Goldman: Macroprudential worth 50bps

Goldman Sachs is out with a study of the effects of macorprudential policy on house prices, via FTAlphaville:

Relative to their previous peaks, nominal house prices are now 43%higher in Norway, 20% higher in Canada, 14% higher in Israel and 7%higher in Australia.

…In Canada, nominal house prices did not fall much during the GFC and have risen steadily since 2009. Homebuilding activities have also exceeded their pre-GFC levels. In response, the Ministry of Finance lowered the maximum LTV ratio from 100% to 95%, reduced the maximum amortisation period from 40 years to 35 years, and set the maximum DSTI ratio to 45% in October 2008. In April 2010, it lowered the maximum LTV ratio further to 90% for refinancing mortgages and to 80% for investor properties. In March 2011, the authorities reduced the maximum amortisation period from 35 years to 30 years and decreased the maximum LTV ratio to 85% for refinancing mortgages.

We find that macro-prudential housing policy, when designed to tighten (ease) mortgage availability, reduces (increases) the annual growth rate of real household credit by 1% and the annual growth rate of real house price by 1% (Exhibit 10). These estimates are statistically significant and are of economically significant magnitudes, given that the average annual growth rate is 5% for real household credit and 2% for real house prices in our sample. Estimating VAR and DSGE models using US data, Jarocinski and Smets (2008) and Iacoviello and Neri (2010) find that a 50bp increase in the federal funds rate leads real house prices to decline 0.75%-1.5%. Putting our estimates in this context, the average macro-prudential housing policy in our sample is equivalent to a 50bp change in the policy interest rate.

If Australia were to implement this and provide cover for an additional two rate cuts, the dollar would probably fall to 80 cents.