The public at large takes a very dim view of banking. When the pre-fix ‘shadow’ is added, matters only get worse. Although ‘shadow’ refers to credit intermediation occurring outside the banking sector, many ‘non-banks’ are integrated within large banking groups. It is only recently that shadow banking has attracted regulatory attention: the Financial Stability Board (established by the G20 in 2009) is concerned by the association of non-banks with ‘systematic risk’ and ‘regulatory arbitrage’. As a prerequisite to weighing those concerns, the structural similarities and differences between banks and non-banks must be understood.

A beginner’s guide might start with the standard representation of fractional reserve banking, where banks hold balances of base money (currency or reserve deposits at the central bank) as a fraction of their illiquid loans. Those balances allow banks to meet demands from their depositors for withdrawals. An associated narrative leads to the ‘money multiplier’ description of money creation. On the basis of reserves obtained from the central bank, banks make loans which then appear as new deposits. In turn, these facilitate additional bank lending so that broad money increases, round-by-round, to an amount that is in the same proportion as the increase in reserves. Broad money is a multiple of base money. The money multiplier is a definitional structure, to which behavioural relationships must be added if any insights are to be gained into the institutional niceties of banking. It is but a start: a ‘handle on the facts’.

As is true for most bank lending, non-bank lending is secured by collateral: a pawn broker lends against the security of (say) gold watches; a building society lends against the security of property. If the borrower defaults, the security can be sold. Now if it were possible to re-pawn gold watches (i.e., if ‘rehypothecation’ were possible) pawn-brokers could obtain additional funds to extend their business. Unfortunately, their clients expect to have their original watches returned.

More generally in the world of shadow banking, re-hypothecation is possible. Suppose Non-Bank I obtains a loan from Non-Bank II against high-quality collateral (say, a sovereign bond): Non-Bank II might repeat that process by re-hypothecating the security to Non-Bank III; and so on. This creates a collateral chain in which multiple non-banks hypothecate the same security as collateral for their respective loans. At each link in the chain a ‘haircut’ occurs, i.e. collateral of given worth secures a loan of increasingly diminished worth.

This invites a representation of shadow banking that is analogous to fractional reserve banking. Non-banks are distinct from banks in having no access to central bank reserves. They use collateral instead. As bank loans are a multiple of reserves, so non-bank loans are a multiple of high-quality collateral. At each round of bank lending, the ratio of broad money to reserves increases, but at a diminishing rate. At each round of non-bank lending, the ratio of loans to collateral increases, but at a diminishing rate. As the money multiplier is the ratio of broad money to base money, so ‘collateral velocity’ is the ratio of loans to hypothecated collateral. And finally, de-leveraging for banks occurs when loans are repaid or reserves accumulate; and for non-banks, when haircuts rise, or collateral falls in value, or the collateral chain shortens. Bear in mind, however, the caveat that these conceptual representations of financial intermediation provide no more than a structure for analysis and argument.

Shadow banking comprises diverse financial intermediaries whose existence can often be traced to state regulation. For example, US money market funds were a means to avoid the prohibition on bank demand deposit interest. Generic to all regulation (as to central planning generally) is that it removes the market’s way of moderating risk, and that it undermines the expectation of own responsibility. Yet, politicians hold fast to the belief that regulators can have the know-how and capability to constrain financial intermediation in the ‘public interest’. Even in achieving some moderation of existing practice, regulators would then be required to counter the reactions and error-learning adjustments that would inevitably follow. A rain-dance is as likely to achieve success.