Ponnala, Venkati (PhD)Adugna, Sara2018-06-292023-11-042018-06-292023-11-042014-05http://etd.aau.edu.et/handle/123456789/4927This s tudy e xamined d eterminants of l oan por tfolios qual ity, us ing pane l dat a of fourteen (14) MFIs from the period 2003- 2012. The study employed four dependent variables as proxies for loan por tfolios qual ity, nam ely: l oan l oss r ate (LLR), por tfolio at r isk o ver 3 0-days ( PAR- 30days), por tfolio at r isk ov er 90 -days ( PAR-90days) and write-off r atio (WO R).This st udy i s crucial f rom t he fact t hat t here i s l imited r esearch on M FIs l oan por tfolios qual ity us ing quantitative appr oach i n E thiopia. Based on t he pool ed o rdinary l east squares (OLS) and random effects generalized least squares, the study finds an i nstitution size (LnTA) is negatively and s ignificantly i nfluences L LR and W OR. O perating e xpense (OPPEXP) has a ne gative significant r elationship with L LR, P AR-30days and W OR. A ge of t he M FIs ha s a pos itive significant r elationship with P AR-90days. P ercentage of women bor rowers (WOMBOR) has a positive s ignificant i mpact on P AR-30days. D eposit t o l oans (DTL) pos itively and s ignificantly influences LLR and WOR. Gross loan portfolio/ total asset (GLP/TA) has a negative significant impact on L LR and W OR. V oluntary s avings ( LnVOLSAV) has a ne gative s ignificant relationship with LLR, PAR-90days and W OR. The regression results also show that Return on equity ( ROE) has a ne gative s ignificant r elationship with L LR, P AR-30days, P AR-90days and WOR. Change in growth national income (CH-GNI) negatively and significantly influences PAR- 30days and P AR-90days. The s tudy finds i nsignificant r esults on l everage and i nflation. Si nce MFIs pr ovide f inancial s ervices t o t he poor pe ople t hose who c annot pr ovide c ollateral, t hus they m ay hi ghly f ace de fault r isk when bo rrowers f ail t o r epay t heir obl igations a s pe r t he agreement. A ccordingly, t he f indings of t he s tudy may ha ve i mplications f or M FIs and pol icy makers in that it provides hint on some important determinants of loan portfolios qualityen-USEmpirical evidenceQualityLoan PortfoliosThesis