So that studying the impact of factors on the bad debt market in Vietnam is an appropriate and necessary research direction, both theoretical and practical

ces thereby determining the prices and output of goods and services on the market (Hal R.Varian, 1999). The market includes all existing and potential customers, who have the same specific needs or desires, able and willing to participate in the exchange to satisfy needs or that human desire (Carlos A. Diaz Ruiz, 2012). Thus, in nature the market is formed by the need to exchange products and services between buyers and sellers. 1.2.2 The theory of market entry behavior According to Sven Ove Hansson (2005), Mark Perlman & E. Roy Weintraub (2003), decision making can be considered a problem-solving activity with a solution that is optimal, or at least satisfactory. Additionally, Czinkota, et al., (2009) have refered that market participation means the participation of enterprises in the process of trading and buying - selling certain goods and services on the market. Inside, market participation is a complex process, including analysis and synthesis of internal and external factors of the enterprise (Pehrsson; 2008a), (Pehrsson, 2008b), which are employed to the strategy and choose the mode of market participation (Gannon và Johnson, 1997; Root, 1987). 8 CHAPTER 2: SITUATION OF BAD DEBT MARKET IN VIETNAM AND EXPERIENCE IN HANDLING OF IT IN SOME COUNTRIES 2.1 Economic development process and causes of bad debts 2.1.1 Vietnam's socio-economic development context 2005 - 2018 Economic development is widespread and requires high investment capital, which lead to low capital efficiency. The large investment demand leads to the investment situation far beyond the accumulation capacity of the economy, high budget deficit. To make up for this gap the economy must depend on foreign investment and borrowing. This fact has caused national debt and external public debt to increase rapidly, which have become chronic illness of the economy. 2.1.2 Operation of the system of Vietnamese credit institutions The economy growing was too hot, the real estate market collapsed, interest rates and inflation increased highly, unstable financial markets, businesses using loans for improper purposes, the capacity of market analysis was limited, bad administration, the knowledge and skills of human resources were not high. Additionally, commercial banks and credit institutions raced to profit, which leads to unrestrained - spreading investment, many loan management principles were ignored, the processes and procedures for borrowing are removed. As a result, the bad debt in the whole banking system and credit institutions increased. 2.2 Situation of bad debt market in Vietnam 2.2.1 Situation of bad debt in Vietnam According to calculations with collected data from the State Bank by the author, end of 2012, NPL ratio peaked at 4.12% and then gradually decreased over the years at 2.55% that was below level of 3% and maintain non-performing loan ratio within a safe level until 2018. However, if including the bad debts sold to VAMC has not been processed, then NPL ratio is still quite highly: Table 2.5: Results of solving bad debts from banks via VAMC Unit: Billion VND Items 2015 2016 2017 2018 Total outstanding credit to the economy 4.655.890 5.505.406 6.509.858 6.509.858 NPL ratio reported by credit institutions 2,55% 2,46% 1,99% 1,89% Bad debts reported by credit institutions 118.725 135.432 129.546 136.291 Bad debt VAMC bought 245.000 282.403 313.132 344.049 VAMC bad debt has been recovered 2.278 50.169 81.488 119.000 Bad debt remaining VAMC has not processed 222.220 232.234 231.644 225.049 Total bad debts reported by credit institutions and debts bad at VAMC 340.945 367.666 361.190 361.340 NPL ratio includes non-performing loans at VAMC 7,32% 6,68% 5,55% 5,55% Source: Calculated from published data of the State Bank and VAMC 9 2.2.2 Situation of bad debt market in Vietnam In Vietnam, only VAMC bought bad debts from credit institutions, the purchase of bad debts was not really vibrant, there have not been many big deals, there was not real market for bad debts like: (i) market participants were limited; (ii) the amount of bad debt handled through trading was still small, the effectiveness of the purchase was not high; (iii) The market has not had a consensus on the pricing of debts, so there was a big difference between the buying and selling prices, especially state credit institutions, leading to the process negotiation in buying debt go to failure; (iv) information about bad debt in the market was the lack of transparency and inadequacies, heterogeneity information of bad debt balance and bad debt rate of credit institutions, banking supervision agency, and according to the evaluation of reputable financial institutions in the world; (v) The current policies and laws on bad debt trading market are gradually being improved according to the market mechanism. 2.4 Experience in dealing with bad debts through the debt market in some countries In order for the activities of purchasing and handling bad debts to be effective, most of these countries handle bad debts in ways like: (i) set up a specialized agency where was under the management and administration of the Government and perform the function of purchasing and dealing with bad debts belonging to state enterprises. This specialized agency operates in the form of purchasing bad debts from businesses at commercial banks and credit institutions through the conversion of time bonds, under the Government's guarantee. After purchasing the debt, the agency proceeds to classify and sell the debt through market auctions, or restructure the business until stable operation, then divest the investment. The conversion price was determined based on the ability to recover the capital of the debts and the market price; (ii) establish reconstruction agencies (industry) that directly run by a ministerial agency, usually the Ministry of Commerce and under the supervision of a Commission including representatives of relevant agencies under financial regulation and the current legal system. The industrial rebuilding agency manages and operates a fund called the reconstruction fund, where were formed by national budgets and capital contributions from various economic organizations. It operates by acquiring businesses' debts at commercial banks and credit institutions, after classify and resell to investors or restructure the businesses until they operate with profit, then divest their investments or enjoy profits. The purchasing price of these agencies was determined according to the market price. 10 CHAPTER 3: MODELS AND RESEARCH METHODS 3.1 Hypotheses of bad debt market 3.1.1 Mechanism to form bad debt market The bad debt market is the place where the exchange and sale of debts are recorded in the list of bad debts at commercial banks or credit institutions. In nature, the bad debt market is formed by buying and selling decisions on the market. Thus, to study the factors that affect the bad debt market, the author reviews the factors affecting the buyers 'market participation decisions, sellers' market participation decisions and social welfare on bad debt market. 3.1.2 Factors affecting the decision to participate in the bad debt market a) Factors affecting sellers' decision to enter the market Diagram 3.2: Factors affecting sellers' decision to enter the market Let X be a set of factors affecting sellers' market participation behavior, P is the negotiated price in the market, we have the function of participating in the seller market, including: Businesses, commercial banks, credit institutions or individuals can be simulated through a general function as follows: Q = f(P, X) (3.1) The function (3.1) is a model to study the factors that influence sellers' decision to enter the market. b) Factors affecting buyers' decision to enter the market Call Y a set of factors affecting the decision on market participation of organizations and individuals purchasing bad debts, P is the agreed price in the bad debt trading market, the decision to enter the market of an organization or individual has the following general form: Q = f(P, Y) (3.2) sellers' market participation decisions Endogenous factors Exogenous factors • Business goals • Business resources • Experience of the business • Seller’s expectations • Features of collateral type • Characteristics of bad debts • Macroeconomic factors • Political - legal factors • Socio-cultural factors • Infrastructure and technology • Accessibility of information • The price level in the bad debt market 11 The function (3.2) is The model studies the factors that influence buyers' market entry decisions. 3.1.3 The measure and factors affecting welfare in the bad debt market Research model of factors affecting social welfare in the bad debt market: θ = F(g(X), f(X, Y)) (3.11) θ is the ratio between market price and book value that can become a measure of the change in welfare in the bad debt trading market and is influenced by factors that influence on buyer’s and seller’s market participation decisions. market of of bad debt markets. 3.2 Experimental analytical model and estimated techniques 3.2.1 Binary Logistic Model a) Analytical models The probability of a buyer engaging in a bad debt market is influenced by a set of Zi factors and the price level of P as follows:  =  +  +   +  (3.16) Where li is a dependent variable that receives two values as [0,1] with 1 being the decision of organizations and individuals willing to participate in buying or selling bad debts at Pi price and 0 is the decision of organizations and individuals not buying or selling bad debt at Pi price.  is a constant,  , γ are coefficients of Pi and Zi, reflects the impact level of Pi and Zi on li. ε$ is the redundant reflects the unobservable variables in the model.. j reflects the number of variables included in the model, j≠i, j=1m. 3.2.2 Multinomial Logistic Model a) Analytical models Call K is a dependent variable including a set of price that a buyer Pi, the ith of buyer willing to pay or a seller is willing to sell on bad debt trading market. Assume that the price the buyer is willing to pay or the seller willing to sell to be influenced by a set of factors Zij, Zij is defined as a vector of independent variables, j=1M. The relationship between Pi and Zi prices can be expressed as a function of the following: Pi=f (Zij) (3.25) Assume that f(Z$&) is a linear function and is defined as follows: f'Z$&( = α +β$&Z$& + & + ε$ (3.28) , wherw  is a constant constant, is the regression coefficient of the variable of ,  is the balance reflecting unobservable variables in the research model. 3.3 Data and collection Used data for analysis inclues primary data and secondary data. To collect primary data, the study conducted a survey with 500 questionnaires that have been randomly sent to 12 businesses, commercial banks and credit institutions in Hanoi city. After removing 33 copies due to incomplete responses from the 456 questionnaires collected, the number of questionnaire used for research purposes was 423 copies. Most of the variables in the questionnaire were designed according to series of ordinal data and assumed that variables were employed in the analysis model, were independent variables. Therefore, to ensure reliability during the estimation process, the study conducted non-parametric tests with data samples collected through questionnaires. Test results have shown that all variables included in the analysis model were independent variables, statistically significant at 1% (P value = 0.00), therefore, the data included in the model was guaranteed the reliability of the estimation process. CHAPTER 4: RESULTS AND DISCUSSION 4.1 Descriptive statistics
Type of business and business sector Table 4.1: Type of business and business sector Variable symbol Type of business Amount Percent VAR65 Company with more than 50% State capital 84 19,9 Co., Ltd. 80 18,9 Other joint stock company 165 39,0 Partnership company 11 2,6 Private enterprise 36 8,5 Foreign-invested enterprises 13 3,1 Business households and individuals 34 8,0 Business sector VAR68 Real estate business 105 24,8 Wholesale and retail business 46 10,9 Construction investment 27 6,4 Manufacturing and processing industry 27 6,4 Transportation & warehousing services 3 0,7 Agriculture and fisheries 8 1,9 Business in the field of education 11 2,6 Information and communication field 6 1,4 Accommodation and dining 47 11,1 Finance, banking and insurance 143 33,8 13
Info rm atio n of th e b u sin ess a nd it’s rep resentatives p a rticip ating in th e interview T able 4 .2 : Info rm atio n of interview ees V a riable sym b ol V a riables Info rm ation of th e interview ee V A R69 N u m b er of em ploy ees labo r ≤10 10 -50 50 -100 100 - 200 200 -300 300< lab o r 43 (10 ,2%) 110 (26 ,0%) 59 (13 ,9%) 66 (15 ,6%) 53 (12 ,5%) 92 (21 ,7%) V A R70 O w n er capital C apital≤1 billio n 1 - 5 5 -10 10 -50 50 -100 100 billio n < C apital 8 (1 ,9%) 93 (22 ,0%) 87 (20 ,6%) 22 (4 ,35%) 49 (11 ,6%) 164 (38 ,8%) V A R72 A cad em ic lev el C olleg es U nd erg rad u ater M asters D o cto r P rofesso r 5 (1 ,2%) 357(84 ,4%) 56(13 ,2%) 5(1 ,2%) 0(0%) V A R71 G end er M en W o m en 239 (56 ,5%) 184 (43 ,5%) 4 .2 F a cto rs affecting sellers ' d ecisio n to enter th e m a rk et Bin ary logistic reg ressio n m od el and M a xim u m Lik elih o od M eth od h av e b een em ploy ed to an aly ze th e facto rs affecting m ark et p articip atio n b eh avio r . Th e an alytical results in table 4 .10 h av e sh o w n th at relev an ce and p redictiv e lev el of an aly sis m od el w as relativ ely high w ith th e v alu e of -2 L og lik elih o od is 119 ,069 , C o x & S n ell R Sq u are w as 0 .491 , N ag elk erk e R Sq u are w as 0 .797 at 1% statistical sig nifican ce and O v erall P ercentag e w as 82 .4 . Th e m ajo rity of v ariables in clud ed in th e m od el w ere fo u nd to b e statistically sig nificant , in w hich th e p rice (P1) h as a p o sitiv ely and stro ngly effect o n th e seller 's ability to enter th e m ark et w ith a co rrelatio n co efficient of 0 .534 at 1% of statistical sig nifican ce . This im plies th at , w h en th e p rice in creases , th e seller 's ability to p articip ate in th e m ark et w ill in crease and vice v ersa . A s L eo nid H u rw icz (1945) h as p ointed o ut th at m ark et p rices play an im p o rtant role in d eterm ining th e p articip atio n and supply of p rod u cts , g o od s and services in th e m ark et , th e in crease in p rices w o uld lead to in creased seller b en efits and , as a result , stim ulate sellers to p articip ate and supply p rod u cts , g o od s and services in th e m ark et . , , 14 Table 4.10: Factors affecting sellers' decision to participate in market -2 Log likelihood: 119,069 Cox & Snell R Square: 0,491 Nagelkerke R Square: 0,797 Variable B S,E Wald df Sig, Exp(B) Variable B S,E Wald df Sig, Exp(B) P2 0,534 0,094 32,329 1 0,000 1,706 VAR27 -0,279 0,356 0,616 1 0,433 0,756 VAR02 -0,838 0,391 4,605 1 0,032 0,433 VAR28 0,407 0,298 1,863 1 0,017 1,503 VAR03 -0,275 0,396 0,480 1 0,048 0,760 VAR29 0,193 0,238 0,657 1 0,418 1,212 VAR04 -0,475 0,397 1,429 1 0,023 0,622 VAR30 0,287 0,262 1,203 1 0,027 1,333 VAR05 0,241 0,360 0,446 1 0,050 1,272 VAR31 -0,363 0,186 3,811 1 0,051 0,695 VAR06 -0,160 0,362 0,196 1 0,658 0,852 VAR32 0,125 0,176 0,504 1 0,478 1,133 VAR08 -0,213 0,393 0,294 1 0,588 0,808 VAR33 -0,537 0,151 12,687 1 0,000 0,585 VAR09 -0,244 0,388 0,397 1 0,529 0,783 VAR34 -0,162 0,162 1,007 1 0,031 0,850 VAR10 1,204 0,422 8,142 1 0,004 3,332 VAR35 -0,084 0,129 0,429 1 0,513 0,919 VAR11 -1,466 0,441 11,038 1 0,001 0,231 VAR36 -0,027 0,199 0,018 1 0,089 0,973 VAR12 0,852 0,367 5,374 1 0,020 2,344 VAR37 0,057 0,185 0,094 1 0,760 1,058 VAR13 1,106 0,340 10,575 1 0,001 3,022 VAR38 -0,069 0,127 0,292 1 0,589 0,934 VAR14 1,073 0,295 13,225 1 0,000 2,924 VAR39 -0,865 0,208 17,276 1 0,000 0,421 VAR15 -0,220 0,345 0,405 1 0,524 0,803 VAR40 0,190 0,134 1,989 1 0,158 1,209 VAR16 -0,707 0,364 3,761 1 0,052 0,493 VAR65 0,270 0,233 1,344 1 0,246 1,310 VAR17 0,134 0,425 0,099 1 0,075 1,143 VAR66 33,994 15,306 4,933 1 0,026 5,8E+10 VAR18 1,126 0,407 7,655 1 0,006 3,082 VAR67 0,173 0,125 1,911 1 0,167 1,189 VAR19 -0,040 0,345 0,013 1 0,909 0,961 VAR68 -0,526 0,149 12,508 1 0,000 0,591 VAR20 1,938 0,462 17,573 1 0,000 6,946 VAR69 -0,027 0,328 0,007 1 0,093 0,973 VAR21 -0,781 0,287 7,409 1 0,006 0,458 VAR70 0,269 0,323 0,692 1 0,405 1,308 VAR22 0,352 0,294 1,438 1 0,023 1,422 VAR71 -0,167 0,737 0,051 1 0,821 0,846 VAR26 0,285 0,348 0,667 1 0,041 1,329 VAR72 -1,382 0,832 2,758 1 0,097 0,251 Constant -30,511 10,942 7,776 1 0,005 0,000 15 4 .3 F acto rs affecting b uyers ' d ecisio n to enter th e m ark et Sim ilar to th e supply in th e b ad d ebt m ark et , Bin ary L ogistic R eg ressio n M od el and M axim u m Lik elih o od E stim atio n M eth od w ere em ploy ed to an aly ze th e facto rs affecting m ark et entry d ecisio n . Th e an alytical results in T able 4 .11 h av e indicated th at th e relev an ce and accu racy of th e fo recasting m od el of th e an aly sis m od el w as relativ ely high w ith th e v alu e of -2 L og lik elih o od w as 186 ,495 , C o x & S n ell R Sq u are w as 0 .477 , N ag elk erk e R Sq u are w as 0 .719 and O v erall P ercentag e w as 87 .4 , Th e m ajo rity of v ariables in clud ed in th e an aly sis m od el w ere fo u nd to b e statistically sig nificant , in w hich th e p rice of P 1 w as fo u nd to n eg ativ ely influ en ce b uy ers ' d ecisio n to enter th e m ark et w ith a co rrelatio n co efficient of -0 .468 at 1% of statistical sig nifican ce . Th e n eg ativ e im p act of p rices o n b uy ers ' m ark et entry d ecisio n s im plies th at w h en th e p rice in creases , b uy ers ' ability to p articip ate in th e m ark et d ecreases . A cco rding to L eo nid H u rw icz (1945) , m ark et p rices play an im p o rtant role in b uy ers ' b eh avio r w h en d eciding to enter th e m ark et . This w as d u e to th at ch ang es in p rices w o uld ch ang e th e interests of b uy ers in th e m ark et . W h en p rices in th e m ark et tend to in c rease th e interests of b uy ers tend to d ecrease , th u s less stim ulating b uy ers to join th e m ark et . 16 Table 4.11: Factors affecting buyers' market participation decisions -2 Log likelihood: 186,495 Cox & Snell R Square: 0,477 Nagelkerke R Square: 0,719 Variable B S,E Wald df Sig, Exp(B) Variable B S,E, Wald df Sig, Exp(B) P1 -0,468 0,087 28,643 1 0,000 0,627 VAR48 1,241 0,458 7,348 1 0,007 3,460 VAR02 -0,312 0,311 1,002 1 0,031 0,732 VAR49 0,163 0,098 2,761 1 0,097 1,178 VAR03 0,313 0,294 1,133 1 0,028 1,367 VAR50 0,185 0,111 2,808 1 0,094 1,204 VAR04 -0,142 0,300 0,224 1 0,063 0,868 VAR51 0,123 0,116 1,122 1 0,029 1,131 VAR05 0,134 0,222 0,363 1 0,547 1,143 VAR52 0,027 0,087 0,100 1 0,752 1,028 VAR06 0,829 0,241 11,841 1 0,001 2,291 VAR53 -0,349 0,119 8,632 1 0,003 0,705 VAR08 1,410 0,351 16,129 1 0,000 4,095 VAR54 0,131 0,100 1,714 1 0,019 1,140 VAR09 0,221 0,303 0,533 1 0,465 1,248 VAR55 0,086 0,123 0,493 1 0,048 1,090 VAR10 0,519 0,354 2,149 1 0,014 1,680 VAR56 0,167 0,147 1,283 1 0,025 1,181 VAR11 -0,650 0,245 7,063 1 0,008 0,522 VAR57 -0,340 0,124 7,523 1 0,006 0,711 VAR12 0,085 0,266 0,102 1 0,749 1,089 VAR58 -0,285 0,120 5,613 1 0,018 0,752 VAR13 -0,122 0,246 0,246 1 0,620 0,885 VAR59 -0,224 0,161 1,942 1 0,016 0,799 VAR14 -0,298 0,197 2,290 1 0,013 0,743 VAR60 -0,546 0,813 0,450 1 0,050 0,579 VAR15 0,318 0,229 1,927 1 0,016 1,375 VAR61 0,587 1,204 0,238 1 0,626 1,798 VAR16 -0,103 0,215 0,231 1 0,063 0,902 VAR62 -0,317 2,212 0,021 1 0,088 0,728 VAR17 -0,011 0,235 0,002 1 0,961 0,989 VAR63 -0,281 0,745 0,142 1 0,070 0,755 VAR18 -0,574 0,223 6,611 1 0,010 0,563 VAR64 0,111 0,426 0,067 1 0,795 1,117 VAR19 -0,100 0,197 0,260 1 0,610 0,904 VAR65 -0,888 0,193 21,279 1 0,000 0,411 VAR20 -0,103 0,193 0,287 1 0,059 0,902 VAR66 24,622 13,178 3,491 1 0,062 4,9E+11 VAR21 0,009 0,166 0,003 1 0,957 1,009 VAR67 -0,591 0,164 12,972 1 0,000 0,554 VAR22 -0,127 0,171 0,556 1 0,045 0,880 VAR68 -0,085 0,101 0,704 1 0,040 0,918 VAR44 2,143 0,928 5,337 1 0,021 8,524 VAR69 0,162 0,208 0,611 1 0,434 1,176 VAR45 -3,138 1,501 4,367 1 0,037 0,043 VAR70 -0,277 0,248 1,246 1 0,026 0,758 VAR46 6,641 2,874 5,341 1 0,021 766,009 VAR71 -0,663 0,515 1,658 1 0,198 0,515 VAR47 -0,827 0,776 1,136 1 0,028 0,438 VAR72 -1,660 0,585 8,058 1 0,005 0,190 Constant -11,232 13,163 0,728 1 0,393 0,000 17 4 .4 F a cto rs affecting so cial w elfa re in b ad d ebt m a rk et Th e M ultin o m ial logistic reg ressio n m od el and th e M axim u m Lik elih o od E stim atio n M eth od w ere em ploy ed to estim ate th e co efficients in th e an alytical m od el . A fter rem o ving v ariables w ith statistical n o -sig nifican ce , th e an alytical results in T able 4 .12 h av e sh o w n th at th e relev an ce and accu racy estim ates of th e an aly sis m od el are relativ ely high w ith -2 LL v alu es of 137 ,207 , C o x and S n ell R of 0 .683 and N ag elk erk e R of 0 .735 w ith th e ob serv ed sam ple n u m b er of 423 . Th e an aly sis h av e also indicated th at th e p rop erxy of m ark et w elfare w ere divid ed into th ree different rang es , in cluding : (i) th e p rop erxy of m ark et w elfare in th e m ark et rang e fro m 0 .68 to 0 .75 , (ii) th e p rop e rxy of m ark et w elfare in th e m ark et rang e fro m 0 .76 to 0 .83 and (iii) th e p rop erxy of m ark et w elfare in th e m ark et rang e fro m 0 .83 to 0 .90 . A t each rang e lev el , th e d eg ree of im p act and im p act dim en sio n of th e v ariables in clud ed in th e an aly sis are relativ ely different . 18 Table 4.12: Factors affecting welfare in the bad debt trading market Variables 0,68 ≤ . ≤0,75 0,76 ≤ . ≤0,83 0,84 ≤ . ≤0,90 B SE Wald df Sig, Exp(B) B SE Wald df Sig, Exp(B) B SE Wald df Sig, Exp(B) Intercept -25,98 7,63 11,58 1 0,00 -16,41 6,37 6,63 1 0,01 5,83 7,61 ,58 1 0,44 VAR02 -0,17 0,24 0,51 1 0,47 0,84 -0,09 0,23 0,14 1 0,70 9,1E-01 -0,34 0,26 1,68 1 0,02 7,1E-01 VAR03 0,32 0,31 1,026 1 0,03 1,38 0,38 0,30 1,65 1 0,02 1,4E+00 -0,56 0,32 3,07 1 0,08 5,7E-01 VAR04 0,68 0,32 4,33 1 0,03 1,98 0,60 0,32 3,60 1 0,05 1,8E+00 0,45 0,35 1,68 1 0,02 1,5E+00 VAR05 -0,61 0,25 5,93 1 0,01 0,54 -0,44 0,23 3,50 1 0,06 6,4E-01 -0,84 0,27 9,63 1 0,00 4,3E-01 VAR06 0,67 0,25 7,12 1 0,00 1,96 0,22 0,24 0,89 1 0,03 1,2E+00 -0,17 0,27 0,39 1 0,52 8,4E-01 VAR08 1,07 0,33 10,18 1 0,00 2,92 1,15 0,32 12,64 1 0,00 3,1E+00 -0,72