1 CHAPTER 1: INTRODUCTION 1.1 Research rationale: Vietnam is the second largest coffee producer with an average volume about 1.3 million tons coffee per year Its areas under crop for producing coffee is approximately 641.7 thousand hectares Most of them are used for export in the form of raw materials Vietnam exported about 1.906 million tons of coffee, with the volume sale of $ 3, 5569 billion in 2014 Although coffee industry is highly effective, it also faces with a lot of risks from the uncertainty factors In addition, coffee producers and sellers lack of competence in risk management Thus, it is essential to a research in the field of risk management in coffee industry The research “The impact of risk factors to production and businesss performance in Vietnam” is carried out to satisfy this demand 1.2 Research objectives: This research is designed to achieve three objectives Firstly, it tries to identify the risk factors, influencing the production and trading of coffee, discover the gap in risk management theory from literature review, and suggest conceptual framework Secondly, the influence of risk factors, measured by appropriate scales, on producing and trading coffee is identified Thirdly, risk management solutions are built/suggested to reduce the impacts of risk factors on coffee industry Fourthly, it aims to provide some complementary knowledge in risk theories 1.3 Research questions: The research aims to address four questions Firstly, what risk factors influence producing and tradingcoffee in Vietnam? And what theoretical gaps, related to risks, address in this research Secondly, what are relationships between risk factors and production and business performance in the context of Vietnam? Thirdly, what are risk management solutions appropriate for coffee producing and selling in Vietnam? Fourthly, How the research findings contribute to the risk management theories? 1.4 Research methodology: Qualitative and quantitative approach are combined in doing this research 1.5 Research object and scope: 1.5.1 Research object: Risk factors influencing coffee production and business performance in Vietnam 1.5.2 Scope of research: Measurement scales and conceptual framework, which present the relationships between risk factors and coffee production and business performance in the period from 2010 to 2015, are built based on the literature review of previous related research and field survey 1.6 Research contributions: (1) Building new measurement scales for risk factors, influenced to production and business performance (2) Assessing all types of risks associated to coffee producing and selling process in detail, thorough, and comprehensive (3) highlighting the two faces of risk, which are chance of losses and gains (4); Adding the new theoretical perspective to risk and risk management theories CHAPTER 2: LITERATURE REVIEW AND CONCEPTUAL FRAMEWORK 2.1 Risk 2.1.1 Traditional approach (negative approach): (1) Hoàng Phê (1995) said that “Risk is bad, not good, and sudden occurrence”; (2) Nguyễn Lân (1998) considered “Risk is unlucky occurrence”; (3) Oxford dictionary defines “the possibility of something bad happening at some time in the future; a situation that could be dangerous or have a bad result”; (4) Hồ Diệu (2002) described “Risk is measured by asset losses or less actual profit than expected profit” 2.1.2 Hedging approach: (1) Frank Knight (1921, page 233): “Risk is measured uncertainty”; (2) Allan Herbert Willett (1951, page 6): “Risk is exposure to uncertainty related to unexpected occurrences”; (3) C Arthur William et al (1964): “Risk is potential deviation in outcomes When risk occurs, people are not able to predict exactly outcome The existence of risk causes uncertainty Risk occurs when any people’s action leads to unpredictable gain or losss possibility; (4) David Apgar (2006): “Risk is any uncertainty outcome which affects actual results and make them different from expected results” 2.1.3 Other definitions: (1) Vân, Đ.T.H et al., (2013, 32): “Risk is measured uncertainty, when managed well, people can get opportunities; on contrary, when managed bad losses will occur”; (2) Aswath Damodaran (2010, p 86 t1): “Risk refers to likelihood to get actual investment rate of return which is different from expected investment rate of return Thus, risk refers to not only bad outcome, when actual is lower than expected investment rate of return, but also good outcome, when actual is higher than expected investment rate of return In reality, we can say that risk is associated a change of gains when it bring good outcome and a change of losses if it accompanies to bad outcome We should consider both of them when assessing risk” (3) Ngô Quang Huân N.Q et al (1998, page 8): “Risk is potential variation in outcomes, the lager number of type of outcomes and the greater variation among them, the greater level of risk Risk is objective concept and measured” 2.1.4 Suggested concept in this research: Risk is occurrence at a given probability and makes variations in outcome of occurrence, and makes a difference from expected or predicted results Otherwise, risk existence may cause unpredicted losses Duplicity of risk concept involves a chance of losses and gains After all, risk is an objective phenomenon, happened out of people’s intention, but it is identified, measured, and controlled Furthermore, human being may be able to transform from a chance of losses to a chance of gains 2.2 Losses: Tuấn, A.N (2006, page 21) defined losses as “damage and lost in term of asset, opportunities, physical and mental human well-being, people’s health, and career development are caused by risks” 2.3 The relationship between potential variations in production and business performance and risks 2.3.1 The concept of variation in production performance: It refers to changes in actual production performance in comparison to expected production performance These changes are invisible and caused by risk factors 2.3.2 The concept of variation in business performance Variations in business performance refer to difference between actual and expected business performance, caused by risk factors 2.4 Risk management: It is defined as “scientific, continuous, comprehensive, and systematic approach process to risks This process is used to identify, control, prevent, reduce damage, losses, and negative effects of risks, and make them to be opportunities” (Vân, D.T.H, et al., 2013, page 66) The suggested definition of risk management in this research is “Risk management is scientific, continuous, comprehensive, and systematic approach process to risks in order to identify, control, prevent, reduce damage, losses, and negative effects of risks based on predicted probability of risk occurrence, and make disadvantage to be advantage risk 2.5 Literature review related to research Table 2.1: Previous researches in foreign countries No RISK FACTORS Previous researches WB (2004) PRODCUCTION Market price, weather, 01 working pest, BUSINESS capital, Market price, international financial production market, working capital, business imbalance pattern, psychological and behavior businessman UNCTAD/WT Market O (2002) international coffee roasted suppliers, 02 price, international business speculators, pattern, market information, working capital, and society Quoc 03 Luong Market price, psychology and and Loren W behavior of producers Tauer (2004) ICO (2014) 04 World coffee trade 1963- Market imbalance , society Thinh Hoang Si production Market price, business pattern, international speculators, international coffee roasted suppliers, international finance markets, and society 2013 05 price, Market price & Huong Nguyen Thi (2015) No RISK FACTORS Previous researches PRODCUCTION WB (2015 BUSINESS Weather, pest Working capital, Market price, international payment 06 political mechanism currencies, foreign exchange rate, working capital 07 Bunn, Christian Weather (M.Sc.) (2015) Table 2.2: Previous researches in Vietnam No 01 RISK FACTORS Previous Researches PRODUCTION BUSINESS Geography and National Resource Center, Thời tiết Vietnam Institute of Science and Technology (1987) 02 Nghị, N.S at al., (1996) Pest and production process Gia, T.B and Market Cường, 03 (2005) price, H.T production political psychology producers working Finance Markets, technology, producers’ psychology and behavior process, institutions, and capital, International society, behavior of No RISK FACTORS Previous Researches PRODUCTION BUSINESS Chi, T.Q T Market price, political institutions, Market price, international payment (2007) 04 society, working capital, production currencies and exchange imbalance political institutions, working capital, rate, society, international finance markets, psychology and behavior of businessmen 05 Tran, (2011) N.T.N Market price 2.6 Conceptual framework and hypotheses: The impact of risk factors to production and business performance in Vietnam 2.6.1 Conceptual framework and hypotheses: The impact of risk factors to production and business performance in Vietnam 2.6.1.1 Conceptual framework and hypotheses: The impact of risk factors to production performance in Vietnam FACTOR DETERMINANTS OF RISK’S IMPACT PRODUCTION PERFORMANCE TO RODUCTION AND BUSINESS PERFORMANCE GTT KTSX CN THT TTSX SDB MCDSX BDKQSX VSX HVNSX IMPACT TO PRODUCTION PERFORMANC-E RISKS’ MPACT TO PRODUCTIO N PERFORMA-NCE FACTOR DETERMINANTS OF BUSINESS PERFORMANCE RISK’S GTT KTKD TTTT QDCQT NRX TTTC DTTTG VKD IMPACT TO BUSINESS IMPACT TO BUSINESS PERFORMAN-CE PERFORMANCE TTKD BDKQKD HVNKD 2-1 Conceptual framework: The impact of risk factors on production and business performance (Source: Văn, L.B) - Potential deviation model in production performance is determined by following multiregression: BDKQSX = β1 + β2GTT + β3KTSX+ β4CN+ β5THT+ β6SDB + β7VSX + β8MCDSX + β9TCCTSX+ β10XHSX+ β11HVNSX + u + Dependent variable is: “Potential deviation in production performance” and denotes as BDKQSX + Independent variables include “Market price” (GTT); “Production process” (KTSX); “Production technology” (CN); “Weather” (THT); “Pest” (SDB); “Working capital” (VSX); “Production imbalance” (MCDSX); “The impact of political institutions on production” (TCCTSX); “Social impact on production” (XHSX); “Producers’ psychology and behavior” (HVNSX) - Loss model in production is defined as follows: TTSX = γ1 + γ2GTT' + γ3KTSX'+ γ4CN'+ γ5THT'+ γ6SDB'+ γ7VSX'+ γ8MCDSX'+ γ9TCCTSX'+ γ10XHSX'+ γ11HVNSX' + u' + Dependent variable is: “Loss in production” and denotes as TTSX + Independent variables are defined as “Market price” (GTT'); “Production process” (KTSX'); “Production Technology” (CN'); “Weather” (THT'); “Pest” (SDB'); “Working capital” (VSX'); “Production imbalance” (MCDSX'); “The impact of political institutions on production” (TCCTSX'); “Social impact on production” (XHSX'); “Producers’ psychology and behavior” (HVNSX') - Hypotheses: There are nine hypotheses, ranged from HSX1 đến HSX9, in potential deviation model of production performance Loss model in production has nine hypotheses, ranged from H'SX1 đến H'SX9 2.6.1.2 Conceptual framework used to examine the impact of risk factors on business performance and derive hypotheses: - Potential deviation model in business performance is as follows: BDKQKD = β'1+ β'2GTT+ β'3KTKD+ β'4TTTT+ β'5TTTC+ β'6VKD+ β'7DTTTG+ β'8QDCQT+ β'9NRX+ β'10TCCTSX+ β'11XHSX ++ β'12HVNKD+ u + Dependent variable is "Potential deviation in business performance”, denoted as BDKQKD + Independent variables include "Market price” (GTT); "Business process” (KTKD); "Market information” (TTTT); "International finance markets” " (TTTC); "Working capital” (VKD); "International payment currencies and foreign exchange rate” (DTTTG); "Speculate Funds” (QDCQT); "International roasted coffee-nut producers” (NRX), “The impact of political institutions on production” (TCCTKD); “Social impact on production” (XHKD); (HVNKD) " Producers’ psychology and behavior " - Loss model in business performance is defined as follows: TTKD = γ'1 + γ '2GTT' + γ'3KTKD'+ γ'4TTTT' + γ'5TTTC'+ γ'6VKD'+ γ'7DTTTG' + γ'8QDCQT' + γ'9NRX' + γ'10TCCTSX'+ γ'11XHSX' + γ'12HVNKD' + u' + Dependent variable is: "Loss in business performance” and denoted as TTKD + Independent variables are "Market price” (GTT'); "Business process” (KTKD'); "Market information” (TTTT'); "International finance markets” (TTTC'); "Working capital” (VKD'); "International payment currencies and foreign exchange rate” (DTTTG'); "Speculate Funds” (QDCQT'); " International roasted coffee-nut producers” (NRX'); “The impact of political institutions on production” (TCCTKD'); “Social impact on production “(XHKD'); " Producers’ psychology and behavior” (HVNKD') - Hypotheses: Potential deviation model in business performance has ten hypotheses, ranged from HKD1 đến HKD10; and loss model in business performance includes also ten hypotheses, ranged from H'KD1 đến H'KD10 CHAPTER 3: RESEARCH METHODOLOGY 3.1 Delphi method Information is collected from experts through the following process: (1) determining the research’s objectives; (2) selecting relevant experts (3) designing questions and sending them to experts; (4) summarizing the experts ‘response and writing the executive report; (5) sending executive report to experts to get comments; (6) Receiving the new revising comments from experts; (7) doing the third step until making a consensus agreement 3.2 Scale validity assessment Assess the scale validity for each latent variable by calculating cronbach alpha The acceptable level of cronbach’s alpha ranges from 0.6 to 0.9 3.3 Assess and revise measurement scale by EFA Carrying out exploratory factor analysis and testing necessary requirement of EFA satisfy two criteria: KMO  50%; and Sig  5% 3.4 Correlation analysis Testing collinearity relationship among independent variables: identify a statistically significant high correlation coefficient at the level of 5% 3.5 Multi regression analysis: Carry out multiregression analysis between dependent variable and independent variables, discovered from EFA 3.6 Sample design 3.6.1 Sampling method: Stratification sampling is used to satisfy a presentativeness of respondents in term of area, size, and type of business 3.6.2 Sample size.: According to Hoelter (1983), a minimum sample size is 200 respondents The sampling process is this research is as follows: 3.6.1.1 Sample size requirement in building measurement scale (Delphi method) Sample size in production sector: n =10; sample size in business sector: n = 10; 3.6.1.2 Sample size in preliminary research - Sample size in preliminary research is determined by qualitative approach as follows:: (1) Production sector: n = 50 (Group discussion: n = 10; Questionnaire survey: n = 40); (2) Business sector: n = 50 (Group discussion: n = 10; Questionnaire: n = 40) - Sample size in preliminary research is determined by quantitative approach is as follows: (1) Production sector: n = 100; (2) Business sector: n = 100 3.6.1.3 Sample size in formal research (quantitative approach): Production sector: n = 200; Business sector: n = 200 CHƯƠNG 4: RESEARCH RESULTS AND DISCUSSIONS 4.1 Research results: 4.1.1 Measurement scale design and testing: 4.1.1.1 Measurement scale assessing by Delphi method: - Production sector: Questionnaires designed based on eleven definition of risk (mention in part 2.1) were sent to ten specialists After rounds discussions and responses, all of them make consensus agreement in defining production risk in terms of potential deviations in production performance and production losses - Business sector: Questionnaires designed based on eleven definition of risk (mention in part 2.1) were sent to ten specialists After rounds of discussions and responses, all of them make consensus agreement in defining production risk in terms of potential deviations in business performance and business performance losses 4.1.1.2 Measurement scale assessment in preliminary research: - Preliminary research with qualitative approach: + Coffee production: (1) Potential deviations in production performance: rate of agreement for each risk factor is as follows: BDKQSX: 96%; GTT: 90%; THT: 80 %; SDB: 82%; KTSX: 76%; CN: 78 %; MCDSX: 72 %; VSX: 84 %; HVNSX: 70 %; TCCTSX: 32 %; XHSX: 36% (2) Losses in production performance: rate of agreement for each risk factor is as follows: TTSX: 88%; GTT: 92%; THT: 82%; SDB: 80%; KTSX: 74%; CN: 70%; MCDSX: 68%; VSX: 84%; HVNSX: 68%; TCCTSX: 28%; XHSX: 30% 10 + Coffee Business in Vietnam: (1) Deviation in business performance: agreement rate for each risk factor is as follows: BDKQKD: 92%; GTT: 98%; KTKD: 80%; QDCQT: 90%; NRX: 72%; TTTT: 84%; DTTTG: 74%; TTTC: 76%; VKD: 84%; HVNKD: 70%; TCCTKD: 36%; XHKD: 34% (2) Losses in business performance: agreement rate for each risk factor is as follows: TTKD: 80%; GTT: 92%; KTKD: 76%; QDCQT: 90 %; NRX: 72%; TTTT: 88%; DTTTG: 74%; TTTC: 76%; VKD: 86%; HVNKD: 70%; TCCTKD: 26%; XHKD: 28% - Preliminary research with quantitative approach: + Production sector * Potential deviation in production performance: ● Measurement scale reliability by cronbach’s alpha coefficient: Two of the total eleven variables, removed from the model after reliability analysis, are TCCTSX, and XHSX ● Results from Exploratory factor analysis: (1) Independent factor: KMO coefficient is 0.734; P-value is 0.00; and total variance explained by the factor is 77.86% Thus, data is fitted to the research model (2) Independent factor: Results from EFA for eight factor are as follows: KMO = 0,675; P-value = 0,00; Total variance explain by factor is 75.544% Data is fitted to the research model * Losses in production performance in preliminary research: ● Scale Reliability assessment by Cronbach's alpha: Two of the total eleven variables, removed from the research model after reliability analysis, are “TCCTSX” and “XHSX” ●Results from exploratory factor analysis ( EFA) (1) Dependent factor: KMO coefficient is 0.734; p-value is less than critical value (0.00 in comparison to 0.05; and total variance explained by this factor is 81.697% Data is fitted to the research model (2) Independent factor: Results from EFA for factors are as follows: KMO = 0.709; p-value = 0,00; and total variance explained =77.096% Data is fitted to the model * Potential deviation in business performance (Preliminary research): ● Scale reliability assessment by Cronbach's alpha: Two of the total eleven variables, removed from the research model after reliability analysis, are TCCTSX and XHSX ● Results from Exploratory factor analysis in preliminary research: (1) Dependent factor: KMO = 0.45; p-value = 0.00; total variance explained by factors = 80.56%; data is fitted to the research model (2) Independent factor: results from exploratory factor analysis for factors are as follows: KMO 11 = 0.666; P-value (Sig) = 0.00; total variance explained by factors = 76.95% Data is fiited? to research model * Losses in business performance (preliminary research): ● Scale reliability assessment by Cronbach's alpha: Two of the total twelve variables, removed from the research model after reliability analysis, are “TCCTSX” and “XHSX” ● Results from exploratory factor analysis ( EFA) (1) Dependent factor: KMO coefficient is 0.742; P-value (Sig) is less than critical value (0.00 in comparison to 0.05); total variance explained by this factor is 80.05% We can conclude that data is fitted to research model (2) Independent factors: Results from EFA for factors are as follows: KMO coefficient is 0.649; p-value (sig) is less than critical value (0.00 in comparison to 0.05); and total variance explained by these factors are 76.932% Data is fitted to the research model 4.1.1.3 Carry out formal research by using quantitative approach - Formal research in coffee production sector: + Potential deviation in production performance: * Scale reliability assessment by Cronbach's alpha : All nine factors satisfy requirement reliability Thus, nothing is removed * Results from exploratory factor analysis: ● Dependent factor: KMO coefficient is 0.744; P-value (Sig) is 0.00; and total variance explained by this factor is equal to 81.581% We can conclude that data is fitted to research model ● Independent factors: Results from exploratory factor analysis for eight factors as follows: KMO coefficient is 0.75; P-value (sig) is 0.00; and total variance explained for these factors is 75.25% Data is fitted to the research model * Correlation coefficient analysis among independent variable: correlation coefficients between independent variable are low and medium at the statistically significance of 5% so there is not collinearity * Multiple regression analysis for relationship between dependent variable (Potential deviation of production performance) and independent variables: The adjusted determinant coefficient is 523; F-statistic value is 28.274.; all variance inflation factors are less than 2; and pvalues for t-statistics are significant at the value of % These results allow us to conclude that all independent variable have impact on the potential deviation of production performance and variation of all dependent variable explains 52.3% of total variance of dependent variable (See table 4.1) 12 Table 4.1: Multiple regression coefficients and collinearity test Unstandardized Standardized coefficients coefficients Model Beta Standard error Collinearity tstatistic Beta Tolerance -.941 358 KTSX 114 043 136 CN 133 057 135 GTT 142 060 122 2.387 053 130 (Constant) THT 126 statistic Sig VIF -2.626 009 2.681 008 933 1.072 2.343 020 725 1.380 018 918 1.089 2.362 019 787 1.271 SDB 211 048 242 4.355 000 776 1.289 VSX 241 051 251 4.754 000 859 1.164 MCDSX 178 053 172 3.356 001 916 1.092 HVNSX 186 043 217 4.320 000 949 1.054 (Note: dependent variable is potential deviation of production performance; source: calculated by author) Multiple regression model: BDKQSX = 0,136 KTSX + 0,135 CN + 0,122 GSX + 0,130 THT + 0,242 SDB + 0,251 VSX + 0,172 MCDSX + 0,217 HVNSX + Losses in production performance: * Scale reliability testing by Cronbach's alpha: Result from scale reliability analysis for nine factors (one dependent factor and independent factors) shows that these factors are satisfied reliability requirement * Results from exploratory factor analysis (EFA): ● Exploratory factor analysis for dependent factor: KMO coefficient is 0.748; p-value (sig) is 0.00; and total variance explained by this factor is 82.609% Thus, data is fitted to research model 13 ● Results from exploratory factor analysis for eight independent factors are as follows: KMO coefficient is 0.740; p-value (sig) is 0.00; and total variance explained by these factors is 75.765% Data is fitted to the research model * Correlation analysis: there is statistically significant relationship between independent variable and dependent variables at the level of 5% In other hand, correlation coefficient among independent variable is significant low so there is not collinearity among them * Results from regression analysis: The adjusted R2 is 0.586; F-statistics is equal 36.166; and all variance inflation factors of dependent variables are less than Thus, the model good of fitness is acceptable (See table 4.2) In addition all p-values of independent variables is statistically significant at the level of 5% so we can conclude that these independent variables have influence on dependent variable (losses in production performance) Table 4.2: Multiple regression coefficients and collinearity test Unstandardized Standardized coefficients coefficients Model Beta - (Constant) 1,368 Standard error Collinearity tstatistic statistic Sig Beta ,335 Tolerance VIF -4,085 ,000 KTSX ,096 ,041 ,112 2,349 ,020 ,911 1,098 CN ,168 ,054 ,169 3,121 ,002 ,713 1,403 GTT ,185 ,058 ,155 3,199 ,002 ,891 1,122 THT ,110 ,052 ,113 2,139 ,034 ,751 1,332 SDB ,216 ,047 ,236 4,583 ,000 ,782 1,278 VSX ,245 ,049 ,246 ,000 ,855 1,170 MCDSX ,216 ,054 ,194 4,041 ,000 ,903 1,107 HVNSX ,208 ,042 ,233 4,919 ,000 ,927 1,079 4,993 (Note: dependent variable is losses in production performance; calculated by author) The multiple regression model may be expressed as follows: TTSX = f(KTSX,CN,GTT, THT, SDB, VSX, MCDSX, HVNSX) Result from regression analysis is: TTSX = 0,122 KTSX + 0,169 CN + 0,155 GTT + 0,113 THT + 0,236 SDB + 0,246 VSX + 0,194 MCDSX + 0,233 HVNSX 14 - Formal research in coffee business sector: + Potential deviation in business production performance: * Scale reliability assessment by Cronbach's alpha: All ten factors are satisfied the reliability requirement because their conbach’s alpha coefficient are in the range between 0.6 and 0.9 * Results from exploratory factor analysis: ● Exploratory factor analysis for dependent factor: KMO coefficient is 0.777; p-value (sig) is 0.00; and total variance explained by this factor is 74 980% Thus, data is fitted to research model ● Results from exploratory factor analysis for nine independent factor are as follows: KMO coefficient is 0.776; p-value (sig) is 0.00; and total variance explained by these factors is 74.798 Data is fitted to the research model * Correlation coefficient analysis among independent variable: There is high correlation coefficient between dependent variable and independent variable and correlation coefficients between independent variable are low and medium at the statistically significance of 5% So there is not collinearity among independent variables * Multiple regression analysis for relationship between dependent variable (Potential deviation of business performance) and independent variables: The adjusted determinant coefficient is 644; F-statistic value is 38.126.; all variance inflation factors are less than 2; and p-values for tstatistics are significant at the value of % These results allow us to conclude that all independent variable have impact on the potential deviation of business performance and variation of all dependent variable explains 64.4% of total variance of dependent variable (See table 4.3) 15 Table 4.3: Multiple regression coefficients and collinearity test Unstandardized Standardized coefficients coefficients Mode l Beta Standard error -1.644 328 130 042 TTTT 161 TTTC VKD 1(Con Collinearity statistic t-statistic Sig Beta Tolerance VIF -5.010 000 144 3.137 002 894 1.119 053 152 3.019 003 736 1.359 133 054 112 2.471 014 914 1.094 119 050 121 2.398 017 737 1.357 135 056 131 2.391 018 623 1.606 210 044 238 4.824 000 769 1.300 NRX 214 048 220 4.431 000 758 1.320 GTT 212 047 203 4.487 000 913 1.095 173 039 199 4.422 000 930 1.075 stant) KTK D DTTT G QDC QT HVN KD (Note: Dependent variable is BDKQKD; Source: Calculated by author) The multiple regression model may be express as follows: BDKQKD=f(KTKD,TTTT,TTTC,VKD,DTTTG,QDCQT,GTT,HVNKD) Result from multiple regression analysis is as follows: BDKQKD = 0,144 KTKD + 0,152 TTTT + 0,112 TTTC + 0,121 VKD + 0,131 DTTTG + 0,238 QDCQT + 0,220 NRX + 0,203 GTT + 0,199 HVNKD 16 + Losses in business performance (formal research): * Scale reliability testing by Cronbach's alpha: Result from scale reliability analysis for ten factors (one dependent factor and independent factors) indicates that these factors are satisfied reliability requirement * Results from exploratory factor analysis (EFA): ● Exploratory factor analysis for dependent factor: KMO coefficient is 0.738; p-value (sig) is 0.00; and total variance explained by this factor is 79 605% Thus, data is fitted to research model ● Exploratory factor analysis for nine independent factors: KMO coefficient is 0.758; p-value (sig) is 0.00; and total variance explained by this factor is 75.2785% Thus, data is fitted to research model * Correlation coefficient analysis: There is high correlation coefficient between dependent variable and independent variable and correlation coefficients between independent variable are low and medium at the statistically significance of 5% So there is not collinearity among independent variables * Multiple regression analysis for relationship between dependent variable (Potential deviation of business performance) and independent variables: The adjusted determinant coefficient is 613; F-statistic value is 36.053.; all variance inflation factors are less than 2; and p-values for tstatistics are significant at the value of % These results allow us to conclude that all independent variable have impact on the potential deviation of business performance and variation of all dependent variable explains 61.3% of total variance of dependent variable (See table 4.4) 17 Table 4.4: Multiple regression coefficients and collinearity test Unstandardized Standardized coefficients coefficients Model t-statistic Beta 1(Consta Collinearity statistic Standard Beta error -1.486 330 KTKD 147 041 TTTT 141 TTTC Sig Tolerance VIF -4.505 000 167 3.582 000 899 1.112 054 133 2.617 010 748 1.337 146 054 123 2.698 008 935 1.070 VKD 110 050 113 2.207 029 738 1.356 DTTTG 151 056 148 2.713 007 654 1.528 QDCQT 204 043 240 4.763 000 768 1.302 NRX 171 048 178 3.583 000 789 1.267 nt) GTT 188 043 199 4.322 000 918 1.089 HVNKD 200 038 237 5.220 000 940 1.064 (Note: Dependent variable is TTKD; Source: Calculated by author) The multiple regression model may be express as: TTKD = f(KTKD,TTTT,TTTC,VKD,DTTTG,QDCQT,GTT,HVNKD) Result from multiple regression is as follows: TTKD = 0,167 KTKD + 0,133 TTTT + 0,123 TTTC + 0,113 VKD + 0,148 DTTTG + 0,240 QDCQT + 0,178 NRX + 0,199 GTT + 0,237 HVNKD 4.1.2 Results from testing research framework: 4.1.2.1 Testing conceptual framework: - Conceptual framework of risk’s impact on coffee production: 18 + Conceptual framework related to potential deviation of production performance: There are nine research concepts, after removed two research concepts (one dependent concept and eight independent concepts) + Conceptual framework related to losses in production performance.: There are nine research concepts, after removed two research concepts (one dependent concept and eight independent concepts) - Conceptual framework of risk’s impact on coffee business: + Conceptual framework related to potential deviation of business performance: There are ten research concepts, after removed two research concepts (one dependent concept and nine independent concepts) + Conceptual framework related to losses in production performance: There are ten research concepts, after removed two research concepts (one dependent concept and nine independent concepts) 4.1.2.2 Testing hypotheses - Conceptual framework related to potential deviation of production performance: Nine hypotheses, ranged from Hsx1 to Hsx9, are statically accepted at the significant level of 5% - Conceptual framework related to losses in production performance: Nine hypotheses, ranged from H'sx1 to H'sx9 are statically accepted at the significant level of 5% - Conceptual framework related to potential deviation of business performance: Ten hypotheses, ranged from Hkd1 cho đến Hkd10, are statically accepted at the significant level of 5% - Conceptual framework related to losses in production performance: Ten hypotheses, ranged from H'kd1 cho đến H'kd10, are statically accepted at the significant level of 5% 4.1.3 Managerial implication: 4.1.3.1 The risk factors, impacted on production process, consist of: (1) fluctuations of market price (2) production process (3) technology ( both of technical production process and postharvest technology); (4) weather; (5) pest; (6) working capital; (7) production imbalance; (8) producers’ psychology and behavior 4.2.3.2 The risk factors, impacted on o business process, consists of: (1) fluctuations of market price; (2) business process; (3) market information.; (4) international finance markets; (5) working capital; (6) international payment currencies and foreign exchange rate; (7) international speculate funds; (8) international roasted coffee-nut producer; (9) psychology and behavior of businessmen 4.2 Assessment of practices in coffee production and business: This assessment based on risk factors highlight their impact on Vietnam’s coffee production and business in practices CHAPTER 5: RISK MANAGEMENT SOLUTIONs 19 5.1 Macro solutions: 5.1.1 Macro solutions in coffee production : 5.1.1.1 Solutions to mitigate the impacts of market price fluctuations 5.1.1.2 Solutions to mitigate the impacts of production process 5.1.1.3 Solutions to mitigate the impacts of production technology 5.1.1.4 Solutions to mitigate the impacts of weather 5.1.1.5 Solutions to mitigate the impacts of pest 5.1.1.6 Solutions to mitigate the impacts of working capital 5.1.1.7 Solutions to mitigate the impacts of production imbalance 5.1.1.8 Solutions to mitigate the impacts of producers’ psychology and behavior 5.1.2 Micro solutions in coffee business 5.1.2.1 Solutions to mitigate the impacts of market price fluctuations 5.1.2.2 Solutions to mitigate the impacts of business process 5.1.2.3 Solutions to mitigate the impacts of market information 5.1.2.4 Solutions to mitigate the impacts of international finance markets 5.1.2.5 Solutions to mitigate the impacts of working capital 5.1.2.6 Solutions to mitigate the impacts of international payment currencies and foreign exchange rate 5.1.2.7 Solutions to mitigate the impacts of international speculate funds 5.1.2.8 Solutions to mitigate the impacts of international roasted coffee-nut producers 5.1.2.9 Solutions to mitigate the impacts of producers’ psychology and behavior 5.2 Macro suggestions: (1) Government should pay more intention on building and developing the coffee future market.; (2) Government should make a long-term master plan regarding to coffee production area and processing to achieve the sustainable development of this industry; (3) Government supports providers’ services, related to coffee production business techniques, to producers and businessmen (4) Government has to give incentives to investors in building and developing production and postharvest production process; (5) Government provide guidelines for development of market information system to coffee producers and businessmen; (6) Developing training programs, under the government’s support, is helpful for coffee business; (7) Government focuses on supporting and giving opportunities for the development of Vietnam’s coffee association 20 5.3 COMBINED SOLUTIONS: (1) Establish the close relationship among government, scientists, credit providers, producers, and businessmen to support sustainable development.; (2) Carry out quality management throughout from production to distribution stage CONCLUSION Findings and contributions + Findings: Determine risk factors which impact on coffee production and business performance in Vietnam + Measurement model: Measurement model take an effect in scanning and removing factors which have no impact on coffee production and business process + Conceptual framework: Conceptual framework provides a theoretical supplement to risk theories in explaining the impact of risk factors on coffee production and business process Limitation: The lack of respondents ‘perceived measurement scale is limitation of this research