The one thing the central banks won’t say about QE

“Only when the tide goes out do you discover who has been swimming naked.”

This oft-cited quote is generally attributed to Warren Buffett. The point is obvious; the high tide can conceal all sorts of worrying sights, such as the lack of credit discipline from an investment manager or the lack of swimwear on your fellow swimmers.

However, underlying the quote is a fact that we think is often lost on the investment community: tides go in and eventually, they go out.

We touched on this idea last month when we introduced Average Jane and highlighted the degree to which her personal circumstances were impacted by fiscal stimulus on the way in. We posed the question as to how her personal circumstances would be impacted on the way out.

This month we want to discuss how markets may be impacted as (if?) the king tide(1) of monetary stimulus is unwound (i.e. goes out). When we engage in these debates with our peers there are generally two retorts:

1. QE will never be unwound(2); or
2. It will be unwound in an orderly fashion.
   

We want to focus on this from one particular perspective that we don’t believe is getting enough attention: the impact of QE on market liquidity. But before we do this, let’s go back to first principles.

A market is liquid when there are a large number of heterogenous participants active in that market. Note the use of the word “heterogenous”. When a group of buyers are heterogenous they are buying or selling for different reasons. This aids liquidity because a change in conditions for one participant doesn’t necessarily result in a change in conditions for the remaining participants. Within fixed income markets, consider the different motivations of banks (buying short dated repo-eligible paper for liquidity purposes), superannuation funds (buying for a hedge with equities), life insurance companies (buying long dated for yield and liability matching), P&C insurers (focused on short dated liquid bonds), foreign domiciled investors (some focussed on unhedged yield, some on hedged yield) and fund investors (some focussed on outright yield, some focussed on spread). All participants are buying or selling for different reasons. This aids market liquidity and ensures prices adjust in a continuous fashion to new information.

There are limited studies considering the effect of QE on market liquidity (as opposed to those that show the impact on outright yields). Those that do have generally shown that transaction costs decline in response to QE programs.(3) So liquidity improves as the tide comes in. But what happens to market liquidity as/if the tide goes out?

A 2020 study by the ECB looked at exactly this question.(4) They highlighted that where the market had a higher share of “preferred habitat investors” i.e. investors focussed on a certain part of the yield curve, market liquidity would eventually suffer as central banks became too high a share of the bond market. This may not be an issue in Australia but certainly is globally. Just look at the change in ownership of the US treasury market where the Federal Reserve alone now owns 22% of Treasury securities having taken 55% of the increase in securities in 2020.

The European study defined preferred habitat investors as insurance companies, pension funds and monetary authorities. By this measure, preferred habitat investors are now 36% of the US treasury market up from 30% at the end of 2019. According to the aforementioned study preferred habitat investors in the Euro area were 32% at the end of 2018 and are undoubtedly higher now.

Major Holders of US Treasury Securities

Source: SIFMA

We were reminded of this study in late February as liquidity conditions in the most liquid parts of the market deteriorated sharply. As the below chart shows, during February/March bid ask spreads in 30 year treasuries tripled from year end 2020 levels and were more than 4 times the 5 year moving average, wider than the taper tantrum of 2011/12.

Bid Yield Minus Ask Yield (in bps) in US 30 Year Treasuries

Source: Bloomberg

Faced with these conditions, central banks responded with largely short term measures, solving the liquidity issue in the short term but potentially exacerbating it further out the curve. To wit, the ECB accelerating the pace of bond purchase programs and the RBA increasing costs of shorting government bonds.

While these policy actions seem to have calmed markets for now, we are growing concerned that the most liquid bond markets in the world are characterised by increasingly homogenous pools of capital. 

Central banks can only be a buyer or a seller; they cannot be both at the same time and other participants recognising this fact, will follow. When the central banks are buying, everyone is buying. When they are selling, everyone is selling. 

These conditions reflect a market where the price reaction function is becoming increasingly dis-continuous and markets, increasingly illiquid.

This has important implications: a key input into our appetite to invest in an asset is our ability to sell out of it on a timely and cost effective basis. Assets that were once liquid on a daily basis with minimal costs, may not be in the future. If you need daily liquidity in your portfolio, make sure it will be there when you need it and don’t trust it will be there simply because it was in the past.

Learn more

We aim to provide diversified sources of income by seeking opportunities in both the public and private sectors, whilst maintaining capital stability. For further insights, please visit our website.

Footnotes

(1) In December, Morgan Stanley noted that in 2020, G10 central banks will have injected US$4 trillion via government bond purchases alone with another US2.8 trillion expected for 2021.

(2) A side note is that most people that make this case don’t tend to think about the implications. Saying QE will never be withdrawn is an equivalent statement to saying QE can never be withdrawn.

(3) Karamfil (2020) found that the Corporate Sector Purchase Program reduced bid offer prices by 6 basis points. Christensen and Gillan (2016) found that QE programs in the United States reduced liquidity premiums in TIPS (Treasury Inflation Protected Securities) by 12-14 basis points.

(4) Ferdinandusse, Freier, Ristiniemi (2020), Quantitative easing and the price-liquidity trade-off