Bankruptcy Prediction Model With Zeta Optimal Cut-Off Score To Correct Type I Errors

BANKRUPTCY PREDICTION MODEL WITH ZETAc OPTIMAL CUT-OFF SCORE TO CORRECT TYPE I ERRORS
MOHAMAD IWAN Yogyakarta Abstract This research examines financial ratios that distinguish between bankrupt and non-bankrupt companies and make use of those distinguishing ratios to build a one-year prior to bankruptcy prediction model. This research also calculates how many times the type I error is more costly compared to the type II error. The costs of type I and type II errors (cost of misclassification errors) in conjunction to the calculation of prior probabilities of bankruptcy and nonbankruptcy are used in the calculation of the Z T! c optimal cut-off score. The bankruptcy prediction result using Z T!c optimal cut-off score is compared to the bankruptcy prediction result using a cut-off score which does not consider neither cost of classification errors nor prior probabilities as stated by "air et al. (#$$%) and for later purposes will be referred to "air et al. optimum cutting score. &omparison between the prediction results of both cut-off scores is purported to determine the better cut-off score between the two' so that the prediction result is more conser(ati(e and minimi)es expected costs' which may occur from classification errors. This is the first research in Indonesia that incorporates type I and II errors and prior probabilities of bankruptcy and non-bankruptcy in the computation of the cut-off score used in performing bankruptcy prediction. arlier researches ga(e the same weight between type I and II errors and prior probabilities of bankruptcy and nonbankruptcy' while this research gi(es a greater weigh on type I error than that on type II error and prior probability of non-bankruptcy than that on prior probability of bankruptcy. This research has successfully attained the following results* (#) type I error is in fact +$'%, times more costly compared to type II error' (-) -- ratios distinguish between bankrupt and non-bankrupt groups' (,) - financial ratios pro(ed to be effecti(e in predicting bankruptcy' (.) prediction using Z T! c optimal cut-off score predicts more companies filing for bankruptcy within one year compared to prediction using "air et al. optimum cutting score' (+) !lthough prediction using "air et al. optimum cutting score is more accurate' prediction using Z T! c optimal cut-off score pro(ed to be able to minimi)e cost incurred from classification errors.
BACKGROUND Uncertain condition in Indonesians economy nowadays put firms in the risk of experiencing financial distress or even bankruptcy. Prediction error towards the continuity of an entity in the future can cause severe loss. There are two types of errors that may occur namely type I error and type II error. If the prediction performed is a prediction about whether a firm is to file for bankruptcy or not in the future then type I error can be comprehended as predicting a firm not to file for bankruptcy while in fact the firm does file for bankruptcy. Type II error on the contrary is predicting a firm to file for bankruptcy while in fact the firm does not file for bankruptcy. !lthough both types of prediction errors inflict a certain amount of financial loss type I errors inflict a greater financial loss compared to that of type II errors. Therefore prediction re"uires a cut#off score which classifies a company in to either the bankrupt group or the non#bankrupt group with a minimum cost of classification errors. $itherto there are no theories that affirm definitely what financial ratios must be used in predicting bankruptcy. %atios used in predicting bankruptcy can vary among different researches. It is merely the sub&ective consideration of the researcher followed by statistical verification on financial ratios applied. This research uses the formula of !ltman $aldeman dan 'arayanan ()*++, to calculate the -.T!c optimal cut#off score. /inancial ratios developed by 0achfoed1 ()**2, are used to distinguish between bankrupt and non#bankrupt groups. The distinguishing financial ratios are eventually opted based on statistical testing to build the prediction model. The prediction model in this research is developed in the same way of !vianti (3444, that is by applying the two#group discriminant analysis. ! cut above or advantage of this research is that it incorporates the type I and II errors and prior probabilities of bankruptcy and non#bankruptcy in the computation of -.T!c optimal cut#off score thus the prediction result is expected to be able to minimi1e cost incurred from classification errors compared to the prediction result if using the cut#off score as the one stated in $air et al. ()**5, and for later purposes will be referred to $air et al. optimum cutting score.
SIMPOSIUM NASIONAL AKUNTANSI VI

Surabaya, 16 17 Oktober 2003
114
Ba kru!t"y Pre#$"t$o Mo#e% &$t' (eta" O!t$)a% *ut+O,, S"ore To *orre"t Ty!e I -rror.
SESI
PROBLEM FORMULATION AND RESEARCH OBJECTIVE This research investigates what financial ratios distinguish between bankrupt and non#bankrupt companies and makes use of those distinguishing ratios to build a one#year prior to bankruptcy prediction model. This research also calculates how many times type I error is more costly compared to type II error. The costs of type I and type II errors (cost of misclassification errors, in con&unction to the calculation of prior probabilities of bankruptcy and non#bankruptcy is used in the calculation of the -.T!c optimal cut#off score. The bankruptcy prediction result using -.T!c optimal cut#off score is then compared to the bankruptcy prediction result using $air et al. optimum cutting score to determine the better cut#off score to apply in terms of less costly in predicting bankruptcy. RESEARCH LIMITATIONS 6ome limitations exist in this research which are7 ). Previous researches in general define bankruptcy as legal bankruptcy and this definition is applied as the dependent variable (8eaver )*99:)*95; !ltman )*95; <hlson )*54; -ain )**2 etc,. $owever this research applies the stock based insolvency definition and uses negative e"uity as its dependent variable (!vianti 3444,. In conse"uence the term bankruptcy in this research means having a negative e"uity. 3. The determination of the cost of classification errors is obtained from a sample of two state owned banks (8ank 8'I dan 8ank 8%I, and two private banks (8ank 'iaga dan 8ank =anamon, for the year of )**9 )**+ )*** 3444. Thus there is a probability that the computation of the cost of classification errors does not represent the entire population as a whole. >. 'o hold out sample is employed in this research; hence prediction is limited to the original sample. The reason for not employing a hold out sample is because of time consideration and lack of data which until the point this research was concluded was not available. RESEARCH BENEFITS This research is deemed to have the following benefits7 ). This research is expected to become a basis to opt for the cut#off score to be used in conducting bankruptcy prediction. 3. This research is expected to benefit stakeholders of a company in assessing and making short# term decisions regarding the company. >. This research is expected to become a sub&ect of information and: or consideration for developments in further researches. 2. This research is expected to start and open a discourse of research in Indonesia regarding bankruptcy prediction that incorporates cost of classification errors and prior probabilities of bankruptcy and non#bankruptcy. THEORITICAL BASIS %atios are among the most popular and widely used tools of financial analysis (8ernstein and ?ild )**5,. The result of the calculation of financial ratios obtained from a set of financial statements is able to determine the economic ability of a company (0achfoed1 )**@ in !vianti 3444,. 0achfoed1 ()**2, in !vianti (3444, states that financial ratios can be used to predict future events by associating financial ratios with economic phenomenas. Prediction is conducted to reduce future uncertainties. /inancial ratios can be used to predict future bankruptcy by developing a bankruptcy prediction model. The model developed in this research is purported to represent financial conditions of companies and to predict whether companies will file for bankruptcy or not (!vianti 3444,. Defini i!n" !f B#n$%&' c( 8ankruptcy can be classified in to two categories which are stock based insolvency and legal bankruptcy (!vianti 3444,. Previous researches apply the legal bankruptcy definition as its dependent variable. The term bankrupt in this research as in (!vianti 3444, applies the stock based insolvency definition. ! company is said to be bankrupt (stock based insolvency, if a company experiences lack of temporary li"uidity and continues to have a larger book value of liabilities than assets thus the e"uity
115
SESI
becomes negative (including minority interest in the subsidiarys net assets,. In such circumstance a company is said to be bankrupt from the e"uity perspective (%oss et al. )**> and 8righam A Bapensky )**> in !vianti 3444,. The reason why this definition is employed is because data of publicly held companies in Indonesia that are legally bankrupt are very hard to get hold of if any. The Commerce Court in Indonesia was formed in )**5 and until the point this research was conducted there were only very few bankruptcy appeals. Conse"uently it is very difficult to find data regarding companies having legally bankrupt status. !ccording to !ct 'umber 2:)**2 an institution is stated to be bankrupt by the &udgment of court if the debtor retain two or more creditors and does not pay at least one overdue and collectible debt. =ifferent researchers often define the term bankruptcy differently and due to certain conditions and limitations thus the term bankrupt employed in this research is companies that have negative e"uities and for that reason other definitions are not cited in this research. PREVIOUS LITERATURE AND HYPOTHESES DEVELOPMENT There have been two types of bankruptcy prediction studies. The first (e.g. 8eaver )*99, looks at the relation between individual accounting numbers or ratios and bankruptcy (the univariate approach,. The other uses several ratios to predict bankruptcy (the multivariate approach,. The univariate approach uses one ratio at a time to predict failure. It is likely that different ratios reflect different aspects of the firms financial position so better predictions can be obtained by using combinations of ratios instead of one ratio. /or this reason the multivariate approach "uickly supplanted the univariate approach (?atts and -immerman )*59,. 6everal researches were conducted in the bankruptcy prediction domain (companies predicted do not include banking and financial sector companies, as the followings 8eaver ()*99; )*95a; )*95b, !ltman ()*95; )*+>, !ltman and Dorris ()*+9, !ltman and 0cBough ()*+2, !ltman $aldeman and 'arayanan ()*++, =eakin ()*+3, Dibby ()*+@, 8lum ()*+2, .dmister ()*+3, ?ilcox ()*+>, 0oyer ()*++, Dev ()*+), <hlson ()*54, 6chiedler ()*5), 6cott ()*5), =ambolena A Ehoury ()*54, -mi&ewski ()*52, 0ensah ()*52, Bentry et al. ()*5+, 8arniv A %aveh ()*5*, Platt A Platt ()**4, -ain ()**2, %ichardson et al. ()**5, Din et al. ()***, 6etyorini A $alim ()***, and !vianti (3444,. 8eaver ()*99, used the univariate approach and the main findings of the research was that accounting data in forms of financial ratios have the ability to predict failure for at least five years prior to the failure. 8eaver called this approach as a profile analysis. !ltman ()*95, used the multivariate discriminant analysis (0=!, to predict bankruptcy. There was a limitation in !ltmans model because the prediction accuracy for predictions over than two years prior to bankruptcy became awfully low compared to the prediction of one and two years prior to bankruptcy. !ltman $aldeman and 'arayanan ()*++, also developed a bankruptcy prediction model using the 0=!. <ne distinguishing point of this research amongst others is the determination of the cut#off score. <ther bankruptcy prediction researches give the same weight between type I and type II errors and prior probabilities of bankruptcy and non#bankruptcy. That approach is in contrast with this research that uses the -.T!c optimal cut#off score which gives a greater weight on type I error than type II error and prior probability of non bankruptcy than prior probability of bankruptcy. <ne of !ltmans essential findings is that type I error is >@ times more costly than type II error. Therefore the cut#off score must weigh type I error and type II error differently. <hlson ()*54, used the logistic regression analysis and developed > bankruptcy prediction models; prediction models for one two and one or two years prior to bankruptcy. The sampling method used by <hlson ()*54, was proportional with the population. <hlson ()*54, also considered the publication date of financial statements because other researches assumed that financial statements for the year of bankruptcy were published before bankruptcy filings occurred resulting in overstatement of prediction power. !vianti (3444, built bankruptcy prediction models using > different methods. .ach method was used to build > bankruptcy prediction models#one year two years and three years prior to bankruptcy; hence * models were successfully developed. The methods used to build the models were Dinier =iscriminant !nalysis Dinier =iscriminant !nalysis combined with Principal Component !nalysis and Dogistic %egression. =ifferent financial ratios were employed as the predictor variable for each year and
116
SESI
each method. %esults showed that the linier discriminant model was superior to the other two methods in performing predictions for one and two years prior to bankruptcy. !s for the prediction of > years prior to bankruptcy the logistic regression model was superior to the other two models. Previous researches have proven that financial ratios can be used to build a bankruptcy prediction model and use different financial ratios in order to predict bankruptcy. It has been proven that different sets of data will result in different models (!ltman ()*95, !ltman $aldeman and 'arayanan ()*++, <hlson ()*54, !vianti (3444,. 6ince financial ratios differ along with different researches that use different data it is necessary here to determine what financial ratios differ from bankrupt and non#bankrupt companies according to the data used in this research (prediction can only be conducted if any characteristics of the ob&ect being predicted do exist which in this case the ob&ect are bankrupt and non#bankrupt companies thus the characteristics that are expected to differ from bankrupt and non#bankrupt companies are financial ratios of both groups,. <n the basis of the argument above the following alternative hypothesis is formulated7 $)7 /inancial ratios differ between bankrupt and non#bankrupt companies. If financial ratios that distinguish between bankrupt and non#bankrupt companies exist then those financial ratios will be used to develop a one#year prior to bankruptcy prediction model. This step is performed to further understand whether those financial ratios can be used to predict bankruptcy beforehand. !s a result the following alternative hypothesis is formulated7 $37 /inancial ratios can be used to predict bankruptcy beforehand.
Researches and Year Type I Cost= Type I Cost= 2xType II Cost Type I Cost= 20xType II Cost NB "1$8 88$8 "5$" "1$4 "5$% "3$5 "5$5 B 86$1 84$% %%$8 81$" %3$6 84$% %5$0 TS "1$% 88$8 "5$6 "1$2 "5$3 "5$3 "5$1 Type I Cost=38xType II Cost NB B TS "1$8 86$1 "1$% 83$3 "3$1 83$5 "0$6 8%$5 "0$1 84$0 "4$4 84$3 83$% "1$6 83$" 8"$6 "3$1 8"$% 86$0 "1$2 86$0
NB B TS NB B TS Bea er !1"66# ""$% 36$1 "8$4 ""$2 43$1 "8$1 &'t(an !1"68# ""$8 6$" "%$" ""$3 33$3 "8$0 B')( !1"%4# ""$8 18$1 "8$" ""$4 38$" "8$2 &'t(an et a'$ !1"%%# ""$8 4$2 "%$" "8$" 38$" "%$% *a(+o'ena,-ho)ry!1"80# ""$% 26$4 "8$3 "8$" 4%$2 "%$% .h'son !1"80# ""$8 8$3 "8$0 "8$8 40$3 "%$% /(01e2s30 !1"83# ""$% "$% "840 ""$3 3%$5 "8$1 Note5 NB 6 Non7+an3r)pt co(pan0es B 6 Ban3r)pt co(pan0es TS 6 Tota' Sa(p'e So)rce5 /(01e2s30 !1"834 Ta+e' 5dan 6#5 pp$32733 0n 8oster !1"86#
T#)*e +, Pe%cen #-e !f c!%%ec *( c*#""if(in- !f . )#n$%&' c( '%e/ic i!n 0!/e*" &"in- 1e 0&* i2#%i# e #''%!#c1 'ecessary to stress on are researches of 8eaver ()*99, !ltman ()*95, <hlson ()*54, and !vianti (3444,. 'either of those considered on prior probabilities of bankruptcy and non#bankruptcy nor cost of classification errors (cost occurring from type I and II errors,. !ccording to -mi&ewski ()*5>, in table ) above if type I error is given a greater weight than type II error then the percentage of correctly predicting the bankrupt companies also becomes greater while the percentage of correctly predicting the non# bankrupt companies becomes smaller. !n increase of correctly predicting bankrupt companies means that the amount of companies predicted to be non#bankrupt from those companies supposed to be predicted as bankrupt companies becomes smaller. <n the contrary a decrease of correctly predicting non#bankrupt companies means that the amount of companies predicted to be bankrupt from those companies supposed to be predicted as non#bankrupt becomes greater. Thus giving a greater weight on type I error than type II error will predict more companies as bankrupt companies and less companies as non#bankrupt companies. !ccordingly the following alternative hypotheses is formulated7 $>7 The percentage of predicting bankrupt companies will become larger by giving a greater weight on type I error than type II error compared to the percentage of predicting bankrupt companies of a prediction that gives the same weight on type I and type II errors. The greater the amount of companies predicted to be bankrupt indicate the lesser the amount of type I errors. Thus expected cost of classification errors can be minimi1ed since type I error is much more
11%
SESI
costly compared to type II error. Conse"uently a model with a cut#off score that gives a greater weight on type I error than type II error is better in terms of more cost effective than a model that gives the same weight on type I and type II errors. /or this reason the following alternative hypotheses is formulated7 $27 8ankruptcy prediction with a cut#off score that incorporates prior probability of bankruptcy and non#bankruptcy and type I and type II errors will cut costs occurring from prediction errors. RESEARCH METHODOLOGY SAMPLE AND DATA S#0'*e !f B&i*/in- 1e B#n$%&' c( P%e/ic i!n M!/e* The sampling method employed in building the bankruptcy prediction model is the matched pair sampling method based on industrial sectors and si1e of companies. Industrial sectors are determined based on the categories in the Indonesian Capital 0arket =irectory 3444 and 344) while the si1e of companies are determined based on the total asset average in accordance with 8apepams regulation sub&ect on /oreign Capital Investment: =omestic Capital Investment, regarding the criterias of si1es of companies based on the amounts of total assets. !verages of total assets that are used to determine the si1es of companies are obtained from the Indonesian Capital 0arket =irectory 3444 and 344). 6teps in determining sample are as follows7 ). =etermine bankrupt companies in years of )*** and 3444 and trace their financial statements one year back which are years of )**5 for bankrupt companies of )*** and )*** for bankrupt companies of 3444. 3. =etermine non#bankrupt companies that match bankrupt companies determined earlier. 'on# bankrupt companies are selected using industrial sectors and si1es as the criteria of matched pairs. The followings are lists of bankrupt and non#bankrupt companies used in building the one#year prior to bankruptcy prediction model. Table 2. Sample of bankrupt and non-bankrupt companies
No. Bankrupt Companies Non-bankrupt Companies 1 9T C0tatah Ind)str0 :ar(er T+3 9T &ne3a Ta(+an; T+3 2 9T 9ras0dha &ne3a N0a;a T+3 9T :)'t0 B0ntan; Indones0a T+3 3 9T Tex(aco <aya T+3 9T Te010n Indones0a 80+er Corporat0on !TI8IC.# T+3 4 9T 9o'ys0ndo =3a 9er3asa T+3 9T B)d0 &c0d <aya T+3 5 9T 9anca >0rata(a Sa3t0 T+3 9T 9)d10ad0 9rest0;e ?0(0ted T+3 6 9T 99 9er3e+)nan ?ondon S)(atra Indones0a T+3 9T &stra &r;o ?estar0 T+3 % 9T &'ter &+ad0 T+3 9T T0(ah T+3 8 9T *a o(as &+ad0 T+3 9T Sar0 @)sada T+3 " 9T S:&RT Corporat0on T+3 9T :ayora Indah T+3 10 9T &r;o 9antes T+3 9T 9anas0a Indosyntec T+3 11 9T 9r0(ar0ndo &s0a InArast)ct)re T+3 9T Sepat) Bata T+3 12 9T S)rya *)(a0 Ind)str0 T+3 9T S)(a'0ndo ?estar0 <aya T+3 13 9T S)ra+aya &;)n; Ind)str0 9)'p T+3 9T S)par(a T+3 14 9T To+a 9)'p ?estar0 !9T Ind0rayon Bta(a# T+3 9T 8a1ar S)rya >02esa T+3 15 9T Tr0 9o'yta Indones0a T+3 9T ?a)tan ?)as T+3 16 9T &r;ha -arya 9r0(a Ind)stry T+3 9T >ahana <aya 9er3asa !9T BC&@&RI# T+3 1% 9T :)'0a Ind)str0ndo T+3 9T S)rya Toto Indones0a T+3 18 9T Tex(aco 9er3asa =n;0neer0n; T+3 9T -o(ats) Indones0a T+3 1" 9T CT -a+e' Indones0a!-a+e'(eta' Indones0a#T+3 9T S)(0 Indo3a+e' !I-I Indah -a+e' Indones0a# T+3 20 9T Ca1ah T)n;;a' T+3 9T &stra Internat0ona' T+3 21 9T Indo(o+0' S)3ses Internat0ona' T+3 9T &stra .toparts T+3 22 9T CT 9etroche( Ind)str0es T+3 9T Bn0ted Tractor T+3 23 9T C0p)tra *e e'op(ent T+3 9T B)30t Sent)' T+3 24 9T *har(a'a Int0'and T+3 9T *)ta 9ert020 T+3 25 9T <a3arta Set0a+)d0 9roperty T+3 9T <a3arta Internat0ona' @ote' , *e e'op(ent T+3 26 9T -a2asan Ind)stry <a+a+e3a T+3 9T <aya Rea' 9roperty T+3
118
SESI
9T Ba3r0e'and *e e'op(ent !='an; Rea'ty# T+3 9T S)rya Se(esta Intern)sa T+3 9T S)((arecon &;)n; T+3
2% 9T ?0ppo ?and *e e'op(ent T+3 28 9T :odern'and Rea'ty T+3 2" 9T S)rya(as *)ta :a3()r T+3
S#0'*e &"e/ in /e e%0inin- '%i!% '%!)#)i*i ie" Prior probabilities of bankruptcy are calculated using the numbers of bankrupt companies divided by total companies listed on the Fakarta 6tock .xchange for years of )**9 )**+ )*** and 3444 while the calculation of prior probabilities of non#bankruptcy are the numbers of non#bankrupt companies divided by total companies listed on the Fakarta 6tock .xchange for years of )**9 )**+ )*** and 3444. S#0'*e &"e/ in c#*c&*# in- c!" !f c*#""ific# i!n e%%!%" 3 ('e I #n/ ('e II e%%!%"4 The sample of banks used in computing cost of classification errors are 8ank 8'I 8ank 8%I 8ank 'iaga and 8ank =anamon. !ll data used are data from years )**9 )**+ )*** and 3444. !ssuming that banking characteristics in Indonesia is similar the two state#owned and two private#owned banks aforementioned are expected to represent the whole banking population. =ata from year )**5 are not included in the determination of the -.T! c optimal cut#off score (calculation of prior probabilities and cost of classification errors, for the reason that in year )**5 the average credit interest rate and the Central 8ank Certificate interest rate boosted very high in comparison with previous and subse"uent years because of the economic crisis thus no new loan at that time was given. The unusual high level of interest rate in year )**5 is deemed to be able to distort the calculation of -.T!c optimal cut#off score. IDENTIFICATION AND OPERATIONAL DEFINITION OF VARIABLES OF PREDICTION MODEL This research uses two types of variables that are dependent variables and independent variables. The dependent variables are categories (non#metric, that are bankrupt companies and non#bankrupt companies. The independent variables used are metric variables that are financial ratios used by <u and Penman ()*5*, 0achfoed1 ()**2, and !vianti (3444,. Disted below are the financial ratios used. T#)*e 5, Fin#nci#* %# i!" &"e/ ! )&i*/ 1e !ne-(e#% '%i!% ! )#n$%&' c( '%e/ic i!n 0!/e*
No. 1$ 2$ 3$ 4$ 5$ 6$ %$ 8$ "$ 10$ 11$ 12$ 13$ 14$ 15$ 16$ 1%$ 18$ 1"$ 20$ 21$ Financial Ratios Short Term Li ui!it" #I$ Cash to c)rrent '0a+0'0t0es !D1# Cash A'o2 to c)rrent '0a+0'0t0es !D2# E)0c3 assets to c)rrent '0a+0'0t0es !D3# C)rrent assets to c)rrent '0a+0'0t0es !D4# Lon% Term Solvenc" #II$ C)rrent assets to tota' '0a+0'0t0es !D5# Net 2orth and 'on; ter( de+t to A0xed assets !D6# Net 2orth to A0xed assets !D%# &ro'itabilit" #III$ .perat0n; 0nco(e to =arn0n;s +eAore taxes !D8# =arn0n;s +eAore taxes to sa'es !D"# Cross proA0t to sa'es !D10# .perat0n; 0nco(e to sa'es !D11# Net 0nco(e to sa'es !D12# &ro!uctivit" #I($ In entory to 2or30n; cap0ta' !D13# Cost oA ;oods so'd to 0n entory !D14# Sa'es to F)0c3 assets !D15# Sa'es to cash !D16# Sa'es to acco)nts rece0 a+'es !D1%# Cash A'o2 to tota' assets !D18# C)rrent assets to tota' assets !D1"# E)0c3 assets to 0n entory !D20# In entory to sa'es !D21#
Abbreviation CC? C8C? E&C? C&C? C&T? N>?T*8& N>8& .IN=BT =BTS C9S .IS NIS I >C C.CSI SE& SC S&R C8T& C&T& E&I I S
11"
SESI
ST& >CT& T?C& .IT? C?T& C8T? S8& >CT& C&S E&T& N>S >CS I T& C8S N>T& C?I T?T& =BTN> NI8& NIN> =BTT& NIT& SC? NIT? C?N> N>T?
22$ 23$ 24$ 25$ 26$ 2%$ 28$ 2"$ 30$ 31$ 32$ 33$ 34$ 35$ 36$ 3%$ 38$ 3"$ 40$ 41$ 42$ 43$ 44$ 45$ 46$ 4%$
Sa'es to tota' assets !D22# >or30n; cap0ta' to tota' assets !D23# In!ebt ness #($ Tota' '0a+0'0t0es to c)rrent assets !D24# .perat0n; 0nco(e to tota' '0a+0'0t0es !D25# C)rrent '0a+0'0t0es to tota' assets !D26# Investment Intensiveness #(I$ Cash A'o2 to tota' '0a+0'0t0es !D2%# Sa'es to A0xed assets !D28# >or30n; cap0ta' to tota' assets !D2"# C)rrent assets to sa'es !D30# E)0c3 assets to tota' assets !D31# Net 2orth to sa'es !D32# >or30n; cap0ta' to sa'es !D33# In entory to tota' assets !D34# Cash A'o2 to sa'es !D35# Levera%e #(II$ Net 2orth to tota' assets !D36# C)rrent '0a+0'0t0es to 0n entory !D3%# Tota' '0a+0'0t0es to tota' assets !D38# Return on Investment #(III$ =arn0n;s +eAore taxes to net 2orth !D3"# Net 0nco(e to A0xed assets !D40# Net 0nco(e to net 2orth !D41# =arn0n;s +eAore taxes to tota' assets !D42# Net 0nco(e to tota' assets !D43# E uit" #I)$ Sa'es to c)rrent '0a+0'0t0es !D44# Net 0nco(e to tota' '0a+0'0t0es !D45# C)rrent '0a+0'0t0es to net 2orth !D46# Net 2orth to tota' '0a+0'0t0es !D4%#
THE BANKRUPTCY PREDICTION MODEL The bankruptcy prediction model is built by using the two#group discriminant analysis because the dependent variables are non#metric and the independent variables are metric. $) is tested by performing the ?ilks Dambda test which is a test of e"uality of group means of financial ratios of both groups (bankrupt and non#bankrupt,. This test is performed to examine whether distinguishing ratios exist between the two groups. The significance level in this research is @G for the / critical value which means that financial ratios having significance level under @G are ratios that distinguish between bankrupt groups and non#bankrupt groups. If any financial ratio does have a significance value under @G then a discriminant function can be developed. 6ubse"uently the financial ratios that distinguish between bankrupt and non#bankrupt groups are selected to come up with the best variate (linier combination, by applying the stepwise selection algorithm. /inancial ratios that constitute the best linier combination are the final financial ratios that are going to be used as independent variables in the discriminant function which is the bankruptcy prediction model. The prediction using the discriminant function is designed to test $3. DESCRIPTON OF HAIR6 e #*, OPTIMUM CUTTING SCORE ! cutting score is necessary in performing prediction if using discriminant analysis. ! cutting score is the criterion (score, against which individuals discriminant score is &udged to determine into which group the individual should be classified. Those entities whose - scores are below this score are assigned to one group while those whose scores are above it are classified in the other group ($air et al. )**5,. !ccording to $air et al. ()**5, the optimal cutting score will differ depending on whether the si1es of the groups are e"ual or une"ual. 6ince this research applies the matched pair sampling method by which
120
SESI
the number of members in each group are e"ual thus the formula to compute the cutting score or the $air et al. optimum cutting score is as follow7 ZA + ZB ZCE = 2 where -C. H Critical cutting score value for e"ual group si1es -! H Centroid for group ! (bankrupt, -8 H Centroid for group 8 (non#bankrupt, Centroids of each group (-! and -8, is computed by calculating the average of the - discriminant scores of each group. DESCRIPTION OF ZETAc OPTIMAL CUT-OFF SCORE -.T!c optimal cutoff score is obtained from the following formula7 -.T!c ='n where ") H Prior probability of bankruptcy "3 H Prior probability of non#bankruptcy Ci H cost of type I error Cii H cost of type II error Prior probability of bankruptcy ("), is computed by dividing the number of bankrupt companies with the total companies listed in F6I for each year. Prior probability of non#bankruptcy (" 3, is computed by dividing the number of non#bankrupt companies with the total companies for each year. 8oth prior probabilities of bankruptcy and non#bankruptcy are computed for years of )**9 )**+ )*** and 3444. Type I error is analogous to that of an accepted loan that defaults and the type II error to a re&ected loan that would have resulted in a successful payoff. Thus Type I and type II errors (C i and Cii, are computed by the following formulas7
C0 &(o)nt oA 'oan 'osses reco ered =1 7 Cross 'oan 'osses !char;ed 7 oAA#
F1C0 F2C00
Type II (Cii, is computed by the following formula7 Cii HrJi r H effective interest rate on the loan which is computed as follow7 ?R ?8 r = x I?R + x I?8 T? T? C% H Doans given in %upiah C/ H Doans given in /oreign currencies (after converted into %upiah, IC% H !verage interest rate of loans given in %upiah IC/ H !verage interest rate of loans given in /oreign currencies TC H Total credit distributed (Total loans given in %upiah and /oreign currencies,. i H effective opportunity cost for the bank (one year average of the >4 days Central 8ank Certificate interest rate for years )**9 )**+ )*** and 3444,. The -.T!c optimal cut#off score gives a greater weight on type I error than type II error and prior probabilities of non#bankruptcy than prior probability of bankruptcy. Type I error is given a greater weight compared to type II error because type I error is more costly than type II error. Prior probability of non# bankruptcy is also given a greater weight than prior probability of bankruptcy since the probability of a company to file for bankruptcy is much more smaller than the probability of a company to not file for bankruptcy. In the formula above ") is multiplied by Ci to incorporate the cost of type I error with the probability of bankruptcy to understand the probability of type I error to occur. /urthermore "3 is also multiplied by Cii to incorporate the cost of type II error with the probability of non#bankruptcy to understand the probability of type II error to occur. In view of that -.T! c optimal cut#off score is a cut#off score used in bankruptcy prediction based on the consideration of probabilities of occurrences of type I and II errors.
121
SESI
Comparison of the one#year prior to bankruptcy prediction results between predictions using $air et al. optimum cutting score and predictions using -.T!c optimal cut#off score is designed to test $>. The prediction results between predictions using $air et al. optimum cutting score and predictions using -.T!c optimal cut#off score will be different. The differences of the prediction results are difference in the occurrences of type I and type II errors by each prediction using different cut#off scores. Therefore $2 can be tested. RESEARCH FINDINGS =escribed below are the research findings that are organi1ed in a systematically order. The findings are described based on the steps first conducted in the research. TEST OF E7UALITY OF GROUP MEANS OF FINANCIAL RATIOS OF BOTH GROUPS 3BANKRUPT AND NON-BANKRUPT4 %esult of the test of e"uality of group means using ?ilks Dambda showed that 33 financial ratios among the 2+ financial ratios tested have significance levels under @G as shown in table 2 below. 8ased on the test result below $) that indicated that distinguishing ratios exist between bankrupt and non#bankrupt groups is accepted. T#)*e 8, Si-nific#n *( /iffe%en fin#nci#* %# i!" )e 9een )#n$%&' #n/ n!n-)#n$%&' -%!&'"
(ariables
Cash to c)rrent '0a+0'0t0es !D1# E)0c3 assets to c)rrent '0a+0'0t0es !D3# C)rrent assets to c)rrent '0a+0'0t0es !D4# C)rrent assets to tota' '0a+0'0t0es !D5# =arn0n;s +eAore taxes to sa'es !D"# Cross proA0t to sa'es !D10# Net 0nco(e to sa'es !D12# C)rrent assets to tota' assets !D1"# >or30n; cap0ta' to tota' assets !D23# Tota' '0a+0'0t0es to c)rrent assets !D24# .perat0n; 0nco(e to tota' '0a+0'0t0es !D25# C)rrent '0a+0'0t0es to tota' assets !D26# >or30n; cap0ta' to tota' assets !D2"# E)0c3 assets to tota' assets !D31# Net 2orth to tota' assets !D36# Tota' '0a+0'0t0es to tota' assets !D38# Net 0nco(e to A0xed assets !D40# =arn0n;s +eAore 0nco(e taxes to tota' assets !D42# Net 0nco(e to tota' assets !D43# Sa'es to c)rrent '0a+0'0t0es !D44# Net 0nco(e to tota' '0a+0'0t0es !D45# Net 2orth to tota' '0a+0'0t0es !D4%#
*ilks+ Lamb!a 0$846 0$845 0$%"8 0$80" 0$"24 0$"2% 0$"24 0$"0" 0$%48 0$%65 0$868 0$%5" 0$%48 0$886 0$4"4 0$4"8 0$"04 0$%61 0$%1" 0$884 0$81" 0$%85
F "$801 "$883 13$6"5 12$%%5 4$445 4$258 4$445 5$434 18$213 16$5" 8$205 1%$18 18$213 6$"44 55$346 54$401 5$%54 1%$005 21$082 %$06" 11$"26 14$%58
Si%. 0$003 0$003 0$001 0$001 0$04 0$044 0$04 0$024 0 0 0$006 0 0 0$011 0 0 0$02 0 0 0$01 0$001 0
STEPWISE SELECTION ALGORITHM The stepwise selection algorithm is performed in order to obtain the best variate (linier combination, from the 33 distinguishing financial ratios above which will be used in the linier discriminant function. %esults of the stepwise selection algorithm showed that only two financial ratios are best to be used as independent variables in the discriminant function that are net worth to total assets (I>9,' and net worth to total liabilities (I2+,. DISCRIMINANT FUNCTION BUILT

122
SESI
The Canonical =iscriminant /unction Coefficient obtained for the one#year prior to bankruptcy prediction can be seen in table @ below. T#)*e :, C#n!nic#* Di"c%i0in#n F&nc i!n C!efficien
D36 D4% !Constant# 8)nct0on 1 10$086 7$824 72$060
Unstandardi1ed coefficients The one#year prior to bankruptcy prediction model built is as follow7 Z ; -<,=>=? +=,=@>A5> B =,@<8A8. where H =iscriminant index (classification score, I>9 H net worth to total assets I2+ H net worth to total liabilities
COMPUTATION OF HAIR6 ET AL, OPTIMUM CUTTING SCORE In exhibit ) the - score for each company has been computed thus the centroids or the average of - scores for each group can be computed as follows7 Centroid for bankrupt group H #).32+ Centroid for non#bankrupt group H ).)+@ 7 1$24% +1$1%5 $air et al. optimum cutting score = H -=,=5> 2 COMPUTATION OF ZETAC OPTIMAL CUT-OFF SCORE Table 9 below shows the computation of prior probability of bankruptcy ("), and prior probability of non#bankruptcy ("3,. T#)*e >, C!0'& # i!n !f P%i!% P%!)#)i*i ( !f B#n$%&' c( 3C+4 #n/ N!n-B#n$%&' c( 3C<4
,ear
NonTotal bankrupt Companies Companies Companies 210 212 225 230 211 22% 234 254
Bankrupt
1""6 1 1""% 15 1""" " 2000 24 & era;e oA F1 and F2
0$004%4 0$06608 0$03846 0$0"44" 0$050"43
0$""526 0$"33"2 0$"6154 0$"0551 0$"4"058
Table 9 shows that prior probability of non#bankruptcy is much more greater compared to prior probability of bankruptcy. There is a @.4*2> G of probability of bankruptcy to happen while the probability of a company to not file for bankruptcy is *2.*4@5 G. Table + below shows the type I and type II errors in metric forms happening in year )**9 )**+ )*** and 3444 in 8ank 'iaga 8ank =anamon 8ank 8'I and 8ank 8%I. T#)*e ., T('e I #n/ ('e II e%%!%" in 0e %ic f!%0" !f e#c1 )#n$ in (e#%" !f +DD>6 +DD.6 +DDD6 #n/ <===
Name o' Banks
,ear
Ci
Cii
123
SESI
0$"41 1$000 0$%4" 0$618 0$838 1$000 0$%48 0$%38 0$""" 0$"54 0$""4 0$"83 0$""% 0$"85 0$"%" 0$%38 /.010$012 0$02" 70$003 0$056 0$045 0$081 70$002 0$045 70$084 70$050 70$00% 0$016 70$015 0$014 0$036 0$065 /./-2
Ban3 N0a;a Ban3 *ana(on Ban3 BNI Ban3 BRI Ban3 N0a;a Ban3 *ana(on Ban3 BNI Ban3 BRI Ban3 N0a;a Ban3 *ana(on Ban3 BNI Ban3 BRI Ban3 N0a;a Ban3 *ana(on Ban3 BNI Ban3 BRI & era;e oA C0 dan C00
1""6 1""6 1""6 1""6 1""% 1""% 1""% 1""% 1""" 1""" 1""" 1""" 2000 2000 2000 2000
/rom table + above how costly type I error is compared to type II error can be computed by dividing the average value of Ci with the average value of Cii hence the result is :D,@5 (4.5*):4.4)@,. !ccordingly the cost incurred by the occurrence of type I error is @*.5> times more than the cost incurred by the occurrence of type II error. The considerable cost that must be incurred if prediction errors occur moreover if type I error occurs reason for the determination of a new cut#off score that is -.T! c optimal cut#off score. Therefore the -.T!c optimal cut#off score where the calculation will be elaborated below is more tight in predicting meaning that more companies will be predicted as bankrupt while less companies will be predicted as non#bankrupt compared to the prediction result that utili1es $air et al. optimum cutting score. In other words prediction using -.T!c optimal cut#off score is more conservative and more cost effective compared to prediction using $air et al. optimum cutting score. =escription of the calculation of ")Ci and "3Cii for each bank and each year of )**9 )**+ )*** and 3444 is shown in table 5 below. 8ased on table * below averages of " )Ci and "3Cii for years )**9 )**+ )*** and 3444 respectively are 4.429 and 4.4)2. !ccordingly -.T!c optimal cut#off score can be computed as
04046 'n e"ualling +,+D. 04014
T#)*e @, C#*c&*# i!n" !f C+Ci #n/ C<Cii f!% e#c1 )#n$ in (e#%" !f +DD>6 +DD.6 +DDD6 #n/ <===
Name o' Banks ,ear Ci Cii Ban3 N0a;a 1""6 0$"41 0$012 0$005 Ban3 *ana(on 1""6 1$000 0$02" 0$005 Ban3 BNI 1""6 0$%4" 70$003 0$005 Ban3 BRI 1""6 0$618 0$056 0$005 Avera%e o' -Ci !an .Cii 'or "ear -113 Ban3 N0a;a 1""% 0$838 0$045 0$066 Ban3 *ana(on 1""% 1$000 0$081 0$066 Ban3 BNI 1""% 0$%48 70$002 0$066 Ban3 BRI 1""% 0$%38 0$045 0$066 Avera%e o' -Ci !an .Cii 'or "ear -116 Ban3 N0a;a 1""" 0$""" 70$084 0$038 Ban3 *ana(on 1""" 0$"54 70$050 0$038 Ban3 BNI 1""" 0$""4 70$00% 0$038 Ban3 BRI 1""" 0$"83 0$016 0$038 Avera%e o' -Ci !an .Cii 'or "ear -111 Ban3 N0a;a 2000 0$""% 70$015 0$0"4 Ban3 *ana(on 2000 0$"85 0$014 0$0"4
.
0$""5 0$""5 0$""5 0$""5 0$"34 0$"34 0$"34 0$"34 0$"62 0$"62 0$"62 0$"62 0$"06 0$"06
-C i 0$004 0$005 0$004 0$003 /.//4 0$055 0$066 0$04" 0$04" /./22 0$038 0$03% 0$038 0$038 /./50 0$0"4 0$0"3
.Cii 0$012 0$02" 70$003 0$056 /./.5 0$042 0$0%6 70$002 0$042 /./51 70$081 70$048 70$00% 0$016 -/./5/ 70$013 0$013
124
SESI
0$"06 0$"06 0$0"2 0$0%0 /./06 0$032 0$05" /./.5
Ban3 BNI Ban3 BRI
2000 0$"%" 0$036 0$0"4 2000 0$%38 0$065 0$0"4 Avera%e o' -Ci !an .Cii 'or "ear .///
T#)*e D, A2e%#-e" !f C+Ci #n/ C<Cii f!% (e#%" +DD>6 +DD.6 +DDD6 #n/ <===
,ear 1""6 1""% 1""" 2000 Avera%e Ci 0$004 0$055 0$038 0$08% /./43
-
Cii 0$023 0$03" 70$030 0$023 /./-4

.
PREDICTION RESULTS USING BOTH HAIR6 e #*, OPTIMUM CUTTING SCORE AND ZETA C OPTIMAL CUT-OFF SCORE The one year prior to bankruptcy prediction using either $air et al. optimum cutting score or -.T!c optimal cut#off score was performed by means of the Kif functionL. 8ankrupt companies were given a 4 (1ero, code while non#bankrupt companies were given a ) (one, code. /or predictions using $air et al. optimum cutting score the Kif functionL made was if the - discriminant score of a company is smaller than J 4.4>9 then that company will be categori1ed in the bankrupt group (code 4, and if incorrect will be categori1ed in the non#bankrupt group (code ),. !s for the prediction using -.T! c optimal cut#off score the Kif functionL made was if the - discriminant score of a company is smaller than ).)*5 then that company will be categori1ed in the bankrupt group (code 4, and if incorrect will be categori1ed in the non#bankrupt group (code ),. In table )4 and )) below prediction results using $air et al. optimum cutting score and -.T! c optimal cut#off score are shown respectively (see exhibit ) for details of prediction results,. Table )4 elucidates that 35 bankrupt companies from the overall of 3* bankrupt companies were correctly classified and 32 non#bankrupt companies from the overall of 3* non#bankrupt companies were correctly classified. Thus the canonical linier discriminant model using $air et al. optimum cutting score predicted bankrupt companies as many as >> companies and non#bankrupt companies as many as 3@ companies. It also can be concluded that ) type I error and @ type II errors occurred from the prediction. T#)*e +=, P%e/ic i!n %e"&* " 9i 1 H#i%6 e #*, !' i0&0 c& in- "c!%e
Classi'ications &re!icte! 7roup 8embership Bankrupt Bankrupt Count Non-bankrupt Bankrupt 9 Non-bankrupt .0 2 13.2 -6.. Non-bankrupt .4 5.2 0..0 Total .1 .1 -//:/ -//:/
Table )) elucidates that 3* bankrupt companies from the overall of 3* bankrupt companies were correctly classified and )@ non#bankrupt companies from the overall of 3* non#bankrupt companies were correctly classified. Thus the canonical linier discriminant model using -.T!c optimal cut#off score predicted bankrupt companies as many as 2> companies and non#bankrupt companies as many as )@ companies. It also can be concluded that no type I error and )2 type II errors occurred from the prediction. T#)*e ++, P%e/ic i!n %e"&* " 9i 1 ZETAc !' i0#* c& -!ff "c!%e
Classi'ications &re!icte! 7roup 8embership Bankrupt
Non-bankrupt
Total
125
SESI
.1 -4 -//./ 40.5 / -2 /./ 2-.6 .1 .1 -//./ -//./
Bankrupt Count Non-bankrupt Bankrupt 9 Non-bankrupt
It is understood from the prediction tables above (tables )4 and )), that predictions using $air et al. optimum cutting score is more accurate than predictions using -.T! c optimal cut#off score. $owever predictions using $air et al. optimum cutting score will imperil more because by putting the same weight between type I error and type II error will result in the same opportunity towards the occurrence of type I error and type II error. !lthough more misclassifications were made in predictions using -.T! c optimal cut# off score it is more safe or conservative because the chance of a certain company to be predicted as bankrupt becomes greater (from >> companies predicted as non#bankrupt using $air et al. optimum cutting to 2> companies predicted as bankrupt using -.T!c optimal cut#off score, and the chance of a certain company to be predicted as non#bankrupt becomes smaller (from 3@ companies predicted as non#bankrupt using $air et al. optimum cutting to )@ companies predicted as bankrupt using -.T!c optimal cutoff score,. Therefore $3 that indicated that financial ratios could be used to predict bankruptcy before hand is accepted based on both prediction results using $air et al. optimum cutting score (35 bankrupt companies from the overall of 3* bankrupt companies were correctly predicted and 32 non#bankrupt companies from the overall of 3* non#bankrupt companies were also correctly predicted, and -.T!c optimal cut#off score (all bankrupt companies were correctly predicted and )@ non#bankrupt companies from the overall of 3* non#bankrupt companies were also correctly predicted, $> indicated that the percentage of predicting bankrupt companies will become larger by giving a greater weight on type I than type II errors compared to the percentage of predicting bankrupt companies of a prediction that gives the same weight on type I and type II errors. Predictions using $air et al. optimum cutting score that puts the same weight between type I errors and type II errors predict bankrupt companies as many as >> companies while predictions using -.T!c optimal cut#off score that puts a greater weight on type I errors as much of @* 5> times more costly than type II errors predict bankrupt companies as many as 22 companies. Conse"uently $> is also accepted. $2 indicated that predictions using -.T!c optimal cut#off score would cut expected costs incurred from prediction errors. The cost cut even though the prediction results were not as accurate as prediction results using $air et al. optimum cutting score can be seen from the success of -.T! c optimal cut#off score in deleting ) type I error that occurred in the prediction using $air et al. optimum cutting score. 6ince type I error is @* 5> times more costly compared to type II error -.T! c optimal cut#off score that successfully deleted the type I error even though only as many as ) type I error compared to $air et al. optimum cutting score is better to use in predicting bankruptcy in terms of more cost effective. !lthough predictions using -.T!c optimal cut#off score result in more type II errors compared to predictions using $air et al. optimum cutting score (from @ type II errors using $air et al. optimum cutting score to )2 type II errors using -.T!c optimal cut#off score, predictions using -.T!c optimal cut#off score is still more cost effective than predictions using $air et al. optimum cutting score. The cost cut incurred in use of -.T!c optimal cut#off score compared to the use of $air et al. optimum cutting score can be seen as follows7
-.T!c optimal cutoff score Type I error Type II errors Type I error H4 H )2 H @*.5> type II error
$air et al. optimum cutting score Type I error H) Type II errors H @
126
SESI
Cost cut
H @*.5> type II error J * type II errors H @4.5> type II errors
8ased on the calculation above the use of -.T!c optimal cut#off score in comparison with the use of $air et al. optimum cutting score will cut costs as much as @4.5> times of the cost of type II error. /or this reason predictions of one#year prior to bankruptcy using -.T! c optimal cut#off score will cut expected costs that may occur from predictions errors and is more conservative in performing predictions. Thus $ 2 has been accepted. CONCLUSIONS In this research four hypotheses have been developed and tested. The conclusions are as follows7 ). /inancial ratios do in fact differ between bankrupt and non#bankrupt companies. The ?ilks Dambda statistical test proved that 33 ratios differ significantly in the level of significance of @G between bankrupt and non#bankrupt companies. 3. Two financial ratios from the overall of 33 financial ratios that distinguish bankrupt and non# bankrupt companies can be used as variables of predictions in the one#year prior to bankruptcy prediction model. The two financial ratios belong to the Deverage and ."uity group which are the net worth to total assets ratio (I>9, and the net worth to total liabilities ratio (I2+,. >. Predictions using $air et al. optimum cutting score correctly predicted 35 bankrupt companies from the overall of 3* bankrupt companies and 32 non#bankrupt companies from the overall of 3* non#bankrupt companies. !s predictions using -.T!c optimal cut#off score correctly predicted all bankrupt companies and )@ non#bankrupt companies from the overall of 3* non#bankrupt companies. Thus it is proven that financial ratios can be used to predict bankruptcy beforehand. 2. Predictions using -.T!c optimal cut#off score that puts a greater weight on type I error as much as @*.5> times more costly than type II error and a greater weight on prior probability of non# bankruptcy (*2.*4@5 G, than prior probability of bankruptcy (@.4*2> G, predicted bankrupt companies as many as 2> companies from the overall sample of @5 companies and predicted )@ non#bankrupt companies from the overall sample of @5 companies. !s predictions using $air et al. optimum cutting score predicted bankrupt companies as many as >> companies from the overall sample of @5 companies and predicted 3@ non#bankrupt companies from the overall sample of @5 companies. /rom the one#year prior to bankruptcy prediction results it can be concluded that the use of -.T! c optimal cut#off score although not as accurate as predictions using $air et al. optimum cutting score is still better to use in terms of more conservative. In view of that a chance of a certain company to be predicted as bankrupt becomes greater and as non#bankrupt becomes smaller. In other words the probabilty of occurrences of type I errors are minimali1ed. @. The use of -.T!c optimal cut#off score in comparison with the use of $air et al. optimum cutting score will cut costs as much as @4.5> times of the cost of type II error. /or this reason predictions of one#year prior to bankruptcy using -.T!c optimal cut#off score is more safe cost effective and conservative in performing predictions. RECCOMENDATIONS ). Up to now independent variables used to distinguish between bankrupt and non#bankrupt companies and then for performing bankruptcy predictions are merely from financial statement data published by companies. There is a great possibility that market data such as stock price market capitali1ation of a company macro#economic condition etc can be used to predict bankruptcy beforehand. /urther research is expected to incorporate market data in building bankruptcy prediction models. 3. The sample of bankrupt and non#bankrupt companies in building the one#year prior to bankruptcy prediction model is limited to companies listed on F6I in years of )*** and 3444. /urther research

12%
SESI
>.
2.
is expected to use wider range of pooled data so that the model built better represents the real condition of the population. The sample of banks in the computation of -.T!c optimal cut#off score only uses two state#owned banks and two private owned banks. It is expected that a specific research is conducted to investigate the cost occurring from type I errors and type II errors so that the computation of -.T!c optimal cut#off score becomes more accurate as conducted by !ltman ()*++b, in his exceptional research K/ending error costs for commercial banks* 0ome conceptual and empirical issues' 1ournal of &ommercial 2ank /ending <ctoberL. This research did not use a hold out sample because of time consideration and lack of data. /urther research is expected to use a hold out sample (a different set of data than the data used in building the bankruptcy prediction model, in order to validate the model and cut#off score built with the intention that the results of the research also becomes more valid. REFERENCES
!ltman ..I %. $aldeman and P. 'arayanan. KZ T! !nalysis* ! new model to identify bankruptcy risk of corporations3 Fournal of 8anking and /inance Fune)**+ pp. 3*#@2. !vianti Ilya. L4odel 5rediksi 6epailitan miten di 2ursa fek 1akarta 7engan 4enggunakan IndikatorIndikator 6euangan3 =isertasi 6#> Program Pasca 6ar&ana Universitas Pad&ad&aran 8andung 3444. 8aridwan -aki. Intermediate !ccounting. .disi Eetu&uh. 8P/. Yogyakarta )**3. 8ernstein Deopold !. Fohn F. ?ild. 8inancial 0tatement !nalysis* Theory' !pplication' and Interpretation. 6ixth edition. 0cBraw#$ill International )**5. /!68 0tatement of 8inancial !ccounting &oncepts 9o.)#<b&ectives of /inancial %eporting by 8usiness .nterprises ()*+5,. /oster B. 8inancial 0tatement !nalysis. 6econd edition. Prentice J$all Inc. )*59 $air Fr. Foseph /. %olph .. !nderson %onald D. Tatham and ?illiam C.8lack 4ulti(ariate 7ata !nalysis /ifth .dition Prentice $all International Inc )**5. Ikatan !kuntan Indonesia. 0tandar !kuntansi 6euangan per # !pril -::-. 6alemba .mpat. 3443. Disetyati .ni. !nalisis Daporan Eeuangan 6ebagai !lat Prediksi Eebangkrutan 8ank. Tesis 6#3 Program Pasca 6ar&ana Universitas Bad&ah 0ada Yogyakarta 3444. Distari Putu. K!nalisa ;asio 6euangan 0ebagai !lat 5rediksi 6ebangkrutan3 6kripsi %eguler !kuntansi /akultas .konomi Universitas Bad&ah 0ada 3443. <hlson F. !. K8inancial ;atios and the 5robabilistic 5rediction of 2ankruptcy3' Fournal of !ccounting %esearch 6pring )*54 pp. )4*#)>). ?atts %oss Dand Ferold D. -immerman 5ositi(e !ccounting Theory 'ew Fersey7 Prentice $all International )*59. ?ilkinson Foseph ?. !ccounting Information 0ystems* ssential &oncepts and !pplications Fohn ?iley A 6ons /ourth edition 3444.
E;hibit Hair, et al C!/e #Constan -/./03 t$ )53 -/.0.4 )46 < ZETAc
optimum optimal cutting cutoff

128
SESI
score 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 1 0 1 1 1 1 1 1 0 1 1 1 1 1 1 1 1 1 0 0 1 1 1 1 1 score 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 1 0 0 0 1 1 1 1 0 1 0 1 0 1 0 0 1 0 0 0 1 1 0 1 1
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
724060 724060 724060 724060 724060 724060 724060 724060 724060 724060 724060 724060 724060 724060 724060 724060 724060 724060 724060 724060 724060 724060 724060 724060 724060 724060 724060 724060 724060 724060 724060 724060 724060 724060 724060 724060 724060 724060 724060 724060 724060 724060 724060 724060 724060 724060 724060 724060 724060 724060 724060 724060 724060 724060 724060
040"% 04036 04128 04083 04002 04144 0424" 04144 041%5 04015 04102 041%3 04086 04100 04065 04036 0400" 04065 04042 04086 0400% 040"5 04134 0408% 04085 04001 040"2 04154 040"1 04%31 04385 04163 04304 042"% 04522 04802 0484" 044%% 04058 046"0 04305 04533 04241 045"6 0422" 04253 04836 04880 04186 04135 04641 043%1 0426% 04505 04502
04108 0403% 0414% 040"0 04002 04168 04332 04168 04212 04015 04114 0420" 040"4 04111 0406" 0403% 0400" 0406" 04043 040"4 04008 04104 04154 040"5 040"3 04001 04101 04181 040"8 24%18 04626 041"5 0443% 04422 140"2 44043 54636 04"11 04061 24223 04438 14141 0431% 144%% 042"8 0433" 54084 %4320 0421% 04156 14%82 045"0 04364 14020 1400%
71416% 714%26 7048"1 714301 724045 704%50 04180 704%51 7044%1 714"22 714121 70448" 7142%0 714146 714462 714%28 714"%% 714464 7146%6 7142%3 714""1 7141"3 70483" 714260 7142%8 724051 714218 704661 714218 340%4 14308 7045%2 04645 04585 24305 246"4 14862 14""8 71452% 34065 04653 243%5 04108 24%3% 04008 04214 241%" 04%82 704365 704826 24"32 141"% 04332 241"3 241%1

12"
SESI
1 0 1 0 0 1
1 1 1
724060 724060 724060
04355 04145 04424
04552 041%0 04%35
140%1 704%34 14608
0 = +an3r)pt 1 = non7+an3r)pt

130

Bankruptcy Prediction Model With Zeta Optimal Cut-Off Score To Correct Type I Errors

Încărcat de

Informații document

Descriere originală:

Titlu original

Drepturi de autor

Formate disponibile

Partajați acest document

Partajați sau inserați document

Opțiuni de partajare

Vi se pare util acest document?

Este necorespunzător acest conținut?

Drepturi de autor:

Formate disponibile

Bankruptcy Prediction Model With Zeta Optimal Cut-Off Score To Correct Type I Errors

Încărcat de

Drepturi de autor:

Formate disponibile

BANKRUPTCY PREDICTION MODEL WITH ZETAc OPTIMAL CUT-OFF SCORE TO CORRECT TYPE I ERRORS

SIMPOSIUM NASIONAL AKUNTANSI VI

SIMPOSIUM NASIONAL AKUNTANSI VI

1""6 1 1""% 15 1""" " 2000 24 & era;e oA F1 and F2

0$004%4 0$06608 0$03846 0$0"44" 0$050"43

0$""526 0$"33"2 0$"6154 0$"0551 0$"4"058

SIMPOSIUM NASIONAL AKUNTANSI VI

Ban3 BNI Ban3 BRI

Cii 0$023 0$03" 70$030 0$023 /./-4

Bankrupt Count Non-bankrupt Bankrupt 9 Non-bankrupt

$air et al. optimum cutting score Type I error H) Type II errors H @

H @.5> type II error J type II errors H @4.5> type II errors

SIMPOSIUM NASIONAL AKUNTANSI VI

optimum optimal cutting cutoff

SIMPOSIUM NASIONAL AKUNTANSI VI

SIMPOSIUM NASIONAL AKUNTANSI VI

724060 724060 724060

04355 04145 04424

04552 041%0 04%35

140%1 704%34 14608

SIMPOSIUM NASIONAL AKUNTANSI VI

S-ar putea să vă placă și