First it was the TBAC's May presentation "Availability of High Quality Collateral" piggybacking on reasoning presented previously by Credit Suisse. Then JPM's resident "flow and liquidity" expert Nikolaos Panigirtzoglou rang the bell on regulatory changes to shadow banking and how they would impact the repo market and collateral availability (and transformation) in an adverse fashion. Now, it is the turn of Barclays' own repo chief Joseph Abate to highlight a topic we have discussed since 2009: the ongoing contraction in quality collateral as a result of transformations in shadow banking and the Fed's extraction of quality collateral from traditional liquidity conduits (i.e., QE's monetization of bonds). To wit: "Several recent regulatory proposals will increase the pressure on banks to reduce assets that carry low risk weights. Repurchase agreements are a large source of banks’ low-risk assets, and we expect banks to reduce their matched book operations in response to these proposals."
- Given inelastic demand for repo, we expect banks to increase margins to offset reduced volumes.
- The implications for market liquidity are potentially more severe, as banks and dealers may become less willing to facilitate short positions and warehouse risk.
- A reduction in repo activity will reduce the supply of investible assets for short-end investors, widen bid-ask spreads and could encourage the development of riskier substitutes for collateralized borrowing.
- Smaller securities inventories could reduce cash market liquidity. In the Treasury market, these banks may become less aggressive bidders at auctions.
- Likewise a reduction in repo balances will limit the market’s ability to transform weaker collateral into the high quality assets necessary to meet new margin requirements.
What this means explicitly for repo volumes is nothing new and has been covered here extensively in the past, and can be seen quite vividly in the periodic slide of OTR repo into "special" status, usually just ahead of Treasury auctions:
We expect repo volumes to fall for both U.S. and foreign banks. Repo in the U.S. has already fallen by 40% since its pre-crisis peak (Figure 1). We believe the U.S. repo market can reasonably be expected to decline 10% and possibly more across domestic and foreign banks as a result of these new rules and their proposed revisions. In fact, we believe this has already begun, although empirical evidence is difficult to find outside of discussions with money market funds regarding the decline in repo availability.
The most concrete evidence that the pressure on repo has already begun is the occurrence at quarter-end of plunging bank repo borrowings as primary dealers tighten their balance sheets ahead of end-of-period reporting. This is particularly acute for money market funds, whose repo balances from primary dealers typically fall 15% ($75bn) in the final day of a quarter, only to recover over the first two months of the following one. Money funds have attempted to replace some of these lost balances from non-primary dealers, but the credit ratings of these other institutions prevent much of this potential repo replacement. We estimate that nonprimary dealer collateral replaces only about $5bn of the $75bn lost on quarter-end from primary dealers. Unsurprisingly, since money fund balances generally have the same amount of cash to invest at quarter-end, the closing out of primary dealer repo positions results in a steep plunge in repo rates on the final day of the quarter.
The danger from the imposition of a "higher supplementary leverage ratio" is that while on the surface banks may demonstrate stronger balance sheets and less counterparty exposure, it will simply force them to seek "near-substitutes" that behave not exactly according to plan. See Lehman.
Lower repo and bill yields may seem like a small price to pay for stronger bank balance sheets, but there is a strain of research that argues that such scarcity encourages the creation of near-substitutes that may not behave quite like the original. Auction rate securities are an example of an imperfect substitute that under normal market conditions was a perfectly adequate liquidity investment, but ultimately became illiquid during the financial crisis.
Digging into the implications of "thinner dealer inventories" which is already a direct result of changes in the repo market:
In addition to maintaining inventories of securities for proprietary trading and their own portfolios, as significant capital market’s players, these banks hold stockpiles of Treasuries and other securities as part of their market making operations. Net positions in Treasuries across all the primary dealers totalled $80bn in mid-July. Market making – along with hedging inventories – are generally financed in the repo market so a reduction in repo balances will translate into thinner market-making inventories. Similarly, with less dedicated balance sheet, we expect banks to become less aggressive participants in the Treasury’s auctions. This could result in sloppier auctions with bigger tails and ultimately, higher borrowing costs for the Treasury.
A significant reduction in repo might reduce the ability of a dealer or other investors to short securities – without fear of delivery fails. In the Treasury market, a higher risk of failing and consequently incurring the 300bp fails charge, might influence how aggressively dealers participate in auctions and might affect actual auction pricing – even though fails are currently very low (Figure 2). Further, the possibility of increased fails could mean greater volatility in rates as dealers’ willingness to make markets and intermediate between buyers and sellers declines. To avoid the risk of fails, banks and dealers will spend more time searching for available supply to short. Higher search costs reduce the ability of the market to match buyers and sellers, further widening bid-ask spreads.
Repo also plays an important role in intermediating between buyers and sellers outside the Treasury market. Any activity in which banks act as securities “distributors” requires some amount of financing. Whether this financing is done via the repo market will depend on the asset’s underlying liquidity. By reducing balance sheet capacity, the leverage rules are likely to reduce market liquidity as banks ration their available inventory capacity. We expect this rationing will increase bid-ask spreads in any market where banks act as market makers between buyers and sellers.
Finally, and getting straight to the heart of the TBAC's May warning, we focus on the biggest issue of all: collateral transformation, or literally getting something from nothing (or absolute junk) courtesy of the murky, off-balance sheet transformations that take place in the shadow banking system:
In a typical transaction, the customer pledges weaker collateral to the bank in return for higher quality collateral, such as Treasuries or cash. These are then used to meet the customer’s margin requirements on derivatives trades. Since these trades are typically structured as repo transactions using the bank’s balance sheet, they are sensitive to the new supplementary leverage requirements, especially as other regulations are expected to increase the demand for this type of collateral transmutation significantly.
Dodd-Frank sections 731 and 764 change the margin rules for cleared (trades over an exchange) and un-cleared or bilateral derivatives transactions. These are expected to boost the demand for high-quality liquid assets over the next six years. According to estimates from the U.S. Treasury, the incremental demand for high-quality liquid assets to meet the new margin rules will be $1.6-3.2trn. These estimates grow considerably larger if, during the six-year interval estimated, there is a financial crisis or shock that results in a sudden, sharp increase in the demand for safe assets.
Although the supply of high quality collateral to meet this new demand source is likely to come from government issuance as deficits persist globally, we expect banks also to play a role via collateral transformation. After all, most of these derivatives customers are likely to be naturally long non-eligible collateral – such as equities or corporate bonds – but somewhat less so with respect to Treasuries and cash. As intermediaries, banks can use their balance sheets and repo transactions to “extend” the supply of eligible margin collateral for their customers – for a price. As noted above, this is the type of higher margin business for which banks might be willing to use their capacity constrained balance sheets, provided the spread between lower and higher risk collateral is sufficiently wide. To the extent that the supply of government collateral is insufficient to meet the new margin demand, transformation trades are an important safety valve if bank balance sheet capacity is available.
Note the highlighted, underlined text above, because it gets to the bottom of the tapering discussion: in a world in which "deficits persist", the Fed has all the liberty to do as much QE as needed, and absorb quality collateral as much as new gross collateral is injected via the Treasury (i.e., deficit funding). However, as deficits decline, as they have been in the US for the past few quarter (at least until the housing picture inevitably deteriorates again and the GSEs go from source of government funding back to use, and the demographic crunch hits in 2015 and deficits explode higher once again), ongoing monetization means collapsing the high quality collateral pool of assets far more than the TBAC advises. In fact, it is the TBAC that is pulling the strings on the Fed and making it clear, as it did in May, that the Fed has no choice but to taper as long as deficits don't return on their normal, upward trajectory (whether this means a war is inevitable to boost contracting defense spending remains to be decided).
However, it is this aspect of the Fed "cornering itself" that is the underlying driver of all behind the scenes tapering discussions, which have nothing to do with the economy. Because while all the rhetoric and regulatory discussion focuses on the economy, the open markets, and deficits - this is all for general popular consumption. What the Fed and the BIS are truly concerned about, and have been for the past four years, is ongoing instability in the repo world, and other aspects of shadow banking. The last thing the Fed, and Treasury, will want to do is override the TBAC on its advice to moderate collateral extraction and plough on with even more quality asset imbalances in shadow banking. And this is why, at least until the Treasury proceeds to spend itself back into "drunken sailor" mode, that all hopes that the Fed will monetize everything that is not nailed down, will have to be deferred indefinitely.
And with Credit Suisse, the TBAC, JPM and Barclays all having now covered this issue, we look forward to that final "liquidity" guru, Citi's Matt King to chime in on precisely this topic, and make it clear just what the real considerations driving the Fed's tapering "logic" are at this moment.