1. Country differences:
There are specific country mechanisms at play. The DMO engages directly in the repo market in most countries (AUS, DE, FI, DEN, NL, PR, Spain, UK) and through the central bank (independent from open market operations) in Brazil and the US. Australia, New Zeeland and Canada use both the DMO and central banks. DMOs in Italy, Japan and Malta do not intervene at all in the repo market for various reasons: no need (Japan), operational barriers (Japan and Italy)
First question raised here: is this a hidden form of coordination central bank/fiscal authority? What rules govern the interaction between the two?
2. Reasons for intervention:
DMOs and Central Banks carry out REPO operations for cash management purposes, to reduce market squeezes on specific bonds, to increase secondary market liquidity.
This is the interesting bit because it can indicate the extent to which the DMO can provide support to the bond market, particularly in times of stress (a la Spain) when liquidity spirals ignited by the private financial sector can threaten the liquidity of the sovereign bond market. The second and third (reduce market squeezes and increase secondary market liqudity) could both be used for this purpose.
Dissagregated data is provided: DMOs in 9 countries use repos for cash management, 5 to reduce market squeeze and 7 to increase secondary market liquidity.
Italy's DMO does not intervene in any form and so has renounced to make use of a possible stabilization tool. It should because its sovereign bond market is the second largest contributor to the European private repo market (ICMA data).
3. Types of repo:
Most DMOs engage in both special repos (using a specific asset) and General Collateral (an agreed basket). With special repos (right-hand graph), DMOs can target particular market segments, and respond to private demand for special repos that reflects high demand/shortage of a particular asset, often for trading/speculative reasons rather than for funding (the GC repo).
4. Sourcing collateral.
This is where DMOs are like no other repo market player. Private players source collateral from their own portfolios or by borrowing (sec lending). The DMO can issue ad-hoc tranches and cancel them after the repo operations: almost half of respondents do. Viewed from the shadow banking literature that treats repos like money-like instruments, there is an important parallel here: DMOs can do for collateralized financial intermediation what the central bank does for traditional banking, creating additional 'shadow' money to ease repo market tensions/shortages.This is when governments (DMOs) become central banks.
In sum, we now know more about the potential role that sovereign debt managers can play to offset the pro-cyclicality underpinning the repo market. The modus operandi across countries are very different, and some questions are left un-answered.
- Do DMOs use haircuts/margin maintanance for repos beyond overnight? If they do, as some central banks (the ECB), then the counter-cyclical function will be diminished.
- Can DMOs afford not to intervene where their sovereign bond market is an important source of collateral (Italy?)
- What explains the difference between different approaches, and what are the benefits/drawbacks of coordinating withe the central bank?