Thread by @i_aldasoro

To understand post-crisis dollar funding markets, the growing heterogeneity of global banks dollar business models has to be the starting point

https://twitter.com/i_aldasoro/status/1133038625137016833

European banks fundamentally changed their operations in the US since 2011, towards short term arbitrage; Japanese banks continued to expand their dollar loan books 3/

This is, inter alia, reflected in the weighted average maturity of their trades with MMFs (JP preference for longer term funding) 4/

The MMF sector is, in turn, very concentrated: a few fund families account for the lion's share of the market. The MMF reform substantially increased concentration in non-repo (CP, CD, ABCP) segment 5/

JP banks' repo borrowing is very concentrated (50% from their top fund family counterparty, 70% from the top 2); non-repo borrowing more diversified than other banks 6/

The markets in which global banks and MMFs interact are OTC --> bargaining power, outside options matter a lot for pricing
Relative bargaining power differs depending market shares, and more
importantly, on how diversification in set of counterparties 7/

We document a "Japan repo premium": JP banks pay a price for repos with MMFs *in excess* of the risk premium reflecting bank and contract characteristics 8/

We identify bargaining power in repos using the window-dressing of European banks at quarter-ends (due to different implementation of leverage ratio), as well as the Overnight Repurchase Facility (ON RRP) of the Fed 9/ @DanielaGabor

@DanielaGabor In the non-repo market similar bargaining power forces are at play. Given their higher diversification, JP banks pay in fact a discount after controlling for relevant bank and contract characteristics 10/

@DanielaGabor We identify bargaining power in non-repo market using the MMF reform as natural experiment, exploiting heterogeneity in fund conversion from prime to government funds --> by just staying in market, funds from top 5 fund families saw market share rise 11/

@DanielaGabor Enter another interesting graph: all but JP banks are repo intermediaries (in particular FR banks); JP banks only do repos; yet MMF-provided repos make up around 10% on average of JP's repos (around 50% for FR, CA banks) 12/

@DanielaGabor Implication: opportunities for repo intermediation emerge (by standing in between MMFs and JP banks, doing maturity transformation) --> estimated spreads from such arbitrage: 16 basis points
Hypothesis: FR banks intermediate repos to JP banks 13/

@DanielaGabor Identification: window-dressing by FR banks.
--> FR banks quarter-end temporary retreat from repo market explains more than 65% of the contemporaneous variation in the JPY/USD cross-currency basis
$10 billion reduction in FR repos --> 6.5 bp wider 1-week (only 1w) yen basis 14/

@DanielaGabor Broader implications from paper: potential unintended consequences of regulation.
#1 MMF reform lead to more concentration (on an already concentrated market); this has pricing implications 15/

@DanielaGabor # 2 Inconsistent implementation of Basel III leverage ratio across jurisdictions has consequences for financial markets; i. quarter-end (as opposed to daily-averaging) reporting creates reshuffling of 100s of USD bln at quarter ends 16/

@DanielaGabor ii. this creates undue advantages for market participants by opening up possibilities for regulatory arbitrage; iii. it leads to complex and opaque dollar repo intermediation networks between global banks 17/17