BRM payments and risk balancing: Potential implications for - PowerPoint PPT Presentation

BRM payments and risk balancing: Potential implications for financial riskiness of Canadian farms Nicoleta Uzea, University of Western Ontario Kenneth Poon, University of Guelph Dave Sparling, University of Western Ontario Alfons Weersink, University of Guelph Agriculture Bio-Economy Research Agri-Business Food

Introduction • BRM programs major component of government-provided subsidy to agriculture – Margin-based whole farm program (CAIS/ AgStab) + ad-hoc payments – Worries that government program may ‘crowd – out’ private risk management strategies • Plant riskier crops (Turvey 2012) • Reduce incentive to use crop insurance (Antón and Kilmura 2009)

Risk Balancing Hypothesis • Gabriel and Baker (1980) suggest operators may manage risk by trading business risk with financial risk Business Risk + Financial Risk ≤ Total Tolerable Risk • Business risk = volatility in income • Financial risk = level of leverage

Implications of Risk Balancing for the Effectiveness of BRM Programs • If farmers indeed balance BR and FR, BRM programs (assuming they do reduce BR) may lead farmers to take on more FR than they would otherwise • This increases the risk of equity loss • Do BRM programs crowd out farmers’ financial risk management strategies? • No studies have looked at the extent of risk balancing on Canadian farms and the impact of BRM payments on the likelihood of risk balancing

Research Problem Is there empirical evidence to suggest BRM programs crowd out farmers' financial risk management strategies?

Data • Ontario Farm Income Database (OFID) – Tax + production data on Ontario CAIS/ AgriStability participants • Detailed income and expenses • Unique farm ID à panel data possible (2003 to 2010) • Focus on 3 sectors: Field Crops, Dairy, Beef Sector ¡ Field ¡Crops ¡ Dairy ¡ Beef ¡ Number ¡of ¡farms ¡ 3,860 ¡ 236 ¡ 1,854 ¡ in ¡panel ¡data ¡

Business Risk vs Financial Risk • Measuring Business Risk (BR) Theory ¡ OFID ¡measure ¡ Coefficient ¡of ¡VariaHon: ¡ σ Earnings ¡Before ¡Interest ¡and ¡ cx Tax ¡(3 ¡years)* ¡ * ¡First ¡BR ¡measure ¡looks ¡at ¡volaHlity ¡of ¡income ¡in ¡2003-­‑2005 ¡ • Measuring Financial Risk (FR) Theory ¡ OFID ¡measure ¡ σ I ¡ ¡ ¡ ¡ ¡Interest ¡expense ¡ ¡ ¡ ¡ ¡. ¡ Earnings ¡Before ¡Tax ¡ cx(cx - I)

Measuring the Extent of Risk Balancing • Correlation analysis - Per farm: correlate between 5 pairs of BR and FR measures (2003-2005 BR to 2006 FR) - Spearman’s rank-order correlation – chosen to go around negative values for BR and FR - Extent of risk balancing behaviour = share of farms with negative significant correlation

Extent of risk balancing Dairy Field Crops 2 1.5 Density 1 .5 0 -1 -.5 0 .5 1 -1 -.5 0 .5 1 r(rho) Graphs by sector • 224 crop farms (5.80%) & 11 dairy farms (4.66%) with significant negative correlation – Small number of pairs means correlation values has to be very high to be significant ( ≤ -0.9)

Factors Associated with Risk Balancing - What Is the Impact of BRM Payments? • Risk balancing: Looking at movement of BR and FR over time – If BR goes down from 05-06, does FR go up from 06-07? • Estimated logit (random effects and fixed effects) and probit (random effects) models Dependent ¡Variable ¡ RISK_BAL ¡ • if ¡FR ¡moves ¡in ¡opposite ¡direcHon ¡of ¡BR ¡in ¡previous ¡period: ¡1 ¡ • Otherwise: ¡0 ¡

Independent Variables Enterprise ¡Diversity : ¡Herfindahl ¡Index ¡of ¡enterprise ¡revenue ¡(from ¡crop, ¡beef, ¡hogs, ¡etc) ¡ Opera/ng ¡profit ¡margin : ¡$ ¡of ¡net ¡income ¡per ¡$ ¡of ¡revenue ¡ Opera/ng ¡expense ¡ra/o : ¡$ ¡of ¡expense ¡per ¡$ ¡of ¡revenue ¡ Interest ¡expense: ¡ (in ¡100,000s) ¡ BRM ¡payments: ¡ (in ¡100,000s) ¡ based ¡on ¡year ¡they ¡received ¡money ¡ Partnership ¡ (dummy): ¡if ¡farm ¡has ¡more ¡than ¡1 ¡operator: ¡1, ¡otherwise ¡0 ¡ Size ¡category ¡ (dummies) ¡ Field ¡Crop ¡Farms ¡ Dairy ¡Farms ¡ by ¡dollar ¡of ¡sales ¡ $0-­‑$10k ¡ $0k-­‑$250k ¡ $10k-­‑$100k ¡ $250k-­‑$500k ¡ $100k-­‑$250k ¡ +$500k ¡ $250k-­‑$500k ¡ +$500k ¡

Regression results !! ! Field!Crop! !! Dairy! ! Logit! !! Probit! Logit! !! Probit! ! ! ! Independent!Variables! Fixed!Effect! Random!Effect! Random!Effect! Fixed!Effect! Random!Effect! Random!Effect! Enterprise!Diversity! ! NS! NS! ! NS! ! NS! 1.295! ! 0.804! Operating!Profit!Margin! ! NS! <0.038! ! <0.022! ! >4.475! <1.066! ! <0.651! Operating!Expense!Ratio! ! NS! NS! ! NS! ! >3.993! NS! ! NS! BRM!payments! ! >0.316! <0.249! ! <0.149! ! NS! NS! ! NS! Interest!Expense! ! >0.476! 0.228! ! 0.136! ! >0.896! NS! ! NS! Partnership!dummy! ! NS! NS! ! NS! ! NS! NS! ! NS! Size!2! ! NS! NS! ! NS! ! >0.975! NS! ! NS! Size!3! ! NS! 0.678! ! 0.417! ! NS! NS! ! NS! Size!4! ! NS! 0.831! ! 0.510! ! >! >! ! >! Size!5! ! NS! 0.859! ! 0.528! ! >! >! ! >! ! ! ! CONSTANT! ! >! NS! NS! >! NS! NS! ! ! !

Main findings • Program payments may crowd out farm’s risk balancing strategy – Farms that received more payments less likely to risk balance – BUT, may be sector-specific • Other factors that influence risk balancing behaviour also seems to be sector specific – Interest expense influence behaviour for field crop operations but not for dairy – Larger field crop farms more likely to risk balance, but size have no effect for dairy operations

Next steps & Challenges Next Steps: • Extend analysis to other sectors (beef) Challenges: • OFID is detailed in tax & production data but does not capture balance sheet information – Especially important in capturing value of asset (land, quota)

Thank You

Appendix – Field Crop Results Dependent'Variable:' riskbal ' Independent'Variable' Fixed'Effects'Logit' Random'Effects'Logit' Random'Effects'Probit' enterprdiv. 0.277'' C0.232'' C0.142'' (0.297) ' (0.151) ' (0.092) ' opprofmrgn. C0.05'' C0.038*'' C0.022*' (0.026) ' (0.014) ' (0.008) ' opexpratio. 0.009'' C0.001'' 0.000'' (0.011) ' (0.005) ' (0.003) ' govpay. C0.316*'' C0.249*' C0.149'' (0.070) ' '(0.06) ' (0.036) ' interestexp. C0.476*' 0.228*' 0.136*'' (0.231) ' (0.089) ' (0.053) ' partnership. C0.74'' 0.010'' 0.006'' (0.592) ' (0.052) ' (0.031) ' size2. 0.002'' 0.246'' 0.153'' (0.207) ' (0.148) ' (0.091) ' size3. 0.139'' 0.678*'' 0.417*'' (0.222) ' (0.152) ' (0.093) ' size4. 0.206'' 0.831*'' 0.510*' (0.243) ' (0.160) ' (0.097) ' size5. 0.267'' 0.859*'' 0.528*'' (0.279) ' (0.171)' (0.104) ' constant. C0.264'' C0.165'' C' (0.206) ' (0.126) ' ' Number'of'observations' 2,912' 3,860' 2,522' LogClikelihood'value' C'4,492.38' C10403.596' C1,708.92' Likelihood'ratio/Wald'chi 2' ' ' ' C value' 43.48' 181.96'' 26.71' C pCvalue' 0.000' 0.000' 0.005' Rho'value' C' .174' .192' ' ' (.011)' (.032)' Likelihood'ratio'test'of'rho=0' ' ' ' chi 2 'value' C C' 338.80'' 42.19' C pCvalue' C' 0.000' 0.000' Notes:'*'denotes'statistical'significance'at'the'5%'level;'i'–'948'observations'dropped'because'of'all' positive'or'all'negative'outcomes'

Appendix – Dairy Results Dependent'Variable:' riskbal ' Independent'Variable' Fixed'Effects'Logit' Random'Effects'Logit' Random'Effects'Probit' enterprdiv. 1.706' 1.295*' 0.804*' (1.055)' (0.472)' (0.292)' opprofmrgn. I4.475*' I1.066*' I0.651*' (1.287)' '(0.516)' (0.309)' opexpratio. I3.993*' I0.317'' I0.203' (.153)' (0.502)' (0.305)' govpay. 0.00740' I0.888'' I0.553' (.861)' (0.714)' (0.441)' interestexp. I.896*' I0.122'' I0.076' (.434)' (0.102)' (0.063)' partnership. .395' I0.194'' I0.121' (1.447)' (0.149)' (0.093)' size2. I.975*' I0.0503'' I0.033' (.498)' (0.185)' (0.115)' size3. .250' 0.259'' 0.159' (.696)' (0.207)' (0.128)' constant. I' I0.0906'' I0.0518' (0.562)' (0.344)' i ' Number'of'observations' 204 236' 236' LogIlikelihood'value' I301.00' I642.00' I641.98' 2' Likelihood'ratio/Wald'chi ' ' ' I value' 30.94' 17.42' 18.00' I pIvalue' 0.0001' 0.0261' 0.0212' Rho'value' I' .017' .021' ' ' (.035)' (.044)' Likelihood'ratio'test'of'rho=0' ' ' ' 2 'value' I chi I' .25' .25' I pIvalue' I' 0.310' 0.311' Notes:'*'denotes'statistical'significance'at'the'5%'level;'i'–'32'observations'dropped'because'of'all'positive' or'all'negative'outcomes !