Singh And Stella on Collateral

A recent paper by Manmohan Singh and Peter Stella for the IMF Research Department (blogged version here) makes an interesting contribution to begin by questioning the money multiplier – one of the key concepts concerning how central banks are supposed to work.

Traditionally, they point out, “credit and money are [considered] counterparts to each other on different sides of the balance sheet”: banks create credit backed by the ‘base money’ which they hold as currency or deposits with the central bank. The money multiplier is the ratio of total monetary liabilities in the economy (the loans extended by the banking system) divided by the monetary base. With bank notes removed from the equation leaving only the ‘liquid reserves’ which banks have deposited with the central bank, the adjusted money multiplier is considered to provide a measure of the performance (or risk) of the commercial banking system in providing credit to the economy.

Through their ability to create or withdraw these reserves of base money by buying and selling financial assets from commercial banks (known as ‘open market operations’) the central banks are commonly considered to control a ‘transmission mechanism’ which determines the money multiplier. To increase the supply of money in the economy, the central bank can buy financial assets from commercial banks in exchange for liquid central bank deposits, which should provide the basis for increased lending both between commercial banks and into the wider non financial economy (the exact quantity of which should be predictable through algebra): QE has essentially been this carried out on a vast scale, and it is why many financial commentators assert that the measures will inevitably create hyperinflation. 

The paper challenges these notions, firstly by pointing out the by now well known fact that in the decades leading up to the financial crisis, the total amount of credit in the US economy increased far faster than the creation of new central bank deposits (see figure 2). Why? Only commercial banks can hold deposits at the central bank, but as the authors point out over the past 30 years the shadow banking system “has accounted for almost the entire growth in US financial deepening”. How? Though the use of liquid assets to be used in collateralized borrowing: highly rated securities can be used again and again as collateral for the creation of new loans in a process called re-hypothecation, and the increased use of securitization created a range of new alternative liquid assets to be used by the shadow banking system to create loans.

Distinctions need to be made between different types of collateral. On the one hand there is high quality collateral in the form of bonds issued by solvent sovereign governments, which can be used in credit creation almost anywhere. Other kinds of securities can be used as collateral under normal market conditions (and their use grew enormously in the run up to 2008) but they lose their utility in this regard in a downturn when markets begin to doubt their value. This happened most spectacularly with securitized housing market assets, resulting in a system wide liquidity crisis as interbank lending seized up.

The real fulcrum of liquidity creation, they argue, is to be found in the market practices relating to judgements of which assets are ‘acceptable collateral’ rather than with the central bank officials executing conventional monetary policy. Since the crisis, this has been reflected in two phenomena. Firstly, there has been intense demand for safe assets to use as collateral – hence the low price of UK and US government borrowing despite their indebtedness. Secondly, despite the unprecedented creation of new reserves at the Fed, there has been no corresponding increase in inflation (see table 1).

 

Since the crisis, chains of re-pledging have shortened from 3 steps in 2007 to 2.4 at the end of 2010, while the total volumes of re-pledged collateral have fallen from $10 trillion in 2007 to $5.8 trillion at the end of 2010. The actions taken by central banks in expanding their balance sheets has not, the authors suggest, been sufficient to make up for these losses: conventional QE substituted central bank deposits for high quality collateral that would have been re-pledged for collateralized borrowing, meaning that QE had nothing like the lubricating effect on the financial system that would have been anticipated under a conventional money multiplier view.

The banks’ liquidity crisis continues and, as the authors note

Unless there is some rebound in the pledgeable collateral market (by either an increase in ‘source’ collateral, or its velocity or re-use rate), the likely asymmetry in the demand and supply of good collateral may entail some difficult choices for the markets and the regulators.

Difficult choices such as whether a return to pre-crisis levels of global liquidity is even possible, let alone desirable.

Monetary policy is in uncharted territory and, they argue, as assumptions about the relative importance of the transmission mechanism against the role of collateral are revised, swapping bad for good collateral may now become a routine activity of central banks. This carries with it questions of risk and accountability which make the independence of central banks from democratic accountability seem more problematic than ever.