Yesterday a comment by Richard Denniss in the AFR caught my eye:
“The No.?1 issue is: why has competition spectacularly failed in the banking sector. We don’t have a competitive banking system. We have an oligopoly,” says [Richard] Denniss, noting that the major lenders account for 90 per cent of the mortgage market. The lack of competition effectively puts a tax on financial products sold to consumers and small businesses, enabling the big banks to generate “super normal” profits, he says.
The argument that the Australian banking sector is not competitive is a progressive staple. The thing is that the empirical evidence simply doesn’t support that view.
A few years ago I made a submission to the Senate on this point – and posted on the issue – but it is well worth revisiting.
One way of evaluating competition in an industry is to make use of the well-known Structure-Conduct-Performance (SCP) model. This approach works on the principle that the structure of the industry (either concentrated or diffuse) leads to conduct (either monopolistic or competitive) which in turn leads to performance (monopoly profits or normal competitive returns). The SCP approach has been criticised by economists such as Harold Demsetz, nonetheless the approach itself is widely used and appears to be intuitively attractive to non-economists.
Jacob Bikker and Jaap Bos (2008) have investigated the SCP approach in a massive analysis of 7,266 banks across 46 countries spanning 1996 – 2005. Their Australian sample includes 36 financial institutions. In their analysis they make use of the well-known Herfindahl-Hirshman Index and various concentration ratios (in the empirical test of the SCP approach they employ a top-three bank concentration ratio). They are able to find no statistically significant relationship between concentration and profitability for Australia (pg. 87) and no statistically significant relationship between the Herfindahl-Hirshman Index and profitability for Australia (pg. 88). They interpret this result as follows, ‘for most countries the impact of market structure on performance is limited’.
Of course, the argument at present is that the banking industry has become more concentrated and less competitive since the global financial crisis. To be sure it does appear that the banking sector is more concentrated and some smaller financial institutions are struggling in the current environment. It is instructive to examine what Australian regulators and public officials were saying about bank competition at the time the crisis first began and to compare that to what they say now. In particular what was reported to the 2008 House of Representatives Standing Committee on Economics Inquiry into Competition in the Banking and Non-Banking Sectors. At that time The Treasury argued
Notwithstanding the prominence of the five major banks in the Australian banking sector, there is evidence that the sector is contestable. Australian banking customers are served by a wide range of banking providers, including: 13 domestic banks; 12 building societies; 139 credit unions; and a subset of Australia’s 10 foreign bank subsidiaries, 31 foreign bank branches, and more than 100 non-ADI financial institutions. According to financial services research firm CANNEX, there are currently around: 140 providers of over 2,400 mortgage products; 70 providers of over 300 different credit cards; and 130 providers of over 2,400 different types of deposit accounts.
Similarly the RBA argued, ‘By international standards, Australian borrowers are offered a wide range of mortgage products and are able to choose from a large number of different loan types, with many of these offering more flexibility than is available in other countries.’
The view of both the RBA and the Treasury seems to indicate that the structure and conduct components of the SCP model do not suggest monopolistic behaviour.
Or for that matter oligopolistic behaviour and certainly no super normal profits.