Prices for a second‐generation biofuel industry in Canada: Market linkages between Canadian wheat and US energy and agricultural commodities

The production of biofuels is limited in Canada, but the availability wheat straw as a second-generation (i.e., cellulosic) feedstock an exciting prospect for future development biofuel industry. success such industry will depend on ethanol prices and related to straw. These are likely be influenced by markets existing first-generation United States. Therefore, motivation this paper investigate relationships between Canadian US corn, ethanol, gasoline prices. We employ DCC-MGARCH enhanced VEC model time-varying among these markets. Results indicate that there positive Our results add better understanding level integration select agricultural energy More specifically, price identified highlight several sources risk may affect financial commercially producing from Canada. This information particular interest prospective investors policymakers. La de biocarburants est limitée au mais la disponibilité paille blé comme matière première une perspective intéressante pour le développement futur d'une industrie des deuxième génération (c.-à-d. cellulosique). Le succès telle dépendra prix futurs l'éthanol et liés à blé. Ces sont susceptibles d'être influencés par les marchés l'industrie actuelle aux États-Unis. Par conséquent, cet article d'étudier relations entre du canadien maïs, l'essence Nous utilisons un modèle amélioré étudier variant dans temps ces marchés. Les résultats indiquent qu'il existe positives l'éthanol. Nos contribuent meilleure compréhension niveau d'intégration certains agricoles canadiens énergétiques américains. Plus précisément, identifiées mettent en évidence plusieurs risque qui peuvent affecter financier commerciale d'éthanol partir informations seront particulièrement intéressantes investisseurs potentiels décideurs politiques. Federal provincial governments Canada require renewable fuels like blended with reduce greenhouse gas emissions (Government 2017). First-generation which uses crop grains corn feedstock, has been linked higher food (Zhang al., 2013). Second-generation residue waste alleviate concern because supply residues not Although market made does currently exist represent potential feedstock. Western provinces Alberta, Saskatchewan, Manitoba) produced 28 million tonnes 2016, was approximately 87% all 2018). Because by-product grain production, factors ethanol. prospects using patterns markets, could interdependences In study, we whether what extent interdependent.11 It assumed Canadians takers respect international reflected potentially ways. Corn they both commodities used food, inputs Ethanol also input due fuel standards. have relevance Though no western current (which produces straw), provide insights into relationships, if develop. contribute literature three general First, our focus influence biofuels. studies relevant oilseed crops. purpose research largely analyze interrelationships context how gasoline, oil, global general, authors find integrated direction magnitude vary. For example, large crude sugar, tend lead smaller Serra (2011) observes Brazilian sugar long run, neither nor oil Instead, mostly dependent local yields. Allen al. (2018) finds causal change depending high or low volatility environment, significant run relationship. With focused countries States, Brazil, Europe. (Chen 2010; Creti 2013; Du 2011; Nazlioglu, 2018; Hameed & Arshad 2009; Chakravorty 2017; Saghaian, Hao Serra, 2011). But second-second generation cause other countries, become more relevant. second contribution Canada; interacts A third dynamic Most focuses dynamics. analyses can useful about adjustments spillovers markets; do interdependences. There substantial degree downside upside producers situations where output relationships. help direct management strategies, hedge designs estimates correlations returns assets under contract. next section specifies econometric approaches quantify Section 3 discusses data used, while 4 presents results. 5 summarizes conclusions analysis implication emerging time-series consists estimating mean equations together system maximum likelihood method.22 Berndt-Hall-Hall-Hausman (BHHH) algorithm estimate system. equation vector error correction (VECM), multivariate generalized autoregressive conditional heteroskedasticity (MGARCH) correlation coefficient (DCC) specification. Engle-Granger (Engle Granger, 1987) Johansen cointegration tests (Johansen, 1991) identify long-run each series. Once cointegrating identified, VECM employed short- specification Specifically, generate time-variant (Engle, 2002). variable, changes percent price) period, t, K number consideration analysis. α constants residuals. considers short-run market. effects captured by, , matrix estimated parameters lagged changes, . error-correction process,. matrix, Π, function matrix44 r dimension represents vectors. (ECPs), θ, transpose μ. operation prices, yield equations. Each element process, interpreted speed at respective variable adjusts deviation equilibrium Equation (5) described (1). (6) process VECM, follows covariance 7. Matrix forms Equations (5)–(7), presented Appendix B (5.1), (6.1), (7.1), respectively. weekly July 4, 2008 April 27, 2018. CW Feed Wheat, Track Thunder Bay, measured US$/tonne. converted exchange rates Reserve Economic Data (Board Govenors System (US), 2020). rest time series, indicators spot Chicago US$/gallon. first-generation, soybeans. Yellow No. 2 US$/bushel. Gasoline summary statistics series Table 1. coefficients variation (CV) suggest least volatile, most volatile. Figure 1 display similar patterns. All experience peak 2008, result yields 2006-2007, growing demand maize (Abbott 2008; Mitchel, Trade Markets Division FAO, Trostle, 2008). High 2010 2012 attributed those caused (Coulibaly, Wheat fell dramatically 2013. adjusting (1) lower than average 2013, were exacerbated drought poor 2012, (2) above-average since 2013 (Bureau Labor Statistics, 2012; Schnitkey, 2019). During eliminated Board sole purchaser wheat. Some (e.g., Brewin, 2014) tended multiple-seller followed. However, difficult separate occurred varying above stocks, bumper harvest 2013–14 (Slade Gray, volatile reacting aforementioned Furthermore, States approached ethanol-gasoline E10 blend wall 2013–2014, policy uncertainty Environmental Protection Agency (EPA) regarding affected (Baumeister Knittel 2015). then declined substantially 2014 remained since. Irwin (2019) rising 2014, demand. decreased 2015. stagnant increased Kilian, 2016; Prest, Preliminary conducted suitable approach modeling process. Augmented Dicky-Fuller (ADF) Kwiatkowski–Phillips–Schmidt–Shin (KPSS) test stationary non-stationary. augmenting lags chosen 26 based Akaike criterion (AIC).88 ADF wheat, 2, 8, 24, KPSS absolute values non-stationary follow unit root Next, log transformed first-difference taken obtain changes. shows periods across generally correspond falling above. series,99 1, 7, 9, 2. stationary. suggests statistically relationship logged 5-percent level.1010 value statistic larger 5% critical value. null hypothesis cointegrated rejected. carried out one series.1111 requires defined lag structure. AIC first determine optimal lag-length underlying auto-regressive length found two lags. 3. trace eigenvalue insignificant when ). being modelled simultaneously normalized affects ceteris paribus.1212 parameter paribus interpretation. 1% increase associated 1.08% price. full estimated. Several notable 4. react consistent prior studies, noted appear little effect (Etienne Trujillo-Barrera 2012). (2012) note reallocated conventional animal feed food) use, able withstand shocks dynamics driven weather conditions, government policies, macroeconomic Battisti, Second, period 0.16%. Third, their own 0.11%. Finally, ECPs. respond deviations values, not. exogenous following shock system, adjust back rate 3.5% 7.3% per week, An important assumption validity errors homoscedastic ARCH-LM performed residuals autocorrelation significant. validate decision DCC-MGACRH 1313 serially correlated equation. statistical significance transmission ECPs remain same. exception longer now 2.8% 4.3% difference adjustment speeds VECM-DCC-MGARCH means simple deterministic disturbances picked up DCC component. One investigating use DCC-GARCH appropriate lambda satisfy constraint, sum equal 0.95 level. supported. VECM-DCC-GARCH residuals, fails reject serial correlation. 6 reports ARCH GARCH coefficients. measure vulnerability persistency volatilities, Persistency previous period. close highly persistent. Vulnerability individual significant, implies somewhat vulnerable DCCs, interdependent. pairwise (decrease), another (decrease). some sensitive economic events others. during 2008–2009 crisis tends quite 2013–2014. From 2015, less correlated, correlated. again, coincide Recall closed August 2012. visual evidence corresponding lost its monopoly powers. towards end through stabilized did arise closure. experienced corresponded dip 2014. agriculture wall. trends correlations; remains relatively flat Following 3, appears logical flow order strongest weakest strength commodity corn) strongest. Similar costs farm inputs, developing livestock explain why (Trostle, corn. would expect strong largest value-added growers US, accounting 96% produce 2011 2017 (Renewable Fuels Association, third, fourth, fifth explanation blending mandates constraints Good (2016) curve inelastic volume intersection point curve, typically 2007. any observed extensive body recent explored understand growth (Filip 2019; Hochman Zilberman, 2016). Little work done Prospective only invest operations it profitable so, heavily investigated gain scenarios emerge implications producers. As precursor analysis, exists four consideration. identified. predictability first- reflect reflective primary ties gasoline. mixed pass along blenders form benefit On hand, decrease follow. Short-run dictated relationship, profits (if relative prices) short-run. return Similarly, increase), decrease). (negative) causes (lower) encourage farmers grow (less) frequently. turn, available production. variability supply, makers consider strategies counteract unfavourable create market, persists. persists 16 weeks after occurs. need adopt adapt quickly disappear. pairs move constant over time. 2009 (during collapse) compared years. (Ji, Bonato, Büyükşahin Robe, 2014). Pairwise involving major disruptions economy producers, indicating counter disruptions. good news fluctuating transmit changing stronger suggesting supplies suffer fluctuations faced suppliers. actively manage movements mitigate risk. mitigations common strategy gross processing margin (GPM), known “corn-crush spread”, distillers-dry-grains, DDG's) (CME Group, GPM cellulosic producer tonne (decrease) environment (a) GPM), (b) GPM). shifts margins, want hedging offset movements. structure new Investigating employing non-linear regime-switching asymmetric adjustments. Future change, conditions financially sustainable. funding First Research Excellence Fund part University Alberta's Energy initiative (Project ID: RES0036609). responsible errors, well opinions expressed paper. Please note: publisher content functionality supporting supplied authors. Any queries (other missing content) should directed author article.