The hidden web of global liquidity
In this paper, Saleem Bahaj, Marie Fuchs, and Ricardo Reis examine the evolution and impact of cross-border central bank liquidity arrangements.
Comprehensive data collection: The authors constructed a novel dataset of 1,589 liquidity line agreements active between 2000 and 2024. Data was collated from central bank annual reports, press releases, and direct correspondence, categorizing lines by currency (USD, EUR, RMB), structure (bilateral vs. multilateral), and terms (collateral, reciprocity).
Network analysis & indirect connections: The paper maps the global network of these lines to identify “indirect connections” i.e. pathways where a country accesses a foreign currency through an intermediary central bank, for example, accessing USD via the ECB. They formally model how these indirect paths create “ceilings” on Covered Interest Parity (CIP) deviations, extending the theoretical framework of direct lines to the broader network.
- To test if network connections reduce financing costs, they built a monthly panel of CIP deviations for 42 countries from 2007–2023. They utilized quantile regressions to test if a country’s “degree of connection,” or distance from the source currency, correlates with the size of CIP deviations.
- To determine if liquidity lines substitute for foreign exchange (FX) reserves, they used a “stacked event study” design. They compared the evolution of a country’s reserves around the timing of a new liquidity line against a control group of similar countries. To establish causality, they isolated “exogenous” improvements in connectivity, namely, where a country’s access improved solely because a third-party partner signed a new deal, rather than through the country’s own actions.
Findings
Evolution and global reach
- The network expanded in three distinct waves: the 2008–2010 crisis (led by the Federal Reserve), a 2010–2015 bilateral expansion (led by the ECB and People’s Bank of China), and a 2020 pandemic response.
- While direct lines are limited, indirect connections allow the USD, EUR, and RMB networks to each cover approximately 80% of world GDP. The USD network is unique in relying heavily on indirect connections .
Network fragility and geopolitics
- The People’s Bank of China (PBoC) and the ECB are critical intermediaries. The PBoC acts as a central hub that, if removed from the network due to geopolitical tensions, 9 countries representing 20% of global GDP would lose access to the USD network.
- The network’s reliance on these hubs makes it fragile. While there is some redundancy, such as alternative paths via Japan or the UK, the exclusion of a major node like China would significantly lengthen connection paths and sever access for many developing economies.
Economic impact
- There is a strong correlation between network proximity and financing costs. Countries with direct or short indirect connections to a currency issuer experience significantly lower CIP deviations (smaller premiums for synthetic foreign currency borrowing) compared to unconnected or distantly connected countries.
- Liquidity lines function as substitutes for holding foreign reserves. The study finds that countries typically sign liquidity agreements following a sharp drawdown in their FX reserves. Furthermore, causal analysis shows that when a country’s network access improves exogenously, it subsequently reduces its holdings of FX reserves, confirming the substitution effect.
The Global Network of Liquidity Lines
Authors: Saleem Bahaj, Marie Fuchs, Ricardo Reis
From: UCL, LSE
The hierarchy of swap lines
Thomas Goda, Laurissa Mühlich, and Barbara Fritz investigate the determinants of swap line access and size, specifically comparing US Federal Reserve and PBoC strategies. The authors constructed a novel panel dataset of 440 bilateral central bank swap agreements (2007–2023), distinguishing between currency denomination, USD vs. RMB, and stated purpose, trade vs. financial stability. To estimate the determinants of swap access and size, they employed a two-step Heckman selection model:
- Probit Model (Selection Stage): Analyzed the probability of signing an agreement based on trade ties (FTAs, exports), financial openness, and political alignment.
- OLS Model (Outcome Stage): Estimated the size of the swap for successful agreements, correcting for selection bias.
The analysis ran separate specifications for Fed USD swaps, non-Fed USD swaps, and PBoC RMB swaps to isolate distinct strategic drivers.
Findings
- A stratified hierarchy: The Global Financial Safety Net is unequal. The Fed provides infinite liquidity to wealthy “core” nations , while the PBoC provides limited support to developing “periphery” nations.
- Divergent motives: Fed swaps are driven by financial stability concerns, in order to protect the US from spillovers, whereas PBoC swaps are driven by trade interests, specifically, promoting exports).
- The scale gap: Fed swaps are significantly larger, averaging 5.4% of recipient GDP, compared to 2.7% for RMB swaps. Recipients of Fed swaps also have double the GDP per capita of those receiving Chinese support.
- Persistent USD dominance: Because the RMB network lacks the scale and financial focus of the Fed’s network, it reinforces rather than challenges the US dollar’s hegemony in global finance.
You are worth that to me – Uncovering the hierarchy of central bank currency swap scale
Authors: Thomas Goda, Laurissa Mühlich, Barbara Fritz
From: Universidad EAFIT, HTW Berlin, Freie Universität Berlin
