Posted 29/11/2010 by Scott Kompo-Harms
Recently, there has been considerable concern amongst politicians, the media and the broader community regarding competition in the Australian banking sector. Specifically, concerns have been raised regarding banks increasing certain lending rates over and above rises in the RBA’s official cash rate, particularly for housing loans. This observation has led to a debate about collusion and ‘price-signalling’ within the banking sector. This post examines data from the
Australian Prudential Regulation Authority (APRA) and the
Reserve Bank of Australia (RBA) on market shares and concentration in the household lending sector to see whether there has been any increase in the
capabilities of financial institutions to use market power to their advantage.
Tables 1 and 2 below show measures of market share and market concentration for
household loans only in September 2007 and September 2010. These dates were chosen so as to encompass the ‘global financial crisis’, which began in late 2007. Non-bank financial institutions, such as credit unions and building societies, are included in the analysis as they often compete with the banks by offering similar products. However, each non-bank market segment is counted as being one 'firm' in the market.
The tables show that whilst the market share of household loans held by banks has only increased slightly (from around 91.5 per cent to 92.9 per cent), the share of the ‘big four’ major banks (ANZ, Commonwealth, NAB and Westpac) has increased substantially, from under 70 per cent to over 80 per cent. The biggest contributors to this were the takeovers of St George and BankWest by the Westpac and Commonwealth banks respectively (in addition to takeovers of some smaller, non-bank lenders). This is reflected by the increase in the
Herfindahl-Hirschman Index (HHI) between September 2007 and 2010.
Table 1: Market shares and concentration: household loans September 2007 | September 2007 |
Household loans ($ million) |
Market Share |
Contribution to HHI |
Building Societies |
15375 |
1.9221% |
0.00037 |
Credit Unions |
29252 |
3.6570% |
0.00134 |
Money Market Corporations |
79 |
0.0099% |
0.0000 |
Finance Companies and General Financiers |
22971 |
2.8717% |
0.00082 |
TOTAL BANKS |
732224 |
91.5393% |
0.12730 |
- Big 4 Banks |
546718 |
68.3482% |
- |
TOTAL OTHERS |
67677 |
8.4607% |
-
|
TOTAL |
799901 |
100% |
0.12983 |
Table 2: Market shares and concentration: household loans September 2010 | September 2010 |
Household loans ($ million) |
Market Share |
Contribution to HHI |
Building Societies |
17721 |
1.52% |
0.00023 |
Credit Unions |
34853 |
2.99% |
0.00106 |
Money Market Corporations |
20 |
0.00% |
0.00000 |
Finance Companies and General Financiers |
30208 |
2.59% |
0.00067 |
TOTAL BANKS |
1 084 222 |
92.9% |
0.16094 |
- Big 4 Banks |
942925 |
80.8% |
- |
TOTAL OTHERS |
82802 |
7.10% |
- |
TOTAL |
1167024 |
100% |
0.16289 |
An increasing HHI over time shows increasing market concentration. The HHI is calculated by summing the squared market shares of all market participants. The HHI can take any value between zero and one and will increase as an individual firm increases market share and/or the number of firms reduces. If the HHI is close to zero, then the market in question is not very concentrated, whereas if it is close to one, then the market is considered concentrated.
So what is considered ‘close to zero’ or ‘close to one’? Both the
Australian Competition and Consumer Commission (ACCC) and the
US Department of Justice (DoJ) use the HHI (they calculate it as a number between 0 and 10,000 as they use market shares expressed as percentages, whereas I have calculated the HHI below using market shares expressed as decimals). The ACCC’s 2008
Merger Guidelines (p. 37) explain the way the ACCC use the HHI:
7.14. As part of its overall assessment of a merger, the ACCC will take into account the HHI, as a preliminary indicator of the likelihood that the merger will raise competition concerns requiring more extensive analysis. The ACCC will generally be less likely to identify horizontal competition concerns when the post-merger HHI is:
- less than 2000, or
- greater than 2000 with a [change in the HHI of] less than 100.
The US DoJ defines a mildly concentrated industry sector as one that has a HHI of below 1000 (or 0.1 in decimal calculations), while anything within 1000 to 1800 (0.1 to 0.18) is considered moderately concentrated. A calculation in excess of 1800 is considered highly concentrated. Any transactions that increase the HHI in concentrated markets by 100 or more points are likely to prompt further investigation.
As the tables reveal, the change in market concentration, as measured by the HHI, is around 330 basis points (0.03306). This is a substantial change. The household lending market has been quite ‘broadly’ defined here. The risk with this approach is that due to a smaller than assumed degree of substitutability between products, the HHI will underestimate market sector concentration. Thus, the analysis here may understate the degree of concentration because of the lack of substitution possibilities for households between, for example, an owner-occupier housing loan from a bank and a credit card from a finance company. If this is the case, then the household and other lending sectors may be even more concentrated than the analysis presented above suggests, although the initial degree of concentration may also be understated.
Whilst this analysis only compares two points in time and only covers lending to the household sector (and so does not give a picture of the markets for deposits or lending to non-financial corporations, for example), it nevertheless provides some evidence that the sector has become more concentrated. It is likely that concentration has increased in other sectors of the financial industry as well. At the very least, these results superficially suggest that significant change has occurred in the financial industry and a deeper investigation is warranted.