DETECTING FINANCIAL DISTRESS: DISCRIMINANT VERSUS LOGISTIC REGRESSION ANALYSIS
A thesis submitted to the Graduate School in partial fulfillment for the requirements f‘01
the degree
Master of Science (Finance) Universiti Utara Malaysia
Abd Halim (3 Hamilton bin A:mad
0 Abd Halim ((z Hamilton bin Ahnnad. 2003. All rights reserved
Sekolah Siswazah (Graduate School) Universiti Utara Malaysia
PERAKUAN KERJA KERTAS F’ROJEK (Certification of Projecf Paper)
Saya, yany bertandatangan, memperakukan bahawa I
(/, fhe undersigned, certify that) I
I ABD. HALIM @ HAMILTON AHMAD
calon untuk ljazah MSc (Finance) (candidate for the degree 00
telah mengemukakan kertas projek yang bertajuk
(has presented his/her project paper of fhe following title)
DETECTING FINANCIAL DISTRESS : DISCRIMINANT VERSUS LOGISTIC REGRESSION ANALYSIS
seperti yang tercatat di muka surat tajuk dan kulit kertas projek (as if appears on the title page and front colder of project paper)
bahawa kertas projek tersebut boleh diterima dari segi bentuk serta kandungan clan me!iputi bidang ilmu dengan memuaskan.
(that the p&f paper acceptable in form and content and that a satisfactory knowledge of the field is covered by the project paper).
Nama Penyelia : Prof. Madya Dr. Nur Adiana Hiau Abdullah (Name of Supervisor)
Tandatangan : (Signature)
Tarikh (Date)
PERMISSION TO USE
In presenting this this in partial fulfillment for the rtquirements ol‘ post y-aduatc degree from Uniw-siti Utara Malaysia. I agree that the IkJniversity Librar~~ may make it fI-eely available for inspection. I further agree that permksion for copying of‘ this thesis in any 17im1it’1~. in whole or in part. for scholarly p~~rposcs may be granted by my
supenisor or. in their abs~ncc. 1~1, the Dean of‘ the Graduate School. It is understood that any copy i Ii,(7 or publication or use of‘this this or parts tlweof’ for fimncial gain shall not be allon cd \\ ithoclt IN>’ L\ ritten pemission. It is also understood that dw recognition shall lx giwn to mc and to I hi\ orsiti Iltara h,lala\ 5ia lTor an! scliola~~l! IISC \\ hicli ma! bc made of an> material from III) thesis.
Requests fkw permission to cl.>p,\’ or to make other llsc of‘ materials in this thesis. in whole OI- in part. should be addressed to:
Dean of‘ Graduate School IJniwrsiti IJtara Malaysia
060 10 Sintok Kedah Dan11 Aman
ABSTRACT (BAI-IASA MALAYSIA)
Kajian ini rnenguji dua ujian statistik iaitu analisis diskriminan dan nmdcl logi/ untuk niengenalpasti kebarangkalian sesebuah syarikat berada d i dahn kctidakstabilan kewangan. Kaj ian ini juga menggunakan nisbah kewangan sebagai petun-i uk kepada ketidakstabilan kcwangan. Kcputusan darip;~da kajian ini menu~~jukkan model logi/
member-ikan ketepatan yang lebih tingi bchmding analisis diskriminan. Jlodel logil uncrnbcrikan ketepatan klasitikasi scbm~~al; 0 1 .$ pcratus dalam k~selu~~han sxnpcl dan
90 p e r a t u s dalam sampel kmvalan. hlanalala untuk analisis di,il;riniitian. ketepatan klasifikasi adalah 84.5 peratus untul; kcselu11111a11 samp~l dan 80 px~tus cmtul, san~pc!
kawalan. LJntuk analisis diskriminan terdapat tiga petunjuk d i dalam nmlcl yng nwmisahkan di antara syarikat Iwq stabil dan s\,arikat ymg tidak stabil dalam ktxanganr iaitu nisbah senma. pendapatan bersih k e p a d a jumlah aset da11 illalan lqada asset seniasa. M o d e l logi/ bjuga ~n~nu~~.j~~l;kan tiga f&or atau pctul$k dalam model
terakliirnya iaitu ekuiti peniqyng sallam kepada jumlah liabiliti.aliran tunai dai$~adn aktkiti pembiayaan kepada jumlah liabiliti dan pendapatan bersih kcpada jumlah aset.
Dapatan daripada kaj ian ini diharap dapat memberikan pcmahaman tentang faktor atau pcnentu kepada ketidkstabilan ktmangan dan seterusn! a s> arikat dapat mtxganhil langkali tertentu wtuk mengelakl;an lqagalan.
ii
ABSTRACT (ENGLISH)
This study examines two statistical tests. which are discriminant analysis and the logit model to predict the probability of financially distress companies. In addition. this stud!
also utilizes the usage 01‘ iinancial ratios as a predictor of a company in a state of financialI! distressed. The findings ho\\ that the logit model sho~\x better prc’diction accurac! ;lian the discriniinant analysis. The logit model correctly classifitx~ 0 1 .1 percent 01‘ the companies in the estimation sample and 90 percent for tk holdout sample.
I hwe~er. thl- discriminant mode!. the oxrall accuracy rate fix the estimation and the holdout samples are 84.5 percent and 80 percent respectively. For discriminant analysis.
there arc three factors found to hal,e significant discriminating poser: currcllt ratio. net income to total assets. and sales to current assets. Similarly. logit model also identified three fjctors but two of’ the factors (shareholders’ equity to total liabilities and cash flo\\
h-0111 financing to total liabilities) are different from those found in discriminant analysis.
The onl!! factor which is identified in both models is net income to total assets. The
findings @ve clear understanding of the relevant factors that can cause financial distress.
Hence. companies could take immediate actions to a\,oid failure to the cornpan!.
. . .
1 1 1
![The Pronunciation Accuracy of Interactive Dialog System for Malaysian Primary School Students - UUM Electronic Theses and Dissertation [eTheses]](data:image/gif;base64,R0lGODlhAQABAIAAAP///wAAACH5BAEAAAAALAAAAAABAAEAAAICRAEAOw==)