| | SLO | ENG | Cookies and privacy

Bigger font | Smaller font

Show document

Title:Presojanje in napovedovanje insolventnosti za panogo gradbeništvo s pomočjo statističnih modelov
Authors:Šinko, Mitja (Author)
Oplotnik, Žan Jan (Mentor) More about this mentor... New window
Files:.pdf MAG_Sinko_Mitja_2016.pdf (4,27 MB)
 
Language:Slovenian
Work type:Master's thesis (m2)
Typology:2.09 - Master's Thesis
Organization:EPF - Faculty of Business and Economics
Abstract:Gradbeništvo v Sloveniji je doživelo pred krizo enormno rast. Rast so podpirali bančni finančni viri. Z nastopom krize so se začele težave na obeh koncih, ob zmanjševanju državnih investicij so se hkrati odtegovali tuji finančni viri. Posledično so krizo najbolj občutila velika, nefleksibilna in prekomerno zadolžena gradbena podjetja. Več kot dve tretjini velikih gradbenih podjetij ni uspelo niti restrukturirati finančnih obveznosti, saj jih je velika večina pristala v stečaju. Posledice stečajev velikih gradbenih podjetij čutijo tako majhni upniki – predvsem podizvajalci kot veliki upniki – banke in ostali finančni kreditorji. Nefinančni upniki v glavnem predstavljajo manjša ali srednje velika gradbena podjetja. Nekatere od njih je spiralo potegnilo v brezno stečajev. Banke so morale oblikovati ustrezne slabitve za dana posojila. Slabitve so sčasoma načenjale kapitalsko ustreznost. V izogib prihodnjim težavam je smiselno iz izkušenj in potrebe po stabilnem poslovnem odnosu vnaprej predvidevati razvoj prihodnjih dogodkov. Za predvidevanje finančnih težav je smiselna uporaba modelov napovedovanja insolventnosti / neplačila / neuspeha. Ti modeli temeljijo na računovodskih kazalnikih in mehkih dejavnikih. Ker se mehki dejavniki oziroma informacije povečini pridobivajo v poslovnem odnosu, smo se osredotočili na preučevanje javno dostopnih računovodskih informacij. Na vzorcu gradbenih podjetij, kjer smo izločili vsa takratna mikro podjetja in sedem največjih gradbenih podjetij v insolvenčnih postopkih, smo preverjali robustnost tradicionalnih statističnih modelov. Zanimala nas je klasifikacijska natančnost prepoznavnih tujih modelov multiple diskriminantne analize in manj znanih lokalnih modelov logistične regresije. Altmanovi MDA modeli Z-score dopuščajo določeno območje negotovosti, ki se je izkazalo še posebej veliko pri aplikaciji modela Z'-score, saj se je v njem znašlo kar 62 podjetij iz našega vzorca. Po eni strani to pomeni, da je iz previdnosti potrebna dodatna pozornost do teh podjetij, po drugi strani pa to zmanjšuje diskriminacijsko vlogo modela. Vendar se je model izkazal za zelo učinkovitega. Bolj natančen ob upoštevanju skupnega števila napačnih razvrstitev je bil le model logistične regresije od Širce. Slabše od pričakovanj se je izkazal model nepogojne logistične regresije Juričiča. Rezultati testiranja pa ne upoštevajo pomembnega dejavnika – neenakih stroškov napak. Pri enakih stroških napak namreč predpostavljamo, da bi bil strošek insolventnosti povprečnega dolžnika enak zaslužku pri poslovanju s povprečnim uspešnim partnerjem.
Keywords:gradbeništvo, insolventnost, presojanje insolventnosti, napovedovanje insolventnosti, statistične metode, Altman, multipla diskriminantna analiza, logistična regresija, stroški napak
Year of publishing:2016
Publisher:[M. Šinko]
Source:Maribor
UDC:519.2
COBISS_ID:12643868 Link is opened in a new window
NUK URN:URN:SI:UM:DK:RMGTCHMT
Views:550
Downloads:69
Metadata:XML RDF-CHPDL DC-XML DC-RDF
Categories:EPF
:
  
Average score:(0 votes)
Your score:Voting is allowed only for logged in users.
Share:AddThis
AddThis uses cookies that require your consent. Edit consent...

Hover the mouse pointer over a document title to show the abstract or click on the title to get all document metadata.

Secondary language

Language:English
Title:Insolvency Classification and Prediction for Construction Industry with Statistical Models
Abstract:Construction industry in Slovenia has experienced enormous growth before the crisis. Growth was supported by the bank financial resources. With the onset of the crisis, beside reducing government investment also foreign financial resources have deducted. Consequently, the crisis being felt most by large, inflexible and over-responsible construction companies. More than two-thirds of large construction companies failed to restructure financial liabilities, since the vast majority ended up in bankruptcy. Consequences of bankruptcies of large construction companies feel so small creditors - mostly subcontractors as large creditors - banks and other financial lenders. Non-financial creditors are mostly small or medium-sized construction companies. Some of them are spiral dragged into the abyss of bankruptcy. Banks should establish appropriate impairment of the loans. Impairment losses are eventually undermined the capital adequacy. To avoid future problems, it makes sense from the experience and the need for a stable business relationship in advance to anticipate future developments. To anticipate financial difficulties, it makes sense to use models to predict insolvency / default / failure. These models are based on financial indicators and soft factors. Since soft factors or information mostly acquired in a business relationship, we focused on the study of publicly available financial information. A sample of construction companies, where we eliminated all the then micro companies and seven of the largest construction companies in insolvency proceedings, we check the robustness of traditional statistical models. We were interested in the classification accuracy of recognizable foreign models of multiple discriminant analysis and less well-known local logistic regression models. Altman’s MDA Z-score models allows a certain area of uncertainty, which has proved particularly large in the application model Z'-score, because in him there is found 62 companies in our sample. On the one hand, this means that as a precautionary measure requires additional attention to these companies, on the other hand, this reduces the discriminatory role model. However, the model proved to be very effective. More accurate taking into account the total number of misclassifications was only unconditional logistic regression model of Širca. Worse than expected proved to be the unconditional logistic regression model by Juričič. Test results do not take into account an important factor - unequal error cost. In the same cost of errors is intuitively assume that the cost of insolvency of the debtor equal to the average earnings from operations by the average successful partners.
Keywords:construction industry, insolvency, insolvency classification and prediction, statistical methods, Altman, MDA, logistic regression, error costs


Comments

Leave comment

You have to log in to leave a comment.

Comments (0)
0 - 0 / 0
 
There are no comments!

Back
Logos of partners University of Maribor University of Ljubljana University of Primorska University of Nova Gorica