Why Has Global Liquidity Crashed Again?

Authored by Michael Howell via CrossBorder Capital,

All around is crashing: Can the centre hold?

  • Global Liquidity falling at its fastest rate since 2007/08 Crisis

  • Over-zealous Central Banks are largely to blame. Not a broken banking system like 2007/08 Crisis

  • Squeeze evidenced by flat yield curves and negative ‘real’ term premia

  • In the US, the impact of ‘reverse’ QE is equivalent to another 10 rate hikes

We will enter 2019 with Global Liquidity tumbling at its fastest rate since the 2007/08 Financial Crisis. Yet again investors are learning the hard lesson that low nominal interest rates are a dangerously ambiguous guide to monetary conditions.

Already risk asset markets are skidding, in response to tight liquidity, and economic slowdown and probable recessions lie ahead. The future looks especially bad for those economies, firms and institutions that have spent the last decade kicking the proverbial debt can down the road. High debt levels always demand high liquidity to facilitate re-financing. Systemic risks rise if debt cannot be re-scheduled.

Figure1 shows the scale of the recent drop in Global Liquidity. Liquidity here refers to funding liquidity, rather than market liquidity, although the two are closely linked. Since end-January 2018, World private sector liquidity has fallen by some US$3 trillion, with roughly two thirds of the drop coming from the Developed economies, while World Central Bank liquidity has fallen by another US$1.1 trillion, with two-thirds of its drop recorded in Emerging Markets, paced by their large foreign reserve losses. Added together, Global Liquidity has in total fallen by just over US$4 trillion to US$124.1 trillion.

This 3% drop looks more serious when set against its 7% ‘normal’ trend. Put another way, after its brief recovery Global Liquidity has fallen back again to stand some 25% below its long-term trend. The table reports the latest weekly activities of the key Central Banks: only China’s People’s Bank (PBoC) is still expanding its balance sheet. And, measured in current US dollars, latest data show World Central Bank money is down by nearly 10% at an annualised 3-month clip.

What to Watch?

We monitor Global Liquidity by closely watching three channels:

  • Central Bank liquidity injections

  • Private sector liquidity provision

  • Cross-border capital flows

The first channel is now fashionably dubbed ‘QE’ or quantitative easing and measures the activities of policy makers in the money, repo and debt markets. The second looks beyond credit at all forms of cash generation by the private sector. It embraces bank credit, shadow bank credit and household and corporate savings flows in retail and, particularly, wholesale markets, and covers a history of financial engineering that extends back to the UK fringe banks in the 1970s and Japanese zaitech in the 1980s. Cross-border flows include all forms of net investment, but they are noteworthy because foreign currency borrowings, e.g. Eurodollars, are often used as collateral and levered up by domestic credit providers. It follows that Central Bank liquidity and cross-border flows represent what we term primary liquidity, while banks and shadow banks provide secondary liquidity.

We define Liquidity broadly to include ‘global’ or cross-border effects, and deeply, insofar that it extends beyond the traditional financial sector, to include corporate cash flows, and beyond retail banking by embracing wholesale money and repo markets.

The link between the volume of liquidity and interest rates was anyway never one-to-one: a fact that is especially true in the post-2008 period. Moreover, the link between bank reserves, money and liquidity has been similarly blurred; with the size of Central Bank balance sheets playing a more complex role in the funding structure since they simultaneously affect both the supply of cash and the availability of collateral. According to Adrian and Shin (2009):“The money stock is a measure of the liabilities of deposit-taking banks, and so may have been useful before the advent of the market-based financial system. However, the money stock will be of less use in a financial system such as that in the US. More useful may be measures of collateralized borrowing, such as the weekly series of primary dealer repos.” Adrian and Shin, Money, Liquidity And Monetary Policy, New York Fed Staff Papers, January 2009 Funding and the specific role played by Central Banks are critical factors explaining the liquidity cycle.

Central Banks have an outsized-effect in deregulated financial systems, where retail deposits are not the sole funding source, because what matters most is the ability to re-finance positions and at the margin Central Banks are the marginal suppliers of liquidity. Put another way liquidity is not fungible in crises, the very times that it matters most, and so Central Bank interventions are required. Since the supply of liquidity to roll-over existing positions matters more than the demand for finance for new projects, the size of the Central Bank balance sheet often outweighs the impact of interest rates. It follows that the relationship between interest rates and the supply of liquidity is rarely one-to-one. Central Bank interventions into the money markets significantly affect the elasticity of the financial system: in short, quantities matter and Central Banks increasingly determine the volume of funding liquidity and often directly impact the amount of market liquidity in modern financial systems.

Figure 2 shows the dramatic expansion in the size of the US dollar money markets to around US$9 trillion and the dominant role played by the US Federal Reserve in the period since 1980. These markets have increasingly supplemented retail deposits and now fund a rising proportion of US credit and liquidity, notably wholesale lending activity. Admittedly, following the 2007/08 Crisis, they have essentially flat-lined in size. Although the money markets are exploited by both traditional banks and shadow banks as financing pools, what sets traditional banks apart from all other financial institutions is their ability to issue liabilities, e.g. demand deposits, that serve the non-bank sector as a means of payment. Consequently, traditional banks do not face the same funding constraints as other financial intermediaries, so making their lending more elastic. In theory, as long as capital requirements are met, the traditional banking system can accommodate additional credit demands by simply creating new means of payment in the process of making new loans.

What shadow banks do is to transform these bank assets and liabilities and refinance them as longer and more complex intermediation chains, e.g. A lends to B who lends to C, etc. In doing this they provide alternative stores of value, e.g. asset backed securities, to institutional investors that do not want to hold all of their liquid assets as (uninsured) demand deposits. However, shadow banks largely repackage and recycle existing savings. By lengthening intermediation chains they became involved in large volumes of wholesale funding, without creating much new lending. The data show that they are involved in 66% of gross funding, but directly account for barely 15% of new lending. Shadow banks, therefore, increase the elasticity of the traditional banking system by relaxing banks’ capital requirements, through, say, selling loans externally to GSEs or internally to off balance sheet vehicles, so boosting the credit multiplier.

A speculative appetite to borrow always exists and seemingly is independent of interest rates. Keynes dubbed this the ‘unborrowed fringe’. Yet, shadow banks could not have originated the credit boom that preceded the 2007/08 Crisis, since they themselves depend on bank credit. The fragility of this wholesale funding model based on short-term repos has heightened systemic risks, not least because it is market collateral-based and highly pro-cyclical, and it always threatens to feed-back negatively on to the funding, as well as the lending books of traditional banks. Traditional economics misses the importance of this gross funding dimension, because it takes every credit as a debt (debit), every debt as a credit: so assets and liabilities must match, and the system always balances to zero. Thus, it never acknowledges just how big these numbers are: regardless of how much credit or debt there is in the system, the net figure is always the same. But knowing this fact is akin to scaling the World’s longest ladder and promising never to fall off!

The vital role that Central Banks play in this funding mechanism is also not well understood, even by the policy-makers themselves. For example, some experts have even claimed that collateral, such as Treasury notes that were purchased by policy-makers as part of QE operations, when released back into the market from the asset side of the Fed balance sheet:

“is a far better lubricant for the financial system than the reduction in reserves balances on the liability side of the Fed balance sheet. ... Thus, a leaner central bank balance sheet... could justify a much higher policy rate in this cycle than currently being anticipated.” Singh, FT April 2017

This is dangerous talk, particularly if it has recently played a role in persuading policy-makers to become more cavalier in their tightening actions. We maintain that the size of the Central Bank balance sheet serves as an unambiguous guide to the availability rather than simply the cost of primary funding. Thus, quantitative tightening is likely to have an out-sized effect in the modern financial system. This can be seen in Figure 3 which highlights the close relationship between movements in the size of the US money markets and changes in the size of the Federal Reserve balance sheet. The latest drop in the balance sheet spells out an ominous warning for US money market conditions.

This Is A Different Crisis To 2007/08

Therefore, we suggest that, unlike the 2007/08 Crisis which was more about a broken banking system involving the sudden collapse of leverage among over-extended banks and shadow banks, the current credit squeeze looks more like the 1997/98 Asian Crisis when Central Banks, led by the US Fed, tightened the supply of primary liquidity and cross-border flows rapidly retreated. This time around financial markets are probably even more interconnected and more global. Consequently, this could be an Asian Crisis-like sell-off, but one not only confined to Asia. This is shown in Figure 4, which depicts the two moving-parts that explain fluctuations in total credit – changes in the credit multiplier (black line) and growth in the monetary base (orange line).

Although the supply of primary liquidity dipped, the 2007/08 Crisis was dominated by a forced de-leveraging as shown by the skidding credit multiplier– the ratio between total credit and the official monetary base. In fact, pace its absolute decline in 2006, the US Federal Reserve’s balance sheet subsequently expanded rapidly as the crisis unfolded. This is shown by the orange line in Figure 4. The explanation for this collapse is that in the run-up to 2007/08 shadow banks had been borrowing against new collateral, such as US dollar deposits, and re-hypothecating existing collateral (i.e. the so-called collateral multiplier) to create what might be dubbed a ‘shadow monetary base’. We can crudely estimate the size this notional shadow base by keeping the credit multiplier fixed at its previous average value and calculating the funding ‘gap’ required to justify rising total credit. This is shown in Figure 5. On the same chart we have added the cumulative flows of cross-border capital to Emerging Markets because these flows were likely dominated by dollar loans that were probably being sourced from the same off-shore wholesale markets. Not surprisingly, the two series co-move closely.

The Collapse of World Central Bank Money

In contrast, today’s monetary problem is more about the other component, namely tight primary liquidity. This has four dimensions:

  • US Federal Reserve tightening

  • Tightening by other major Central Bank (e.g. ECB and BoJ)

  • USD Area tightening (e.g. Emerging Market Central Banks)

  • Legislative onslaught against the Eurodollar/ off-shore wholesale markets

Figure 6 examines the growth in Central Bank money, broken-down into three parts: the US Federal Reserve, Central Banks in the non-US Developed markets and Emerging Market Central Banks. The scale of the current liquidity squeeze goes beyond the US Fed raising the bar on its ‘reverse QE’ to US$50 billion per month and the ECB halting its QE from year-end. The chart shows how the US Fed has been tightening QE policy since 2015. The major non-US Central Banks started easing later and only began tightening earlier this year. Similarly, with Emerging Market Central Banks and as Figure 7 highlights, the EM cycle has been (as usual) largely dictated by underlying foreign exchange reserve movements. This, in turn, likely reflects an additional negative spill-over from US Fed tightening. Summarising the aggregate story, the chart shows that all three groups are now contributing to the sharpness of the overall liquidity slowdown.

The fourth component of tightening is harder to pin-down because data is scarce. However, we suggest that the offshore wholesale markets are under fire from the US Federal Authorities, who seem keen to regain control of US dollar liquidity.

The Eurodollar markets, which lie outside of US Fed or Treasury control, were a major factor behind the wayward shadow banking boom ahead of the 2007/08 Crisis. There have been three moves made to regain control: (1) the planned replacement of (uncollateralised) LIBOR with the new secured SOFR on-shore money market interest rate; (2) the 2018 tax amnesty that facilitated the repatriation of off-shore US corporate deposits, and (3) the recent removal of the tax allowance for interest payments on off-shore inter-bank loans. These official directives should substantially reduce the future attractions of using the Eurodollar markets. One way to show their impact is in Figure 8, which plots the amount of net funding that US-based banks receive from international banks. Since this represents dollar funding, the likely foreign sources are the Eurodollar markets. Figure 8 highlights the sizeable decline from the US$750 billion peak in 2015.

Other Evidence of Tightness

The scale of this tightness can be seen in two other statistics: (1) the ‘real’ term premia on US 10-year Treasuries, and (2) the flat slope of the G4 yield curve. The 10-year Treasury is the canonical ‘safe’ asset for World investors and a low term premia suggests a high demand for ‘safe’ assets. To remove the distorting effects of inflation expectations on term premia, we calculate an inflation-adjusted series or ‘real’ term premia. When the net demand for safe assets is ‘normal’, the real term premia should be around zero: greater demand for safety pushes it lower, and greater risk appetite among investors should push it higher. Consequently, these real term premia movements provide another way of judging whether monetary conditions are too loose or too tight.

Figure 9 plots the ‘real’ term premia for 10-year Treasuries against Fed Funds. It shows that because of the effect of reverse QE, the ‘true’ Fed Funds rate is nearer to 5% than the prevailing 2.5% target: in other words, tight liquidity conditions are equivalent to the Fed undertaking around 20 rate hikes rather than the nine it has so far implemented this cycle.

In Figure 10, we show the relationship between Global Liquidity and the term structure of World interest rates as depicted by the slope of the G4 government yield curve. Term premia again play a role here because tight liquidity conditions, by forcing investors into demanding more ‘safe’ assets, push up bond prices and simultaneously pull down their yields and term premia. Thus, the general ‘flatness’ of yield curves across the major economies again testifies to generally weak Global Liquidity conditions. It is simply not the case, as many suggest, that Central Bank QE lowers bond yields. Rather, because government bonds are ‘safe’ assets, ‘reverse QE’, in fact, causes lower yields, and QE raises bond yields. This follows because term premia widen as the extra liquidity persuades investors to take more risks, so shifting asset allocation from ‘safe’ bonds to risky equities.


We expect a major policy easing in 2019. However, only China’s PBoC (People’s Bank) among the majors is so far easing monetary policy. We anyway see China as leading this cycle. The US Fed is likely to follow given the scale of tightness in domestic and Global Liquidity and this must involve greater liquidity injections, rather than simply interest rate cuts. We have no view whether this takes the formal shape of an explicit ‘QE4’ policy or if it involves an unannounced increase in the size of the Fed’s balance sheet.

Whichever, the immediate implications will be a yield curve steepening and ultimately a weaker US dollar. Financial shares and Emerging Markets may prove the main beneficiaries.