SYDNEY: 23 January 2014 - The notion that some institutions were ‘Too Big to (be allowed to) Fail’ was very influential during the global financial crisis.  It contributed to an ideational milieu that justified the injection of substantial amounts of public money into banks and other financial institutions around the world.  These institutions were so ‘important’ that it was deemed preferable to prop them up, notwithstanding the deficiencies in their business models exposed by the crisis, rather than attempt to weather the economic dislocations forecast to occur if market forces were allowed to take their course.

What were the dimensions of that importance?  As the dust settled, the Basel Committee on Banking Supervision identified five: size, cross-jurisdictional activity, interconnectedness, substitutability and complexity.  Notably brute size was just one of the dimensions.

APRA recently applied this methodology to the Australian financial system.  (It dropped the cross-jurisdictional dimension because of its domestic focus).  Not surprisingly, it identified the four major trading banks; Westpac, CBA, ANZ and NAB, as being systemically significant at a domestic level.  Few surprises there.

What was a surprise, and not a pleasant one, was the myopic approach taken to the task itself.  The banking system is undoubtedly an important one.  It facilitates transactions across the economy, stores a substantial stock of private savings, employs a great many people and constitutes almost a quarter of the market capitalisation of the Australian sharemarket.  But banks today, especially in Australia, are not simply deposit-taking institutions.  They own businesses that span the breadth and depth of the financial system: insurance companies, stock brokers, superannuation fund trustees, investment managers and financial advisers amongst them.  Some of these are self-evidently critical to real world ‘systems’ with which the financial system interacts integrally.  But there are other roles that they play that are equally critical; nodes that lie beneath the surface and attract much less attention.

Take the superannuation system as a good example.  There are around 200 major superannuation funds and dozens of specialist investment managers.  Most outsource a large number of core activities and rely heavily on advisers.  They are, if you will, a perfect example of the ‘virtual’ institution so eagerly anticipated by management theorists in the 1990s.  This might not be a problem except for the fact that there is considerable market concentration in a number of key outsourced activities.  So the diversity the superannuation system would appear to have is considerably undermined by the interconnectivity arising from common reliance on underlying service providers, including custodians, administrators and insurance providers. A number of these service providers are owned by Australia’s banks and are thus themselves interconnected.

Reference back to the Basel Committee (and FSB) approach.  It is true in this context also that systemic importance is not purely a function of size.  Interconnectivity also matters; a single small node could prove to be the weak link for the system as a whole if it plays a critical role in one or more processes on which other, larger institutions rely.  Substitutibility also matters, and not just availability of alternatives but the practical feasibility of substitution.

There is however one difference that is important.  The nature of the linkages is different here.  Failure can propagate through this system without an insolvency or asset deflation cascade.  Informational errors, ideational misconceptions, cashflow freezes can all propagate in a system-threatening manner.

Which brings us back to the banks.  Australia’s banks (and analysis suggests that Macquarie and IOOF, at least, should be added to the list in respect of the types of risks identified here) play many roles across the superannuation system and the broader financial systems with which it interacts.  To limit ‘systemic’ importance in the way that APRA has done in its recent analysis misses this completely.  When Barings failed in Singapore in 1995, clients of Barings’ asset management, broking and custody businesses found their assets frozen until auditors and investigators could confidently distinguish assets held on trust from those belonging to Barings in its own right.  More recently, the process of unravelling the tangle of claims on the assets in the Lehmans bankruptcy is now well into its fifth year.  The fact that trust and nominee assets are quarantined at law is of some solace, unless you need to value a portfolio, accept contributions, meet benefit payments, manage a stable of active investment managers or report to a regulator (or your clients).  Then the failure could assume an altogether more ‘real’ quality.

This is but one example.  Australia’s banks interact with the superannuation system in a complex but integral manner.  They are critical to its effective and efficient operation.  Entities within the one banking group could at any time be trustee, asset manager, custodian, adviser, insurer, member benefit administrator, broker, OTC derivatives counterparty, JV partner, tenant and portfolio position of a large superannuation fund.  No doubt they have a similar web of interactions with other sectors of the financial markets and the economy, relationships too complex and nuanced to model as simply debtor:creditor relationships and too important to ignore.  If nothing else, this should highlight why it is essential that the Financial System Inquiry takes a more holistic view of the systemic risks posed by the banking sector than does APRA, the regulator specifically mandated to supervise and prudentially regulate the sector.