Three QE controversies, two theories

BNP Paribas Asset Management‘s head of macro research, Richard Barwell, explores three controversies surrounding quantitative easing through the lens of the quantities and signals theories, and considers whether all asset purchase programmes should be considered monetary policy.

Central banks creating reserves to purchase financial assets, or quantitative easing (QE), was once referred to as ‘unconventional’ monetary policy. But now it is an established part of the central bank toolkit. However, there is still no consensus around key questions such as: How does it work? How should it be done? And why?

How it works

Unfortunately, the same fundamental problem that the distinguished former chair of the Board of Governors of the US Federal Reserve System, Ben Bernanke, identified almost a decade ago remains. There is still no consensus on how QE works in theory. We are still debating the same two basic theoretical mechanisms – quantities versus signals – through which QE could significantly and sustainably influence bond yields, and therefore broad financial conditions, and ultimately output and inflation.

The quantities mechanism is the application of a basic economic argument that a shift in demand will affect the price. QE involves the central bank boosting market demand for a specific financial asset – typically government bonds. The supply of bonds is assumed to be unchanged, at least in the short term. The impact on prices (or equivalently, yields) therefore depends on how large the QE operation is, relative to the size of the market and how sensitive the market demand for each security is to a change in price.

If investors view different financial assets as close substitutes for one another then, as soon as QE starts driving up the price of government bonds, the current holders will look to sell, realise the capital gain and invest in other similar securities. Asset purchases would not have a major impact on asset prices as investors rebalance their portfolios. If, instead, market demand for government bonds is price-inelastic, then the price of those securities may have to rise significantly before the existing holders are willing to sell. In this case, QE can drive bond prices up and bond yields down, squeezing the compensation investors typically demand for bearing interest rate risk. However, sceptics argue that any significant distortion in asset prices would soon be arbitraged away – except in febrile market conditions.

An alternative school of thought is that QE impacts bond yields via a signalling mechanism. The decision to buy bonds supposedly reveals information about the state of an economy or its central bank’s strategy. Investors then revise their expectations about the future path of monetary policy and that triggers a shift in bond yields. However, the power of QE to move markets via signalling rests on asset purchases revealing incremental information. It is unclear how valuable that hidden signal could be, given that central banks habitually provide extensive guidance on the future path of policy and pride themselves on the credibility of their commitment to price stability. What is clear is that QE is an expensive way to send the market a signal and even then the central bank is relying on the market to decipher the message accurately. Moreover, the relevance of any signal is likely to decay through time and, hence, so too should the impact of QE on asset prices via the signalling mechanism.
 

How it should be done

Views on how QE works should guide how QE is implemented. Let’s start with the question of what assets the central bank should buy. Policy-makers that believe in the signalling mechanism should not be overly concerned about what assets they buy. After all, according to that view, the actual transactions in capital markets are irrelevant. In contrast, policy-makers that believe in the quantities mechanism should be more interested in what they buy because the location of asset purchases can dictate the impact on asset prices.

The choice of how much QE to undertake will reflect a view on how powerful the QE is. Unfortunately, there is huge uncertainty about the quantitative impact of a round of asset purchases on inflation, and there is no clear guidance on what that means for the conduct of QE. The folk wisdom in macroeconomics is that policy-makers should take a gradualist approach with uncertain tools: do less than looks necessary, and do more later as required. But there is no hard and fast rule in the academic literature. There are moments when it pays to be activist, and err on the side of doing too much.

The Bank of Japan’s innovation in 2016 challenged the central bank orthodoxy on the most effective way of implementing QE. The Japanese central bank cut through the uncertainty around how QE works by setting a target for long-term yields and then adjusting bond purchases to deliver that outcome. That yield curve control strategy has proved successful: the Bank of Japan has maintained the yield peg through infrequent use of fixed-rate purchase operations while gradually reducing the underlying pace of purchases. However, that is not to say it is easy to control yields under all circumstances and, in particular, if investors believe the peg is no longer consistent with fundamentals. The Reserve Bank of Australia, for example, was unwilling to maintain its peg in the face of rising yields in late 2021.

There are constraints on QE but these also depend on how QE is supposed to work. Policy-makers that believed in the signalling mechanism would likely cease and desist QE once the expectations of the future path of the policy rate that are implicit in long-term yields collapsed on the lower bound on the short rate. The maximum signal would have already been received. In contrast, policy-makers that believed in the quantities mechanism might continue buying risk assets to ease financial conditions. That brings us to the final controversy.

Why it should be done

QE has been used on an industrial scale by central banks to achieve price stability for more than a decade. But it does not follow that all QE programmes are monetary policy operations. There are important exceptions.

The size of the QE multiplier for any given announcement depends on the message the market wants to hear and so is inherently uncertain

As previously discussed, even those sceptical about the power of the quantities mechanism will concede that central banks can move prices by adding to demand in dysfunctional markets. It is not reasonable to expect financial institutions to arbitrage away the impact of QE by shorting the bonds the central bank buys in such febrile market conditions. There is consensus that QE is therefore most effective in illiquid, dysfunctional markets, which has morphed into a consensus that QE is the appropriate tool for stabilising dysfunctional markets. This is not quite the same thing as stabilising inflation.

There is nothing new about using asset purchases for financial stability purposes, even if those purchase programmes are often misunderstood as classic monetary policy operations. In a 2009 speech, Bernanke famously distinguished between QE and credit easing (CE). While the ultimate objective of both programmes might be the same – to ease financial conditions, stimulate aggregate demand and thereby restore price stability – he argued that the focus of the latter is different. The intermediate objective of CE is to reduce credit spreads and improve the functioning of credit markets. One way to think of this is that QE is using asset purchases to inject stimulus via the monetary transmission mechanism to preserve price stability, while CE is using asset purchases to repair the monetary transmission mechanism to preserve the quantum of monetary stimulus chosen by policy-makers.

It has long been understood that central banks should act as lenders of last resort (LOLRs) to support systemically important and solvent financial institutions that suffer a transitory loss of liquidity. The idea that a central bank could act as a market-maker of last resort  (MMLR) – the capital markets analogue of LOLR – came back into fashion during the financial crisis that began in 2007–08. The central bank stands ready to purchase securities – or, more accurately perhaps, charge a smaller bid/ask spread than constrained market-makers – when liquidity drains out of systemically important capital markets. The MMLR leans against bloated liquidity risk premia embedded in financial assets, and thereby prevents a tightening in financial conditions that would otherwise threaten the pursuit of price stability. That doesn’t make an MMLR intervention a monetary policy operation.

Beyond the MMLR is the possibility of the central bank acting as a risk-taker of last resort (RTLR). Now the central bank is using asset purchases to lean against unwarranted widening in credit risk premia in dysfunctional markets in circumstances where the market starts to price disaster scenarios, and asset prices crater. One could potentially think about the European Central Bank’s Outright Monetary Transactions scheme as an RTLR operation: designed to prevent unwarranted redenomination risk premia becoming embedded in euro-denominated assets leading to a self-fulfilling crisis. One can also think about the asset purchases that took place in March 2020 in this context. Indeed, a senior official at the Bank of England described the purchases that took place during the early stages of the Covid-19 pandemic as central banks acting as “buyers of last resort”. Once again, there is no doubt those operations were successful in stabilising markets and, ultimately, the economy. Equally, it seems clear those purchases should be thought of primarily as financial stability operations.

It is interesting to imagine what would happen if these problems re-emerged in a scenario in which inflation was too high. The case for alleviating stress within a pocket of financial markets would not disappear just because the central bank wanted to achieve a broad tightening in financial conditions in an attempt to bear down on inflation. This hypothetical scenario is not entirely unrealistic; indeed, it is theoretically possible that a sharp tightening in monetary policy could trigger stress within pockets of financial markets. Careful thought, therefore, needs to be given to the governance of asset purchases for financial stability purposes so this situation can be managed effectively.
 

Conclusions

QE is an established part of the central bank toolkit. However, significant questions remain about how it works, how it should be done and why it should be done. The focus is currently on how QE should be reversed, or so-called quantitative tightening (QT), but the same fundamental issues apply. For example, does QT influence the economy via the quantities mechanism or by transmitting an unspoken signal and, if so, what signal? What is the purpose of QT? To restore price stability or restore policy space to tackle the next episode of financial stress? Attention will soon swing back to these QE controversies in the next downturn. 

Sponsored content