Central bank balance sheets in the major economies now range from 20 to 30 per cent of GDP, two-thirds of which is due to the emergency measures that have been taken to stimulate economies since the crisis began. As far as I am aware, the global scale of this action is completely unprecedented.
Critics see these bloated central bank balance sheets as the culmination of a decades-long process in which periods of excessively easy monetary policy have caused inflation, asset price bubbles and misallocation of investment flows in the economy. For example, this FT piece by former Republican presidential candidate Ron Paul concludes as follows:
“The world is awash in US dollars, and a currency crisis involving the world’s reserve currency would be an unprecedented catastrophe.”
Such an outcome is perhaps conceivable, but is far from inevitable.
Let us examine the monetary consequences of what the central banks have done. The common feature in the US, the UK and the eurozone has been a large rise in the monetary base compared to other monetary aggregates such as M1 to M4. It is important to realise that these are very different types of “money”. The monetary base is mostly the reserves of the commercial banks held at the central bank. M1-M4 are mainly deposits of varying maturity held by the public at the commercial banks. The monetary base can behave very differently from the wider aggregates, and with very different consequences for the economy at large.
The recent rise in the monetary base has occurred because the central banks have purchased sovereign debt from the commercial banks, and have credited the banks with reserve balances at the central banks to settle these transactions. Since the commercial banks have simultaneously wanted to increase their holdings of liquid balances in the safest possible form, in order to secure their future funding requirements, these balances have simply remained at the central bank doing nothing.
Persistent increases in the monetary base have, in the past, generally been accompanied by higher consumer price inflation. Mervyn King pointed this out in a forceful speech in 2001, analysed in this blog. How might this happen again? Monetarist models of the economy generally assume that there is a fixed ratio between the monetary base and M1-M4. In fact monetary textbooks often do not distinguish between these variables at all, referring to them simply as “M”. This obscures an important difference between various types of money.
The fixed relationship between base money and M1-M4 applies when bank lending is constrained by reserve requirements and banks are eager to increase their lending. In those circumstances, a rise in the monetary base or banks’ reserve assets leads to an automatic “multiplier” rise in bank lending, and then in the bank deposits which comprise the M1-M4 monetary aggregates. None of this is happening now, since bank lending is not constrained by reserve requirements and banks do not want to lend.
Until this changes, there is only one remaining monetary route for recent central bank action to cause inflation. That would involve a direct link arising between the growth of base money and inflation expectations. If this were to occur, it would represent a major catastrophe, since central banks would be faced with a nightmare choice between accommodating inflation and risking financial instability.
This link has arisen in some grim historical episodes – Germany and France in the 1920s, and Argentina in the late 1980s, for example – but all of these episodes have occurred during severe economic environments in which the central bank has been seen to be subservient to insolvent fiscal authorities. This could conceivably happen in current circumstances, and that is where many of today’s inflationary risks ultimately could lie. But for the moment a direct link from the monetary base to inflation expectations seems rather far fetched.
Japanese experience in the 1990s and 2000s suggests that the creation of base money may have little effect on the economy, other than to keep banks liquid. In that sense, base money may be the “wrong” sort of money, or at least an “insufficient” sort of money. The “right” sort of money involves the creation of bank lending and bank deposits, as measured in M1-M4. The question which then arises is whether the central banks can create more of the “right” sort of money.
Tim Congdon is the UK’s leading specialist in this area, and he has been arguing for several years that this is possible. His area of expertise has come into its own in the current climate of shrinking bank balance sheets and constrained growth in broad money. (See for example this pamphlet and his comprehensive book about monetary policy, “Money in a Free Society”.)
Tim’s key point is that the central bank should buy government bonds and other assets not from the banks, but from the non-bank private sector. This results in a direct increase in bank deposits held by households and corporations and, as a side-effect, it also increases the monetary base so all types of money increase simultaneously. This does, however, assume that banks are still able to meet their capital ratios as their balance sheets rise. If this assumption does not hold, then any immediate rise in M1-M4 might eventually be offset by lower bank lending as banks struggle to restore their capital ratios.
Tim argues that this type of operation was carried out successfully in both the US and the UK in the 1930s, and that it helped end the Great Depression. The Bank of England’s version of quantitative easing has in fact followed this approach, but its success may have been limited by the stringency of the Basel III capital contraints. Meanwhile, neither the Fed nor the ECB seem interested in this approach.
If the central banks were to refocus the direction of their bond purchases, and also ensure that banks’ capital ratios were not too burdensome, they might make quantitative easing more effective than it has been so far.