Essays on Credit, Asset Prices and Macro-Prudential Policy

Chapter one of this thesis examines the behaviour of credit volumes, asset prices and output

for US data using a dummy-augmented vector autoregression model. The paper contributes

to the literature in three ways. Firstly, statistical analysis indicates that from 1985 onward

house prices exhibit phase-shift in leading output. Over the same period, data shows that

household credit becomes noticeably less volatile relative to output, whereas ?uctuations in

business debt become more pronounced. Secondly, Granger-Causality tests show that there

is signi?cant feedback between house prices and household credit. This result was robust to

periods of high and low household debt service costs. Interestingly, changes in interest rates

are shown to Granger-Cause changes in house prices and credit volumes only during periods

when debt service costs are above the long-run average rate. And third, the magnitude of

the responses of credit and asset price variables to variety of shocks are sensitive to whether

debt service costs are higher or lower than average. Speci?cally, when ?nancial obligations

are high a shock to the residual in the interest rate equation produces an ampli?ed response

in credit and asset price variables.

Chapter two constructs a New-Keynesian DSGE model with multiple credit constrained

agents in order to examine whether expectational shocks to the loan-to-value (LTV) ratio

can create cyclical behaviour in output and house prices. The contributions of this paper

are threefold. Firstly, when agents anticipate a future loosening of the LTV ratio and this

expectation turns out to have been incorrect, house prices, consumption, investment and

output all su?er a sharp decline. Secondly, the paper shows that in order to generate

downward comovement of output and house prices in response to a frustrated expectational

shock to the LTV ratio, agents must be able to post a substantial quantity of the total value

of their housing asset as collateral. And thirdly, the ability to post capital as well housing

as collateral signi?cantly ampli?es the cumulative loss of both output and house prices in

the face of a frustrated expectational shock to the LTV ratio.

Chapter three examines the e?cacy of two types of macro-prudential policy in an es

timated DSGE model with a banking sector. Both in the case of counter-cyclical capital

requirements (CCR) and a lean-against-the-wind (LATW) type Taylor Rule, policy is an

chored to house price growth. Using a conventional central bank loss function as the metric

to gauge the e?cacy of macro-prudential policy, the model indicates a reduction in the vari

ation of output of up to 7.38% in the case of a technology shock, and 22.14% in the case

of a monetary policy shock when the CCR is the instrument of choice. However, policy in

this form exacerbates ?uctuations in in?ation in the case of technology shock in the region

of 4.99%. If the source of the shock is through housing preference, loan-to-value or bank

capital, policy in this form unambiguously increases the variance of both output and in?a

tion. If policy takes the form of LATW, the output-in?ation trade-o? is more pronounced

in the case of a technology shock. The improvements in reducing the variance of output and

in?ation are lessened when LATW is active rather than CCR for a monetary policy shock.

The model indicates that any improvement in stabilising output and in?ation is signi?cantly

o?set by a pronounced increase in the variance of credit growth when policy is in the form

of CCR