Cross-border bank flows, funding liquidity and house prices 1 Kate Phylaktis 2 1 2 Cass Business School
Overview Developments during and after the financial crisis have highlighted the need to understand liquidity and its dynamics: What triggers systemic liquidity dry-ups? Is liquidity fragile? Where is the liquidity going? Is it fueling asset price bubbles? There has been a renewed attention on global liquidity That is, liquidity that crosses the border and affects, directly or indirectly, financing conditions abroad (Domanski, BIS 2011)
Bank flows & house prices
Global liquidity The ease of financing in global financial markets materializes as cross-border credit Identified as vector of factors that affect the supply of cross-border credit (Cerutti et al, 2014) Cetorelli and Goldberg (2011), Bruno and Shin (2015), Cerutti et al (2014) We focus on the private component of liquidity from the major wholesale funding markets for financial intermediaries, the repo market in major financial centers
Global liquidity and house prices Do changes in the financing availability of financial institutions in main financial centers affect housing markets around the world? Prior evidence of the impact of US-based global liquidity on prices: Chudik and Fratzscher (2011), Darius and Radde (2010), Tillmann (2013), Cesa-Bianchi et al (2015)
Global liquidity and house prices Do changes in the financing availability of financial institutions in main financial centers affect housing markets around the world? Prior evidence of the impact of US-based global liquidity on prices: Chudik and Fratzscher (2011), Darius and Radde (2010), Tillmann (2013), Cesa-Bianchi et al (2015) Funding markets may transmit changes in financing conditions abroad via bank flows (Bruno and Shin, 2014) and portfolio allocation of international investors (Baklanova et al, 2015) Do bank flows channel this effect? Or does the effect run via financial markets?
Global liquidity and house prices Do changes in the financing availability of financial institutions in main financial centers affect housing markets around the world? Prior evidence of the impact of US-based global liquidity on prices: Chudik and Fratzscher (2011), Darius and Radde (2010), Tillmann (2013), Cesa-Bianchi et al (2015) Funding markets may transmit changes in financing conditions abroad via bank flows (Bruno and Shin, 2014) and portfolio allocation of international investors (Baklanova et al, 2015) Do bank flows channel this effect? Or does the effect run via financial markets? Certain characteristics reduce countries exposure to shifts in external factors (Cerutti et al, 2014) Are countries able to mitigate this impact on local markets of shifts in financing conditions abroad?
Summary of main findings Funding liquidity is a key component of global liquidity Changes in funding liquidity affect bank flows towards advanced and emerging economies Global effect for advanced economies, rather regional for emerging economies (except Latin America) House prices respond to shocks to funding liquidity not only from the US, but also from Europe and UK Bank flows channel the effect in emerging economies, whereas financial markets trasmit the effect in advanced economies Countries domestic policies and institutions can alter their exposure to liquidity shocks from abroad Domestic monetary policy can offset the liquidity effect in advanced economies, whereas it only attenuates it in emerging economies More stringent bank regulation, flexible exchange rates and controls on real estate investments reduce the effect significantly
Funding liquidity proxy Funding liquidity is measured by the amount outstanding of repurchase agreements of the financial sectors in US, UK, EU, and JP US data reported by primary dealers collected weekly by the FRBNY UK monthly amounts of MFIs repos with private and public sectors by the BoE Euro area monthly amounts of credit institutions repos by the ECB Japanese monthly payable under repos in the A/L of domestically licensed banks by BoJ All data is converted in US$ via IFS monthly exchange rates Why repos? Repos are a main source of wholesale funding for financial institutions Given the presence of collateral, repos are a relatively cheap and stable source of funding In crisis, severe uncertainty in the future value of collateral led to a near collapse of the repo market in the US (Gorton and Metrick 2012) Aggregated amount instead of costs better captures the actual availability of financing
Funding liquidity & costs
Cross-border flows to banking sector Bank foreign claims in all currencies from all reporting countries vs banks in local economies from BIS locational banking statistics Countries included are: AEs: Australia, Canada, Denmark, New Zealand, Norway, Sweden, and Switzerland Asia: China, Hong Kong, Indonesia, Malaysia, Philippines, Singapore, and Thailand Europe (&co): Czech Republic, Hungary, Poland, Russia, Israel, and South Africa Americas: Argentina, Brazil, Chile, Colombia, Mexico, and Peru In a preliminary exercise, we use monthly changes in banks foreign liabilities in each country from IFS (Shin et al, IMF 2014)
Are funding conditions relevant for cross-border flows? Bank i,t = β Fundt s + δvix t + θ Mt s + γ i + ɛ t Bank i,t = β 1 ( Fundt s d+ t ) + β 2 ( Fundt s d t ) + δvix t + +θ Mt s + γ i + ɛ t s = [US, UK, EU, JP]
Advanced subsample Emerging subsample
Emerging subsample by regions
Impact of global liquidity on house prices Having documented the impact of liquidity on bank flows, we now investigate whether funding shocks affect the local housing market Data on house prices available at quarterly frequency up to 2012 from Cesa Bianchi et al (2015) PVAR of liquidity, bank flows, GDP growth, short-term rates, and house prices We estimate each country VAR via OLS and then employ the mean group estimator for the aggregated responses of house prices to liquidity shocks across countries (Pesaran and Smith, 1995) To identify if bank flows are transmission mechanisms, we estimate the PVAR with muted bank flows in the price equation and compare the responses with the baseline model If banks are channel, the impact of global liquidity on house prices is reduced when bank flows cannot directly affect house prices
Liquidity shocks from the US
Liquidity shocks from the US And with muted bank flows...
And other liquidity shocks?
And other liquidity shocks? With muted bank flows...
Financial market channel Similarly to bank channel, we test if financial markets are transmission mechanisms for global liquidity on house prices We estimate the PVAR with real estate companies stock price index GPR index available for the majority of the sample of countries We then cancel the effect of the financial market on house prices If the financial market is channel, the impact of liquidity on house prices is reduced when the real estate index is muted in the VAR
Role of monetary policy Given that local house prices respond to global liquidity, we now investigate if monetary authorities can use monetary policy to offset this effect Preliminary: irrespective of the origin of liquidity shocks, general negative reaction of house prices to shocks in domestic monetary policy We perform forecast error variance decomposition to assess the relative role of global liquidity versus domestic monetary policy on house prices
Country characteristics Following the insights of Cerutti et al (2014), we consider the role of country characteristics on the exposure of house prices to liquidity shocks We separate countries in two groups and estimate the aggregate responses to liquidity shocks for countries with higher or lower than median key country characteristics Responses of countries with characteristics above median: Responses of countries with characteristics below median:
Conclusion We find that global liquidity drives house prices around the world, controlling for domestic factors This impact is strong irrespective of the origins of liquidity for developed countries, whereas it is more regional for emerging markets We document that bank channel is important for the transmission of global liquidity shocks on house prices in emerging markets Financial markets transmit liquidity shocks from the US on house prices in advanced countries We find that domestic monetary policy cannot fully offset global liquidity effects But, certain country characteristics, such as bank regulation and real estate investment restrictions for non-residents, mitigate the impact of global liquidity on house prices