Long-term investors in next-generation biofuel companies have been mostly wiped out since the latters' respective IPOs. Despite this poor record, many investors attribute a high degree of probability to the future profitability of these companies. Such an assessment does not consider the high degree of uncertainty that is inherent in novel pathway technologies such as those employed by next-gen biofuel producers. This article describes the different types of inherent uncertainty and presents an example of their impact on investment returns. Production costs presented as point estimates and narrow ranges do not adequately account for this uncertainty and should be given diminished weight by investors performing their due diligence.
The next-generation biofuels (i.e., biofuels other than starch ethanol and biodiesel) sector has been one of the worst market performers in recent years (see figure). Whereas the S&P 500 has increased by 30% over the last two years, investors in next-gen biofuel companies have been largely wiped out over the same period. The best performer has been Syntroleum (NASDAQ:SYNM), which operates a renewable diesel facility as part of a joint venture and has lost 75% of its share value since February 2011. The worst performer over the same period has been biobutanol producer Gevo (NASDAQ:GEVO), which has suffered a share price fall of 92% since its February 2011 IPO. Renewable hydrocarbon producer Amyris (NASDAQ:AMRS) has lost 84% of its value since its September 2010 IPO, as has cellulosic biofuel blendstock producer KiOR (NASDAQ:KIOR) since its June 2011 IPO.
Shareholder equity has fared no better, with Amyris, Gevo, and KiOR all reporting declines of at least 52% and up to 99.7% since their IPOs (see table). The three companies have overseen a combined reduction of $549.2 million of shareholder equity since then, or an average reduction of 80% (I've excluded Syntroleum from this calculation because its IPO predates its entry into the next-gen biofuel sector by several years).
|IPO s/h equity ($MM)||Latest s/h equity ($MM)|
Sources: Company quarterly earnings reports
The extremely poor performance of these companies since their IPOs can be attributed to a number of factors. First, each of these companies was created to develop and commercialize a novel biofuel pathway. The next-gen biofuel sector is very similar to the biotechnology sector - Amyris and Gevo employ novel engineered microorganisms to produce biofuel, while KiOR employs novel catalysts to do the same - so it comes as no surprise that it is experiencing the biotech sector's extreme volatility as well. Furthermore, by employing novel biofuel pathways that have yet to be commercialized, each of these companies is operating at the very beginning of the learning curve and thus achieving very poor operating efficiencies. Costs are higher and yields are lower than they will be after a decade or more of commercial-scale experience. Finally, the companies in the table above are all start-ups and, as investors know all too well, most start-ups fail to last more than a few years, with failure almost always involving steep shareholder losses.
Despite the characteristics of these companies and the extremely poor performance of their share prices, however, I've found that many investors attribute a very high probability to their long-term success. This is unusual both because of the speculative nature of the securities and the performance of the underlying companies, all of which have overseen unexpected shutdowns, lengthy delays, and/or redesigned commercialization strategies. In discussions with these investors one reason in particular is repeatedly given for this optimism: the respective company's estimated pathway production costs. Determining the economic feasibility of a biofuel pathway is straightforward: if it can yield biofuel on an unsubsidized basis that is less expensive than its petroleum counterpart, the pathway will be profitable. If it cannot, then the pathway will ultimately fail. So when next-gen biofuel bulls cite a company's estimated production cost in support of their long position, they're staking their investment on the ability of the company to produce the biofuel for less money than the petroleum counterpart and to sell it for the same amount of money as the petroleum counterpart.
Next-gen biofuel production cost estimates are frequently cited by the companies and their investors alike. In its recent Q3 2013 quarterly earnings call, KiOR's CEO stated that it expects to achieve a production cost of $2.20-$2.30 per gallon over the short-term. Similarly, this recent bullish SA article on Gevo cites a statement by the company's CEO that its biobutanol production costs are 36-40% lower than that of the petroleum-based version. Assuming that their respective biobased products are sold for roughly the same price as the petroleum-based versions, investors can expect based on these statements to see the companies achieve substantial gross margins after commercial-scale production begins.
The problem with production cost estimates
I will be blunt: these production cost estimates provide little value to investors. In fact, they can actually be dangerous to investors by lulling them into a false sense of security. The estimation of next-gen biofuel production costs is an inexact science (some would prefer to call it an art) that is fraught with uncertainty. Let me explain.
Unit production cost estimates operate primarily as a function of input costs, capital costs, and co-product market prices (other factors also play a role but cost estimates generally exhibit little sensitivity to them). In academia we refer to the estimated unit production cost as a "minimum fuel selling price", or MFSP. As the name suggests, the MFSP is the minimum price that the biofuel (or bioproduct) can be sold for while achieving specific profitability thresholds (generally a net present value, or NPV of $0 and an internal rate of return, or IRR of 10%). The pathway analyzed is considered to be "economically feasible" when the MFSP is lower than the market price of the primary product and vice versa. This methodology makes the MFSP result extremely sensitive to the assumed factors and making an unrealistic assumption can greatly skew the result. I recently read an academic paper reporting a MFSP for a novel biofuel pathway that was a fraction of the cost of the petroleum version. Upon further investigation, however, I found that the authors had assumed that the hypothetical biorefinery could be built for a fraction of the cost of a petroleum refinery on a volumetric basis, which is extremely unlikely. Using a more realistic installation factor generated a far less attractive result.
I generally assume that the estimated production costs reported by companies employ more realistic (albeit unpublished) assumptions simply because the penalties that CEOs face for failing to do so are much more severe than those faced by academics (I have yet to hear of academics being exposed to class action lawsuits due to the contents of their publications). This methodology still makes the assumption that the individual input costs and output prices are known with 100% certainty, however. Investors in particular know that such an assumption isn't realistic given the high degree of volatility exhibited by most commodity prices.
As an example of how this volatility can affect unit production cost estimates, take two papers that several colleagues and I have recently published in the journal Fuel. The first paper calculates a MFSP for "green" gasoline and diesel fuel produced at a hypothetical fast pyrolysis and hydroprocessing facility of $2.52/gal. The second paper builds on this by calculating a probability distribution for the same facility's 20-year NPV based on historical fossil fuel price volatility and employing Monte Carlo analysis. (The hypothetical facility employs natural gas as an input and its outputs compete with petroleum-based gasoline and diesel fuel.) NPV calculations employ the same variables as MFSP calculations, with the only difference being that NPV becomes an endogenous variable and primary product price becomes an exogenous variable (these are reversed for MFSP calculations). A median NPV of $117 million was calculated for the hypothetical fast pyrolysis and hydroprocessing facility (see figure). However, there was only a 22.5% probability that the facility's NPV would be within +/-5% of this median ($112 million - $122 million). In other words, while I could accurately state that the most likely estimated NPV for the facility is $117 million (or MFSP of $2.52/gal, since MFSP is simply another way of presenting NPV), left unsaid in this statement would be that there is a 77.5% probability that the result would not be within this $112 million - $122 million range.
Source: Brown and Wright (2013)
Furthermore, the above NPV range assumes that uncertainty is only affected by fossil fuel price volatility. It does not account for capital cost estimation uncertainty (which can be extremely high for alternative fuel facilities), feedstock price uncertainty and volatility, yield uncertainty (which can also be high, especially during the facility startup phase), or future price estimate uncertainty (which is different from price volatility). These factors could be expected to further reduce the 22.5% probability calculated above if accounted for (and they are certainly encountered by next-gen biofuel producers).
I do not mean to say that production cost estimates are without merit. They provide analysts with a useful initial benchmark, particularly when used to compare the economic feasibility of multiple novel pathways. Production cost estimates can take several months to complete and, as such, they also represent a first step that future analyses can build upon. When presented as point estimates or narrow ranges, however, production cost estimates are of little (if any) net benefit to investors and can even be harmful. This is particularly true when older estimates continue to be used after the original conditions have changed.
The narrow production cost ranges presented by next-gen biofuel producers must be considered in light of the high degree of uncertainty of their respective pathway technologies. No matter how rigorous the methodology underlying the stated estimates, they are in pertain to technologies that have yet to be demonstrated at scale and are therefore very uncertain. Even were the estimates in reference to tested pathways, the significant degree of price volatility underlying the pathway outputs and inputs adds further uncertainty to them; see, for example, the behavior of the corn ethanol crush spread YTD in 2013, not to mention over the last several years. Unit production costs presented in the form of point estimates or narrow ranges are not appropriate for investments in companies that rely heavily on novel pathway technologies. Investors in next-gen biofuel companies are therefore strongly encouraged to properly account for uncertainty when performing their due diligence. At the very least they should consider that investments in such companies are highly speculative by their very nature, regardless of the potential attractiveness of the underlying pathway.
Disclosure: I am long KIOR. 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.