Why Do Property‐Liability Insurers Destroy Liquidity? Evidence from South Africa

• DOI: http://doi.org/10.1111/saje.12212
• Date: 01 March 2019
• Author: Abdul Latif Alhassan,Nicholas Biekpe
• Published date: 01 March 2019

© 2019 Economic Society of Sout h Africa. doi : 10.1111/ saje .1221 2

22

South African Journal of Economics Vol. 87:1 March 2019

WHY DO PROPERTY-LIABILITY INSURERS DESTROY

LIQUIDITY? EVIDENCE FROM SOUTH AFRICA

ABDUL LAT IF ALHASS AN*,† AND NICHOL AS BIEKPE†

Abstract

This paper examines liquidity creation behaviour in the property-liability insurance market in

South Africa. Using annual data on 76 insurers from 2007 to 2014, the paper employs the three-

stage approach to estimate liquidity creation. The results suggest that property-liability insurers are

characterised by liquidity destruction by transforming liquid assets in cash and investable securities

into illiquid reserves liabilities. The findings also indicate that the R1.32 billion in liquid assets

were transformed into illiquid reserves liabilities in 2014, an increase from the R700 million

liquidity de-created in 2007. The increases were mainly driven by large insurers which accounted

for about 70% liquidity de-created. The results of panel regression analysis provide evidence in

support of the “risk-absorption” hypothesis which argues that high levels of capital increases

liquidity creation. In addition, size, leverage and reinsurance were also identified as the firm-level

factors that explain liquidity creation. The policy implications of the findings are discussed.

JEL Classification: D22, G22, N27

Keywords: Liquidit y creation, insurance, capital, South Af rica

1. INTRODUCTION

Property-liability (P&L) insurers underwrite policies with shorter durations and unpre-

dictable loss development; therefore, they are constrained by investments in long-term

projects and illiquid assets1 as they seek to promptly settle cla im demands by policyhold-

ers (Lambert and Hoff lander, 1966 in Choi et al., 2016). Unlike banks and other depos-

itory financial institutions,2 insurers have relatively more liquid assets3 in c ash and bank

deposits and illiquid liabilities in technical reserves (Colombini and Ceccarelli, 2004;

1 Due to expected and une xpected liquidity demand for policyholders.

2 This difference reflects the inver ted production cycle (Dhaene et al., 2017) where insurers re-

ceive premiums upfront before investing in (liquid) assets to retire the short-term and uncertain

claims b y policyholders.

3 Banks il liquid assets a re mostly in loans wit h liquid liabilities are in the form of customer de-

posits. In South Afr ica, cash and bank deposits ac counts for over one-third of short-term insurers

total invested ass ets. Together with other trada ble financial securities l ike stocks and shares, they

account for over 60% of invested asse ts (Alhassan, 2016:24).

*Corresponding author: Sen ior Lecturer, Development Finance and I nsurance, Development

Finance Centre (DEFIC), Graduate School of Busine ss, University of Cape Town, Cape Town,

South Africa . E-mail: latif.alhassa n@gsb.uct.ac.za or lateef85@yahoo.com

† Development Finance Centre (DEFIC), Graduate School of Business , University of Cape Town

South African Journal

of Economics

23South African Journal of Economics Vol. 87:1 March 2019

© 2019 Economic Society of Sout h Africa.

Choi et al ., 2016). This reflects the differences in the production cycle of insurers and

banks. However, for the same purposes of avoiding bank runs, the management of li-

quidity by insurers is also import ant to avoid the incidence of a “claim-run”4 in the event

of catastrophic losses and potential bankruptcy.

Theoretically, the amount of capital firms holds has been posited as one of the most

significant explanatory factors for liquidity creation behaviour among financial institu-

tions. In line with this, two competing hypotheses have emerged to explain the effect of

equity on liquidity creation. First, the “financial fragility crowding-out” theory (Gorton

and Winton, 1995 and Van den Huevel, 2008 cited in Berger and Bouwman, 2009)

suggests that increa sing equity capital reduces liquidity creation of financial i nstitutions.

This results from the reduction in the monitoring functions of firms that are associated

with high levels of capital. This affects the ability of f irms to extend credit or create

financial securities. The second hypothesis referred to as the “risk-absorption” theory

(Berger and Bouwman, 2009) posits that f irm’s high capital levels have the capacity to

absorb risk, hence increasing (decreasing) liquidity creation (de-creation). While these

theories have been employed to examine liquidity behaviour in several industries, few

evidences exist for insura nce markets, especially from the emerging market perspect ives.

This paper exami nes the hypothesis that, un like banks and depository financia l insti-

tutions, insurers are liquidity destroyers in an emerging insurance market. Since there is

a difference in the levels of equity capital maintained by insurers compared with banks

and depository financial institutions,5 this paper also explores the role of capital struc-

ture (debt and equity) in explaining the liquidity creation behaviour of property-liability

insurers. The analysi s expands on the developing literature on liquidity creation in finan-

cial markets by focusing on the property-liability in an emerging insurance market.

Against this background, this paper analyses firm level data (between 2007 and 2014)

from 79 South African property-liabilit y insurers using the three-stage liquidity creation

methodology of Berger and Bouwman (2009).

This paper makes t wo main contributions to the insurance literature. First, this study

provides the first comprehensive assessment of liquidity creation behaviour in an emerg-

ing (Africa’s largest) insurance market. This is achieved using the Berger and Bouwman

(2009) liquidity creation model adapted to insurance markets by Choi et al . (2013, 2016).

The second contribution relates to the examination of liquidity determinants, which are

important in highlighting areas of concern for regulatory actions. Unlike the US where

the insurance liquidity creat ion behaviour has been ana lysed by Choi et al. (2 013, 2016 ),

the regulatory framework of the South African insurance market in the Solvency and

4 This refers to the situat ion where insurers are una ble to pay out claims made by polic yholders

arising out of unforeseen large number of loses arising of natural disasters. T his situation could

be worsened where highly illiquid investments must be liquidated below face value and greatly

hampers insurers abi lity meet their claim obligations.

5 Banks are by reg ulatory requirement expected to maintain equity level of 10% of their tota l

assets. In insurance markets, about 45% of their asset s ate made up of equity from shareholders.

For instance, Choi et al. (2013) reports an average sur plus (equity) ratio of 45.17% in the US

property-liability from 1999 to 2008 whi le Alhassan and Biekpe (2015) also report an average of

42.26% in the non-life insur ance market in SA from 2007 to 2012. Chang and Tsai (2014) also

find an averag e of 44.42% among US property-liabil ity from 2006 to 2010.