House Prices and the Credit Impulse

• Tweet

• Tweet

Today’s fig­ures for the fall in house prices over the last quar­ter and year were larg­er than the usu­al bull-side pun­dits expect­ed: a 1.7% fall for the quar­ter (ver­sus the 0.5 per­cent medi­an pre­dic­tion of 17 econ­o­mists sur­veyed by Bloomberg) and 0.2% fall for the year (ver­sus expec­ta­tions of a 1.6% gain for the year by the same group of econ­o­mists, accord­ing to Chris Zap­pone’s arti­cle).

Fig­ure 1

I won­der if the same econ­o­mists will now assert that declin­ing immi­gra­tion and/or pop­u­la­tion is behind the fall? After all, pop­u­la­tion growth and a sup­ply short­age rel­a­tive to pop­u­la­tion growth were the rea­sons they gave for a rise in prices in the first place—surely the same argu­ment must work in reverse?

In fact, as I’ve argued ad nau­se­am on this blog and else­where, the real expla­na­tion for ris­ing house prices was ris­ing cred­it. To be more pre­cise, what dri­ves the change in house prices is the accel­er­a­tion of mort­gage debt.

This is an empir­i­cal exten­sion of the argu­ment I’ve made about the role of cred­it in the macroeconomy—for back­ground, see these three posts on Delever­ag­ing, Aus­tralian Debt, and Aus­tralian banks and house prices). In a cred­it-dri­ven econ­o­my, aggre­gate demand is the sum of income plus the change in debt—and there­fore the change in aggre­gate demand is the sum of the change in income plus the accel­er­a­tion of debt, a phe­nom­e­non dubbed “The Cred­it Impulse” (Michael Big­gs et al., 2010).

In the past I’ve cor­re­lat­ed this with changes in employ­ment, and shown that the rapid change from accel­er­at­ing to decel­er­at­ing debt was the cause of the Great Reces­sion (or the Glob­al Finan­cial Cri­sis, as it’s called in Aus­tralia). But the Cred­it Impulse affects asset prices too—since we use cred­it not mere­ly to pur­chase new­ly pro­duced goods, but also to buy exist­ing assets. Today, I’ll take a look at the cor­re­la­tion of the Cred­it Impulse with change in Aus­tralian house prices, and com­pare this to the prop­er­ty spruik­er’s argu­ment that pop­u­la­tion growth and sup­ply con­straints jus­ti­fy Aus­trali­a’s astro­nom­i­cal house prices.

This cor­re­la­tion is the “smok­ing gun” in the Aus­tralian prop­er­ty debate.

Fig­ure 2

Not only does the accel­er­a­tion in debt cor­re­late with changes in house prices—the cor­re­la­tion of the accel­er­a­tion in all pri­vate cred­it with change in house prices is 0.28, and the cor­re­la­tion of accel­er­a­tion in mort­gage debt with change in house prices is 0.58—the accel­er­a­tion in debt also leads changes in house prices by about 3 to 6 months.

Fig­ure 3

How about the cor­re­la­tion of changes in population—and the ratio of pop­u­la­tion to dwellings—with changes in house prices?

Fig­ure 4

Not only are these demo­graph­ic fac­tors far less volatile than house prices—and than media and pop­u­lar obses­sion with them would imply—their cor­re­la­tion with changes in house prices is actu­al­ly neg­a­tive. The cor­re­la­tion of change in house prices with change in pop­u­la­tion since 2000 is ‑0.34, and the cor­re­la­tion with change in pop­u­la­tion per dwelling is even worse, at ‑0.41.

The pic­ture gets worse when you con­sid­er leads and lags: the cor­re­la­tion of pop­u­la­tion and pop­u­la­tion den­si­ty 6 months ahead of price changes is low­er than the con­tem­po­ra­ne­ous cor­re­la­tion, and in either direction—leading or lagging—the cor­re­la­tion is neg­a­tive.

Pop­u­la­tion dynamics—even immi­gra­tion dynamics—have noth­ing to do with house prices. What deter­mines house prices is not the num­ber of babies being born, or immigrants—illegal or otherwise—arriving, but the num­ber of peo­ple who have tak­en out a mort­gage, and the dol­lar val­ue of those mort­gages.

For changes in house prices, what mat­ters is the accel­er­a­tion of mort­gage debt, and that’s why the First Home Ven­dors Boost was instru­men­tal to the turn­around in house prices in 2009: it turned a nascent decel­er­a­tion in mort­gage debt into an accel­er­a­tion once more. That accel­er­a­tion has now run out and decel­er­a­tion has resumed–and house prices have start­ed to tum­ble as a result.

Fig­ure 5

The fact that the Cred­it Impulse leads changes in house prices also gives some indi­ca­tion of where future prices are like­ly to go. The mort­gage Cred­it Impulse shown above is for the accel­er­a­tion in mort­gage debt over a year: the change in the change in mort­gage debt com­pared to the pre­vi­ous year. This brings in an inevitable lag in the series—matched by the lag in the change in house price data, which also shows the change in house prices over the pre­vi­ous year—so that the turn­ing points in each series line up in the graph: lows in the mort­gage cred­it impulse are asso­ci­at­ed with lows in house price change, and vice ver­sa. With the mort­gage cred­it impulse still head­ed south, and lead­ing falls in house prices by 3–6 months, that implies that there are at least two more quar­ters of neg­a­tive house price move­ments com­ing up.

Fig­ure 6

Of course, there could always be a change in gov­ern­ment pol­i­cy that entices peo­ple back into debt—as the First Home Ven­dors Boost did in 2008. How­ev­er, gov­ern­ments might huff and puff to try to keep the house price bub­ble inflated—as the Vic­to­ri­an Gov­ern­ment is doing in its cur­rent bud­get, with its sup­posed boost to First Home Buy­ers that in real­i­ty is a sup­port scheme for Vic­to­ri­an house prices—but the like­li­hood of the lit­tle pig­gies rush­ing back into the straw house of debt is min­i­mal, when Aus­tralian mort­gage debt is already at lev­els that dwarf those in the USA. The Aus­tralian house price bub­ble is over.

Fig­ure 7

Big­gs, Michael; Thomas May­er and Andreas Pick. 2010. “Cred­it and Eco­nom­ic Recov­ery: Demys­ti­fy­ing Phoenix Mir­a­cles.” SSRN eLi­brary.