Tuesday, September 12, 2017

Housing and financial market instability

Is real estate trends the biggest force for macroeconomic distortion? I have blogged earlier about the work of Mathew Rognlie which highlights the central contribution of housing prices to widening inequality. 

Now, following on from the work of Atif Man and Amir Sufi, √íscar Jord√°, Moritz Schularick, and Alan Taylor construct historic housing price data for the 1870-2012 period for 17 countries and find,
First, we discuss long-run trends in mortgage lending, home ownership, and house prices and show that the 20th century has indeed been an era of increasing “bets on the house.” The strong rise in aggregate private debt over GDP that can be observed in many Western economies in the second half of the 20th century has been mainly driven by a sharp increase in mortgage debt. Mortgage credit has risen dramatically as a share of banks’ balance sheets from about one third at the beginning of the 20th century to about two thirds today. As a result, the intermediation of savings into the mortgage market has become the primary business of banking, eclipsing the stylized textbook view of banks financing the capital formation of businesses.


Second, turning to the cyclical fluctuations of lending and house prices we... show that throughout history loose monetary conditions were closely associated with an upsurge in real estate lending and house prices... Broadly speaking, when countries peg to some base currency they effectively import the base economy’s monetary policy, even if it is at odds with home economic conditions. Exchange rate pegs therefore provide a source of exogenous variation in monetary conditions. By conditioning on a rich set of domestic macroeconomic controls, we are able to isolate exogenous fluctuations in the short-term interest rate imported via the peg and trace the effect of these fluctuations over time on other variables. Third, we also expose a close link between mortgage credit and house price booms on the one hand, and financial crises on the other. Over the past 140 years of modern macroeconomic history, mortgage booms and house price bubbles have been closely associated with a higher likelihood of a financial crisis. This association is more noticeable in the post-WW2 era, which was marked by the democratization of leverage through housing finance.
Their data construction allows them to identify causal pathways and converge on two important findings, 
Loose monetary conditions are causal for mortgage and house price booms, and this effect has become much more dramatic since WW2... Mortgage and house price booms are predictive of future financial crises, and this effect has also become much more dramatic since WW2.
The case study of how in the aftermath of EMU when countries lost their monetary policy autonomy, monetary accommodation increased mortgage lending, and engendered housing bubble in Spain and Ireland is fascinating. 
As can be seen, for the 1999-2007 period, coinciding with the first years of the monetary union, the optimal monetary policy rates for Ireland and Spain should have been much higher than the ECB policy rates, given the rapid growth rates experienced by them during that period. The loose policy boosted mortgage lending which doubled in eight years, and in turn inflated housing prices by 65-75%. This contrasts with Germany, with its moderate growth and resultant tighter monetary conditions, stable mortgage lending, and declining property prices. 

Its implications are clear. One, policy actions, like macroprudential tools, have relevance in the regulation of the market failures arising from the inevitable excess of mortgage lending. Two, central banks should be cognisant of the fact that the side effects of monetary accommodation can destabilise financial markets, with housing being a major channel of instability. Three, it may be time to re-examine the dominant role of dollar and the attendant effect of its pegging other currencies, which creates the channel for importing financial instability. In simple terms, the dominance of dollar effectively invalidates the Mundell-Fleming trilemma and makes monetary policy always significantly dependent on the US Fed's policy actions. 

No comments: