Read on Twitter
Have revised my paper on collateral supply and central banking to incorporate March 2020.
March events raise significant questions abt viability of 'dealer of last resort', CCP for Treasuries, & standing repo facility proposals 1/
cc @PMehrling @dandolfa
https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3545546
March events raise significant questions abt viability of 'dealer of last resort', CCP for Treasuries, & standing repo facility proposals 1/
cc @PMehrling @dandolfa
https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3545546
1. To stabilize Treasury market, over the course of 18 days, Fed had to purchase 5% of all marketable Ts outstanding: $812 billion. Nobody anticipated that 'safe' assets would require DoLR support to this extent. 2/
Vast sums needed to support market also throw doubt on claims that incremental regulatory relief or a shift to CCP for Ts can address the problem of ephemerally illiquid market-based finance. $812 B in 2.5 weeks is more than the dealers can handle. 3/
2. Mechanics of the DoLR are also a problem in the US. Fed purchases necessarily add to bank reserves. But $0.8T can't be added to reserves without impairing bank capital adequacy. In short, bank intermediation of DoLR is a big problem. 4/
For the present the solution is regulatory forbearance. In US leverage ratio has been converted for 1 yr into a RWA measure (& we know how well those work, cf Basel II). In short, current financial regulatory structure is poorly thought-out and relies on 'duct tape' solutions. 5/
If Fed is really supposed to be DoLR, then need public FedAccounts (as per @MorganRicks1 @LevMenand) + Duffie's CCP for Ts, so the DoLR can bypass dealer banks. Then Q is whether resulting direct monetary stimulus every time DoLR is needed is good policy. 6/
3. Note that before DoLR was tried, Fed expanded massively its repo facility, but as per FOMC, repo offered wasn’t taken up, and this did nothing to improve T market liquidity. This is clear evidence that a Standing Repo Facility will not address repo mkt problems. 7/
My preferred solution: recognize that modern repo markets -- and in particular their monetization of long-term Ts -- is a destabilizing mistake, built on the illusion that markets provide liquidity. In fact, market liquidity is inherently ephemeral (Keynes 1936). 8/
Thus, a principal policy goal should be to use regulatory policy, such as minimum haircuts on long-term debt, to establish an incremental policy of eliminating use of anything over 7 year debt in repo markets and discouraging repo of 2-7 year debt. 9/9
