizen were available for direct investment (it is traded on the Brazilian stock exchange), I would seriously consider it. I am even more optimistic about Rumo, as management has turned around a company that had many operational problems and where there is still a pressing need for more efficient logistics to get key products like grain and fertilizers to Brazilian ports.

The rest of the business is more of the “it's OK, I guess” variety. I don't have any particular issues with Compass (natural gas distribution) or Moove (lubricants), but they are not thesis changers for me.

I expect Cosan's revenue to grow by around 3-4% in the long term, and with further development of Rumo (including a new port terminal project the company announced earlier this year) it could go even higher. EBITDA margins are showing a choppy upward trend, and I expect further improvement with a margin of 13% this year, over 14% next year, and around 15% in FY2026.

However, given the company's significant reinvestment needs and low margins in the raw materials business, I do not expect FCF margins to significantly exceed the low single-digit range in the long term and sustainably.

Even with modest expectations of FCF leverage, discounted cash flow suggests a fair value of around $15. EV/EBITDA suggests even more undervaluation, with a multiple of 4.5 suggesting a fair value of over $17. While this is a lower multiple than those I have used in the past, I believe it is reasonable compared to how global SE and agricultural companies have traded in the past, especially given the higher interest rate environment.

The conclusion

I think there is a good argument that Cosan shares are too cheap and I think the share price already contains a lot of pessimism related to the SEE business. Likewise, I think Rumo is a very attractive asset and another one I would buy if it were readily available.

However, investors looking to invest in Raizen and Rumo must deal with all the additional baggage and challenges that come with Cosan's corporate structure and capital allocation/investment philosophy. While I respect Cosan's management very much and do see value here, I find it increasingly difficult to argue that the value I see is worth the effort, but of course readers should decide for themselves.