Towards Understanding Macrofinancial Impacts of Loan‐to‐Value Ratio Policy in New Zealand: A General Equilibrium Perspective

• Author: Daniel Snethlage,Lucy Greig,Martin Fukac
• Date: 01 March 2018
• Published date: 01 March 2018
• DOI: http://doi.org/10.1111/1467-8462.12256

Policy Forum: Macroeconomic Policies After the Global

Financial Crisis

Towards Understanding Macrof‌inancial Impacts of Loan-to-Value

Ratio Policy in New Zealand: A General Equilibrium Perspective

Martin Fukac, Lucy Greig and Daniel Snethlage*

Abstract

We use a dynamic stochastic general equilib-

rium model as a framework for thinking about

the transmission mechanism of loan-to-value

macroprudential policy. We analyse the key

channels through which the caps on loan-to-

value ratios work to limit the speed of asset and

credit cycles. We further analyse the mecha-

nisms through which the caps support f‌inancial

system resilience during asset price downturns

that are of suff‌icient magnitude to cause

f‌inancial and macroeconomic instability.

1. Introduction

In October 2013, the Reserve Bank of New

Zealand (RBNZ) introduced new macropru-

dential regulations aimed at moderating f‌inan-

cial stability risks associated with rising house

prices and growing debt on the back of an

already highly indebted household sector. The

measure involved a cap on the size of new

home loans relative to the collateral value

(loan-to-value ratios / LVRs). The regulation

targeted growth in highly leveraged home

loans and helped alleviate pressures on the

RBNZ’s interest rate policy that was, until

then, used to lean against the building f‌inancial

imbalance. The speed limits established a new

f‌irst line of defence and allowed the re-focusing

of monetary policy back to its primary mandate

of achieving price stability.

The banks reacted quickly, reducing the

origination of high LVR loans to substantially

below the regulatory limit of 10 per cent of total

residential mortgage lending.

1

Figure 1 shows

this sharp reduction in the supply of high-

leverage loans. The initial impact on the supply

of high-leverage home loans, f‌low and stock of

credit, and house prices is also evident in the

data (see Bloor and McDonald 2013 and Price

2014 for counterfactual evidence of the initial

effectiveness of the policy).

Our article aims at dee pening the under-

standing of how this regulation may have

affected the wider New Zealand economy by

employing a structu ral general equilibr ium

model to shed more light o n the transmission

mechanism of f‌inancia l crises and interac-

tions between f‌inancial and macroeconomic

* Fukac: Inclusive Analytics NZ, Wellington 6012, New

Zealand; Greig: Economics Department, Reserve Bank of

New Zealand, Wellington 6140, New Zealand; Snethlage:

The Treasury, Wellington 6140, New Zealand. Corre-

sponding author: Fukac, email <mfukac@inclusive

analytics.com>. This work would not have been possible

without input and support from Jaromir Benes. Only his

characteristic modesty and his employer’s publication

regulationsprevented us listinghis name among the authors.

We also wish to thank Jennie Kerr,Tim Robinson, May Li

and seminar participants at the NZ Treasury,Reserve Bank

of New Zealand, Ministry of Business, Innovation and

Employment, the IMF Workshop on Modelling Macro-

prudential Policy in Washington, DC (February 2015) and

the MelbourneInstitute Policy Meetings(October 2015) for

their helpful and encouraging comments. The authors

gratefully acknowledge the support of the New Zealand

Treasury, where the authors were employed when the

majority of this work was f‌inished. Important disclaimer:

This article is written to positively contribute to the

discussion aboutthe workings of loan-to-value ratio policy

in New Zealand.The views in this article are solely those of

the authors. The views herein do not represent the off‌icial

views of the Reserve Bank of New Zealand, or the New

Zealand Treasury andthe New Zealand Government.

The Australian Economic Review, vol. 51, no. 1, pp. 99–131

°

C2018 The University of Melbourne, Melbourne Institute of Applied Economic and Social Research

Published by John Wiley & Sons Australia, Ltd

stability. We also attempt to co ntribute to

broader, related qu estions: Does the LVR

policy actually work as intended? Do LVR

caps help prevent a f‌inancial crisis? Does the

policy help improve ove rall macroeconomic

and f‌inancial stability?

The key challenge of this kind of analysis is

that we do not know the counterfactual—the

actual state of the world had the policy not been

imposed. In that sense, having a structural

model as an analytical tool is critical to

obtaining that kind of insight.

We designed a structural model to capture an

internally coherent economic narrative around

the working of LVR speed limits. The housing

market, demand for loans and the banking

sector all play prominent roles in the model.

The design captures our thinking about how

asset prices affect f‌inancial stability and how

f‌inancial and macroeconomic stability interact

with each other. Specif‌ically, the model can

provide insights into the fundamental reasons

for the introduction of macroprudential policy;

the role of the banking sector; the aspects of

f‌inancial imbalance that affect macroeconomic

stability; and the properties and drivers of asset

price cycles in New Zealand that are important

for f‌inancial stability.

Credit in the model is supplied by the

banking sector inf‌lating its balance sheet. The

banking sector is designed after Benes,

Kumhof and Laxton (2014) in the spirit of

how credit money is created in reality

(McLeay, Radla and Thomas 2014). How

credit money is created is the fundamental

difference between our model and the loanable

funds (saver–investor) model, with which the

reader may be familiar. In the loanable funds

Figure 1 Selected Indicators of the Mortgage Market

Note: We thank Chris Bloor for providing us with an updated version of the top left chart, which was previously published in

Bloor and McDonald (2013, p. 4, Figure 2).

Sources: RBNZ estimates, registered banks’disclosure statements, Quotable Value Ltd.

100 The Australian Economic Review March 2018

°

C2018 The University of Melbourne, Melbourne Institute of Applied Economic and Social Research

model the credit supply is constrained by the

amount of savings the economy can generate.

That also limits the possible size and severity of

f‌inancial booms and busts these frameworks

can predict.

In our model, banks can inf‌late their balance

sheets by extending lending to households, and

are only constrained by regulatory require-

ments on capital adequacy and what is

considered prof‌it maximising in the context

of potential regulatory penalties. As a result,

bank lending can outgrow the total amount of

available savings. Banks’ability to inf‌late their

balance sheets helps generate lending booms,

and creates a chance of balance-sheet driven

crises and prolonged economic recoveries,

accompanying loss in living standards after

f‌inancial crises, as documented by Reinhart and

Rogoff (2014).

We motivate an asset price boom in housing

primarily through migration inf‌lows, which

are historically associated with house price

pressures in New Zealand. The second compo-

nent of house price booms is created by

market sentiment that can drive prices above

their fundamental values (that is, in breach

of rational expectations). We are agnostic

about the proportion of fundamental and non-

fundamental drivers of the house price cycle;

we only acknowledge these two factors for

house price growth.

Using a New Keynesian general equilib-

rium framework, we are mindful of its

limitations.

2

Primary amongst these are the

reliance on the rational expectations hypothe-

sis, the use of representative agents, and

various properties that constrain the existence

of f‌inancial crises or critically downplay their

f‌inancial and macroeconomic consequences.

These limitations need to be borne in mind

when deriving policy lessons from this

analysis. Macroprudential policy and its

architecture is a relatively new area. There

is, uncomfortably, a lot we do not yet

understand. This article is an invitation to

other researchers and policy analysts to work

in this f‌ield.

The rest of the article is structured as

follows. In Section 2, we draw together

selected stylised facts on the New Zealand

house price cycle. We then use these to

motivate an experiment in which we study

the macroeconomic and f‌inancial stability

impacts of a house price boom and a sudden

bust. Next, in Section 3, we outline the

structural model used to simulate the experi-

ment. The results of the experiment and policy

implications are discussed in Section 4. Section

5 concludes the article.

2. Initial Thoughts on Modelling Strategy

This section summarises initial thoughts and

empirical facts that motivated our modelling

strategy. These priors are embedded in the

model structure and scenario designs in

Sections 3 and 4, respectively.

2.1 The Macrof‌inancial Signif‌icance of

Housing

Housing is the single most important asset for

many households in New Zealand, and the

largest single purchase most households make.

Housing loans represent around 12 per cent

of household net wealth, and housing and

land assets represent around 55 per cent.

3

In New Zealand, as in many places, home

ownership is a f‌inancial decision particularly in

the presence of sustained rises in house prices

over time. However, it is also a cultural/social

aspiration, with many families aspiring to own

their own home. In addition to this, the New

Zealand rental market often involves relatively

short notice periods and f‌luid tenancies, and

therefore the benef‌its of stability from home

ownership are a strong driver. As such, owner-

occupied housing is both a consumption good

and an investment good.

The aggregate debt of New Zealand house-

holds increased signif‌icantly in the decade

leading up to 2008, reaching around 100 per

cent of household annual income. The increase

was large and historically unprecedented, but

not exceptional compared to other countries’

experience over the same period. Credit growth

was very weak from 2009 to 2012, however,

more recently, as the economy has strength-

ened, household debt relative to income has

begun to moderately pick up again.

°

C2018 The University of Melbourne, Melbourne Institute of Applied Economic and Social Research

Fukac, Greig and Snethlage: Towards Understanding Macro-Financial Impacts of LVR Policy 101