The relationship between household debt and interest rates: a South African perspective

Abstract

One of the challenges that South Africa has been facing lately is the high level of household debt. Since the dramatic drop from 12 to 5.5 percent in 2008, interest rates have been on a constant hiking cycle (SARB, 2015). The current repurchase rate stands at 6.75% and the prime rate at 10.25%. The volatile interest rate and its impact on the borrowing behaviour of households is something that needs to be investigated. The review of the theoretical and empirical literature on the demand for credit and interest rates revealed different results from authors. The theoretical literature underpinning the study is the loanable fund's theory, the life cycle hypothesis as well as the credit channel of the monetary transmission mechanism. The aim of this study was to determine the short-run, co-integrating and causal relationship between household debt and interest rates. A quantitative approach was chosen as the best method to measure the variables of interest and to deduce from the applicable theories. The study employed the autoregressive distributed lag (ARDL) approach to co-integration, ordinary least squares model for a deterministic relationship as well as the Wiener Granger-Causality test for bilateral causality between household debt and interest rates. The study used secondary data from the South African Reserve Bank for the period 1990Q1-2016Q4. The results from the regression analysis showed a negative but insignificant relationship between household debt and interest rates. The results from the ARDL revealed no short-run but only the long-run relationship between household debt and interest rates. The causality tests showed no causality between household debt and interest rates. The results between household debt and inflation, income, GDP, consumption, savings and wealth were also discussed. The implication of this study is that policymakers and credit regulators should be alert to the effects of policy-induced changes.

Nngwe ya dikgwetlho tse di lebaneng Aforikaborwa ke seelo se se kwa godimo sa dikoloto tsa mo malapeng. Fa e sale go tloga ka kwelotlase e e boitshegang ya seelo sa merokotso go tswa go 12% go ya go 5.5% ka 2008, seelo sa merokotso se ntse se lebile kwa godimo ka tsepamo (Monetary Policy Review, 2015). Seelo sa ga jaana sa go reka gape se eme mo go 6.75% mme seelo sa poraeme ke 10.25%. Seelo se se bogale sa merokotso le ditlamorago tsa sona mo mekgweng ya go adima ya malapa se tlhoka go tlhotlhomisiwa. Tshekatsheko ya dikwalo tsa tiori le tsa maitemogelo tse di ka ga topo ya sekoloto le seelo sa merokotso e senola gore bakwadi ba ba farologaneng ba fitlheletse ditshwetso tse di sa tshwaneng. Dikwalo tsa tiori tse di tshegetsang thutopatlisiso di totile tiori ya matlole a a adimisegang, haepotesese ya sediko sa botshelo le kanale ya sekoloto ya sedirisiwa sa tsamaiso ya madi. Maikaelelo a thutopatlisiso eno ke go lebelela kamano ya paka e khutshwane, paka e telele le sebako sa sekoloto sa magae le seelo sa merokotso. Go tlhophilwe molebo o o lebelelang dipalo gonne o tlaa nna botoka go letla mmatlisisi go lekanyetsa dipharologantsho tsa merokotso go bona dikarabo go tswa mo ditioring tse di dirisiwang. Thutopatlisiso e dirisitse molebo wa autoregressive distributed lag (ARDL) malebana le kopanyo ya paka e telele, sekao sa tlwaelo sa sekwere sennye malebana le kamano, le teko ya sebako ya Wiener–Granger malebana le sebako sa sebedi magareng ga sekoloto sa magae le seelo sa merokotso. Thutopatlisiso e dirisitse data ya bobedi go tswa kwa Bankeng ya Rasefe ya Aforikaborwa ya sebaka sa 1990Q1 go fitlha 2016Q4. Dipholo go tswa mo tokololong ya phokotsego di bontsha kamano e e senang bokao magareng ga sekoloto sa mo magaeng le seelo sa merokotso. Dipholo go tswa mo ARDL di bontsha go se nne gona ga botsalano jwa pakakhutshwane (k.g.r. ke jwa pakatelele fela) magareng ga sekoloto sa mo magaeng le seelo sa merokotso. Diteko tsa sebako di bontsha sebako magareng ga sekoloto sa mo magaeng le seelo sa merokotso. Ditlamorago tsa sekoloto sa mo magaeng le infoleišene, lotseno, tlhagiso yotlhe ya naga ya ditirelo le dithoto, tiriso, dipoloko le lehumo le tsona di a tlhagisiwa.

’n Hoë vlak van huishoudelike skuld is een van die uitdagings wat Suid-Afrika in die gesig staar. Ná die drastiese daling in rentekoerse van 12% tot 5,5% in 2008 was rentekoerse in ’n voortdurend opwaartse siklus (Monetary Policy Review, 2015). Tans is die terugkoopkoers 6,75% en die prima uitleenkoers 10,25%. Die onbestendige rentekoers en die uitwerking wat dit op huishoudings se leengedrag het, moet ondersoek word. ’n Oorsig van teoretiese en empiriese literatuur oor die vraag na krediet en rentekoerse toon dat verskillende navorsers tot verskillende gevolgtrekkings kom. Die teoretiese literatuur waarop hierdie studie gebaseer is, fokus op die leenbarefondse-teorie, die lewensiklus-hipotese en die kredietkanaalmeganisme vir geldoordrag. Die studie se doel was om die korttermyn-, koïntegrerende en kousale verbande tussen huishoudelike skuld en rentekoerse te bepaal. ’n Kwantitatiewe benadering is gevolg omdat dit die navorser in staat sou stel om renteveranderlikes te meet en antwoorde met behulp van toepaslike teorieë af te lei. Die studie het koïntegrasie ondersoek deur te kyk na outoregressief verspreide agterstande (ARDL). Die deterministiese verband is ondersoek aan die hand van die gewone kleinstekwadrate-model, en die Wiener–Granger-kousaliteittoets is uitgevoer om bilaterale kousaliteit tussen huishoudelike skuld en rentekoerse te bepaal. Die studie maak gebruik van sekondêre data van die Reserwebank vir die tydperk 1990V1 tot 2016V4. Die resultate van die regressie-ontleding toon ’n negatiewe maar onbenullige verband tussen huishoudelike skuld en rentekoerse. Die ARDL-resultate dui op geen korttermynverband tussen huishoudelike skuld en rentekoerse nie (dit dui dus slegs op langtermynverbande). Die kousaliteittoets toon dat daar kousaliteit tussen huishoudelike skuld en rentekoerse is. Die uitwerkings van huishoudelike skuld en inflasie, inkomste, bruto binnelandse produk, verbruik, spaar en welvaart word ook bespreek.