International Funding Cost and Mortgage Interest Rate Transmission in Australia

• Main Author: Pham, Quynh
• Format: Text
• Language: English
• Published: Griffith University 2018
• Subjects: Australian mortgage market; Mortgage Interest Rate Transmission; International Funding Cost; Banking; Iceland
• Online Access: https://dx.doi.org/10.25904/1912/3414; https://research-repository.griffith.edu.au/handle/10072/385133

This document is constructed from four interrelated empirical papers examining the mortgage interest-rate pass-through mechanism in the Australian banking market over the period 1997:1–2015:12 employing an anonymous sample of 20 banks anonymised. In particular, the transmission from both the bank international funding cost and the cash rate to the effective interest rate on the variable home loans for owner-occupiers, both asymmetries (downward and upward) and symmetries (homogeneity or heterogeneity), are investigated in parallel in the short run and in the long run. Furthermore, the effects of the recent Great Recession (GFC) on this transmission mechanism are carefully considered in different manners while controlling for nonlinearity, slope heterogeneity, complex cross-sectional dependencies arising from both observed and unobserved common factors in panel data. The economic imperative of this examination focuses on seven key reasons: (i) the paramount importance of the Australian mortgage market to monetary policy transmission, housing and financial soundness in Australia, and to the stability of global banking and financial markets, (ii) the salient characteristics of the mortgage products, (iii) the high oligopoly of the mortgage market and banking sector, (iv) the heavy reliance of Australian banks on wholesale funding to finance their mortgages, (v) the increased public concern about bank mortgage rate-setting behaviour and a proliferated attention of regulators (Australian Prudential Regulation Authority [APRA], Reserve Bank of Australia [RBA] and related bodies) towards the enhancement of banking competition and financial soundness, (vi) the small quantity of the literature on the policy rate–mortgage rate transmission using micro-data, considering the substantially dire effect of the GFC on this asymmetric transmission in Australia, and (vii) a lack of knowledge about the role of bank international funding cost on mortgage rate pass-through mechanisms in Australia. The first empirical essay explores Australian mortgage interest-rate pass-through: does international funding cost matter? This study contributes significantly to the existing literature on the policy rate–mortgage rate association, with two key novelties. This is the first empirical analysis of the transmission from changes in international funding cost to mortgage interest rates at individual Australian banks and provides fresh convincing evidence for this cointegrating relationship. This essay is the first comprehensive analysis of the cash rate transmission to mortgage rates for Australia with regards to impact, short-term cumulative and long-term asymmetries. From a wider governing perspective, this bank-based interest-rate finding is supportive of the development of a comprehensive proxy for bank marginal funding cost, as opposed to the implementation of the cash rate. This study further contributes to the descriptive literature that argues ‘international funding cost affects bank mortgage rate-setting behaviour’ by providing new insights: that the significant positive asymmetries of mortgage rate adjustments to changes in the international funding cost and the cash rate exist in the long term for virtually all individual banks. The instant and short-term cumulative asymmetries have also occurred in several sampled banks, but they show wide discrepancies in size and sign. The long-term positive asymmetry and short-term heterogeneous asymmetry in the funding cost transmission provide empirical support to the market power of Australian banks in pricing. These findings of the first study suggest that, while the regulators revise and reform their banking governance regimes, they should seriously consider the key factor of bank market power, as well as bringing stronger customer protection procedures into effect. Finally, this essay explores three components, international funding cost, the cash rate, and mortgage rate pass-through, showing for the first time a strong interactive effect of foreign funding cost and the cash rate on mortgage rates at an individual level. The finding suggests that the prudential bodies, APRA and RBA, should carefully consider the international funding cost–mortgage rate nexus if they are striving against potential housing finance and economic fragility. In this study, I conclude that mortgage rate adjustments are being significantly adjusted in response to international funding cost shocks for individual banks, while the cash rate drives mortgage rates in both short term and long term. The nature of these transmission mechanisms is asymmetric in three dimensions: immediate, in the short run and in the long run. The second empirical essay, building on the first study, provides a comprehensive analysis of the asymmetric association between bank funding costs and mortgage rates, considering account the effect of the crisis and the bank-specific and time effects on this transmission. This initial study examines such a relationship in a small open economy but its mortgage market is one of the eight world-leading mortgage markets, based on the EMF (2016) statistics. The Australian ratio of total outstanding residential loans (home loans) to GDP is the fourth highest (60%), just behind the Netherlands (95%), the UK (68%), and the US (63%), but well above the four others, Spain (52%), France (44%), Germany (42%) and Japan (41%). However, based on their ratios of total mortgages (home loans and housing loans for investors) to GDP in the IMF (2017) database, Australia (105%) stands closely behind only the Netherlands (110%), but far above most of the others, the US (80%), the UK (90%), and the rest (around 50–60%). Therefore, it is of paramount importance to identify whether the crisis has any impact on the transmission of funding costs to mortgage rates in Australia, given the collapse of the US sub-mortgage market and the 2007–11 systemic banking crises in the US, Greece, Spain, Belgium, Iceland, Ireland, which are associated with a housing credit boom similar to Australia’s current position (Laeven & Valencia, 2013). Moreover, it is crucial to assess whether international funding cost attributes to the recent ‘out-of-cycle’ mortgage price-setting conduct of Australian banks after the GFC. This essay contributes further econometrically to the literature by investigating the effect of both bank-specific and time dimension on the interest-rate transmission mechanisms that have been disregarded in the prior literature. This empirical essay finds that, although the cash rate transmission in Australia has been weakened since the GFC, banks have still continued setting their mortgage rates in an upward asymmetric manner, increasing faster than falling. Importantly, this study ascertains that the post-crisis effects of foreign funding costs on mortgage rates are stronger than the pre-crisis and GFC influences. The findings indicate that the cash rate has been unable to reduce the bank markups charged over funding costs after the crisis. This study concludes that, since the GFC, Australian banks’ pricing behaviour has remained positively asymmetric in the long term, but it has significantly transformed from a cash-rate-based model to a foreign-funds-rate-based model. Therefore, this study proposes several important policy implications—for significant reforms of the RBA, and with stronger competition and customer protection schemes for the APRA and Australian Competition and Consumer Commission [ACCC]. The third empirical essay further explores heterogeneity in mortgage interest-rate pass-through at an Australian bank level by answering three research questions: (i) Do unobservables affect the interest-rate transmission, both symmetric and asymmetric, when considering both bank-specific effect and time dimensions? (ii) Is the nature of the long-term interest-rate transmission, both symmetry and asymmetry, accurately homogeneous? (iii) Does the international funding cost still affect the mortgage rate when controlling for cross-sectional dependence? This is the first study in the literature on monetary policy rate transmission to empirically investigate the effect of unobserved factors on the transmission while carefully testing the cross-sectional dependence properties of the macro panel data. As finance and banking environments are highly integrated, locally and internationally, it is important to examine the role of unobservables in bank price-setting behaviour, considering cross-sectional dependence inherent in panel data. Thus, this study provides convincing evidence to understand far more deeply and accurately the bank funding cost–mortgage rate nexus. A richer line of research has hitherto investigated heterogeneity in the interest-rate transmission mechanism outside and in Australia; however, these studies have mostly disregarded the presence of complex cross-sectional dependencies, both observed and unobserved common factors which arise from national and global economic shocks. As far as could be ascertained, none of the prior studies nationally and globally investigate the effect of the unobservables and cross-sectional dependences on these monetary policy transmission mechanisms, both symmetric and asymmetric. Intuitively, the literature assumes that interest rate properties are cross-sectionally invariant. However, this study argues that the omission of unobservables and cross-sectional dependence in the existing literature can produce seriously biased estimated results, with their subsequently misleading inference. The empirical results of this essay show that the transmission mechanisms, both symmetric and asymmetric, have been significantly affected by unobservables which are attributed to the strong presence of complex cross-sectional dependencies in panels. Contrary to intuition, and as hypothesised, this essay finds that, at the bank level, the nature of these long-term transmission mechanisms is heterogeneous. The estimated results of the international funding cost effect on mortgage rate pass-through have been significantly robust. From a practical viewpoint, these findings raise questions about the standard design of monitoring focused homogeneous transmission recommendations. Thus, this study calls for an innovative configuration in modelling interest-rate pass-through. Consistent with the second essay, this study documents the strong impairment of the long-term cash-rate transmission since the crisis, while bank mortgage price-setting behaviour has become closer to their foreign funding cost developments. The fourth essay is motivated by the findings of the cash-rate transmission breakdown in these two previous essays. The researcher conducts an empirical examination of the effects of macroeconomic conditions and credit risk premium on the mortgage rate-setting behaviour of Australian banks prior to, during and after the GFC to provide deep insight into the third literature strand of this transmission mechanism. This leading work finds significant evidence of the influential roles of the unemployment, changeability in the Australian dollar value, volatility in the bank foreign-funding markets, and business and consumer confidence in affecting the asymmetric pass-through of the mortgage rates since the crisis. The findings of this paper therefore shed light on the implementation of forthcoming monetary policy reforms. These outcomes are beneficial for depository institutions and prudential supervisors as well. It is fundamental for commercial banks to deeply understand these determinants when they design and pursue their pricing strategies because these determinants of the mortgage rates could affect the asymmetric adjustments of other bank rates differently (Perera & Wickramanayake, 2016). The findings are useful for the APRA and ACCC in maintaining and enhancing market competition, banking soundness and financial stability. The findings confirm a strong integration between financial instability and monetary transmission that would benefit prudential legislators in designing and configuring more appropriate schemes of maintaining financial stability. This thesis also extends the prior literature by employing a bank-based interest-rate dataset at high weekly-frequency in a single-country study. The use of micro-data in the study of asymmetry and heterogeneity in interest-rate pass-through is actually important for making the findings relatively advantageous to those of the macro-data analyses. This research therefore provides cooperative insights to complement monetary policy that in aggregated-data work are possible to miss or mask (ECB, 2017a). This thesis based on the findings of these interrelated studies concludes that: (i) international funding cost is a statistically driving factor of bank mortgage rates in Australia for individuals or bank-specific groups given the presence of cross-sectional dependence, (ii) the nature of the transmission from international funding cost to mortgage rates is upwardly asymmetric in the long term, (iii) wide discrepancies in mortgage price-setting occur in the short run, (iv) since the crisis, banks have partially repriced their mortgage rates only in the monetary easing episodes; however, their mortgage rates have, for the most part, fully responded to an international funding cost shift, decreasing at a greater rate than they increase, (v) these long-term transmission mechanisms, both symmetric and asymmetric, are significantly heterogeneous, (vi) when controlling for macroeconomic conditions, bank mortgage rates have responded asymmetrically and sizably to instant shocks in the cash rate, but they insignificantly respond to cash rate shocks in the long run. The mortgage rates however have considerably repriced towards the deteriorating labour market, the appreciation and depreciation in the Australian dollar value, the volatility of banks’ foreign-funding markets, as well as changes in business and consumer confidence. This research provides empirical support to the transmission of the cash rate, considering (i) the nexus of international funding cost and mortgage rates and (ii) the nexus of macroeconomic environment and mortgage rates; therefore, proposes important policy implications.