Observations On Reverse Repos Coupled With A Record Excess Bank Reserve Balance

Tyler Durden's picture

Those who follow the meandering permutations of the Fed's balance sheet must have observed with great irony the proclamation by the NY Fed on October 19th that it is prepared to commence tightening liquidity via reverse repo operations, even as 48 short hours later later the Fed announced a new all time high in bank reserves, which for the first time ever hit a level over $1 trillion. The glaring discrepancy between these two observations has left many wondering not only about the veracity of any statements coming out of the Fed, but to consider what the best trades to front-run the Federal Reserve may be for that time when, whether it likes it or not, the NY Fed is forced (politically or otherwise) to start extracting its pound of flesh from the banking system.

The chart below demonstrates the dramatic pick up in excess reserves, which after lying relatively dormant around the $800 billion level, have shot up by over $200 billion in less than two months.

What is particularly notable is that the excess reserves hit a record level hours after rumors swirled that the Fed's reverse repo test conducted this Monday was a failure, and that instead of attempting to interact with banking institutions in a test attempt to raise $200 billion in liquidity, the Fed was forced to approach money market funds instead. Indeed, as Dow Jones reported, money market activity for the latest week would confirm speculation about major MM outflows, with $41 billion withdrawn from such funds, even as MM levels have continued to decline persistently, and hit a recent low of $3.339 trillion. With a 20 bps risk-free arb courtesy of the Fed deriving from MM record low rates at 0.05%, while the target fed-fund rate is at 0.25%, it is no surprise that capital will continue to flow toward the Federal Reserve.

It would probably require Sheila Bair to come out with another YouTube video in which she placates fears that money market investors are now collateralized by precisely the same toxic MBS garbage that banks, domestic investors, China and, most recently, Pimco have all decided to prudently keep far away from.

Not surprisingly, the primary source for bank reserve expansion continues to be the Fed's ongoing monetization of Treasurys and MBS. And even as the first is winding down, the second still has around $400 billion in dry powder left. Add to that the nearly $200 billion in additional liquidity that is forthcoming courtesy of the wind down of the Supplemental Financing Program, which is the Fed's latest "two birds with one stone" program of keeping the US debt ceiling from being breached, and allowing banks to purchase yet more risky assets, and Americans are easily looking at an excess reserve balance well north of $1.5 trillion within 3-6 months: money that should be lent out, but isn't. After all why take risks when you can make risk free money courtesy of the US taxpayer (the interest the Fed pays on the reserves comes straight out of the taxes Americans pay. Damn it feels good to subsidize Wall Street's kleptocracy class with every biweekly federal tax withholding). As for the party line, here is what the Fed says about funding excess reserves:

Why is the payment of interest on reserve balances, and on
excess balances in particular, especially important under current conditions?

Recently the Desk has encountered difficulty achieving the operating target
for the federal funds rate set by the FOMC, because the expansion of the Federal
Reserve’s various liquidity facilities has caused a large increase in
excess balances. The expansion of excess reserves in turn has placed
extraordinary downward pressure on the overnight federal funds rate. Paying
interest on excess reserves will better enable the Desk to achieve the target
for the federal funds rate, even if further use of Federal Reserve liquidity
facilities, such as the recently announced increases in the amounts being offered
through the Term Auction Facility, results in higher levels of excess balances.

It is sad that the only way to keep deflation in check is to effectuate this kind of wealth transfer out of America's working class into the kleptocratic oligarchy. 

Yet perhaps the reverse repo approach will at some point prove successful (even if it did not this past Monday). Here is what market participants read between the lines from the NY Fed's October 19th reverse repo announcement:

1. Ensuring reverse repos will be ready when and if the FOMC decides to use it
2. Possibly expanding the set of counterparties in reverse repos
3. Expanding reverse repos to ‘triparty’ settlements

While the statement is traditionally broad and non-committal, Morgan Stanley does highlight one aspect of it:

We stress that neither did the NY Fed statement confirm that no reverse repurchase operations were in the works before year end, as it again has no capability to make such decisions. This is an important nuance that the market likely missed, which may cause it to misprice the tail risk of a near-term drain.

In other words the Fed may or may not do something or another, depending on what the callers from speed dial 1 through 5 (name your top 5 banks as sorted by declining amount of interest rate swaps in the trillions) instruct Bernanke to do.

An issue meriting further attention is the concept of triparty reverse repos. As Zero Hedge recently pointed out, the NY Fed has been so careless in using these in the past, that in the case of Lehman Brothers it actually used the stocks of numerous bankrupt companies as collateral for the $40 billion triparty repo backing up Lehman's balance sheet in the days after it filed for bankruptcy.

Here is Morgan Stanley's interpretation on what the use of triparty reverse repo will mean:

A tool that the NY Fed may ultimately rely on to drain large quantities of reserves is a triparty reverse repo. This is the least operationally intensive way for the NY Fed to drain large quantities of reserves from a large number of participants (currently only dealers but in time may include GSEs, money market funds & other institutions).

 

One effect of the Fed’s potential reliance on triparty reverse repos to drain excess reserves is that large quantities of collateral will be made available to the market – so if the Fed decides to drain say $200 billion of excess reserves, approximately $200 billion of extra collateral will be available for GC [General Collateral]. This will cause GC rates to rise, potentially above overnight fed funds or above OIS when speaking of term.

And in case investors would like to gradually step in against prevailing popular thought and take the other side of the Fed trade, on the assumption that at some point the Fed will be forced to stop printing trillions of new dollars, MS shares the following:

The fact that banks are not arbing out the risk-less spread between the overnight fed funds rate and the 0.25% overnight deposit rate at the Fed is another example of an ‘arbitrage’ that is often left on the table (Exhibit 2). In this case, one of the reasons that this spread has not collapsed to zero as banks compete to borrow below the deposit rate and invest it overnight at the Fed at 0.25% from the supply side is because credit line limits may be reached that prevent cash-rich non-bank participants from lending beyond a certain limit to the commercial banks. From the demand side, commercial banks may not want to increase the size of their balance sheet that would be involved in earning this spread. We mention this example in passing only to demonstrate that when and if the triparty repos are done in size , they may disrupt the historical OIS / GC spread relationship and seeming ‘arbitrage’ opportunities may not entirely keep this spread from inverting.


Therefore, if investors are not too comfortable with going short risky assets on the bet of liquidity tightening (the topic of a later post), a levered bet on an OIS-GC spread tightening may be one much less risky approach to play the contra-Fed strategy.

And some additional thoughts on this trade from MS, which believes that the market is certainly not discounting for reverse repos hitting the market any time soon:

Talk of reverse repos has added some uncertainty to the market, but front-end rates have barely budged. Since bottoming out at 0.13% at the start of October, the 6-month US T-Bill rate has remained near its lows at 0.16% (Exhibit 3). Meanwhile, the 3-month US T-Bill rate has not moved at all and GC to the end of the year is now lower than it was at the beginning of the month.

 

Further, most of the volatility remains further out the curve between the 1s/2s Treasury spread (Exhibit 4 - light blue) and not in the 6m/1y Treasury spread (Exhibit 4 - dark blue). This indicates that the market is no more worried about higher front-end rates in the very front end of the curve than it was at the beginning of 2009 – and continues to be more worried about higher rates / the level of curve further out in time, say in 1-years time.

 

This implies that investors who believe OIS / GC spreads will tighten can take advantage of current entry levels.

 

What the market is saying is that not only is the Fed not "in danger" of tightening any time soon, but that its disclosed preferred mechanism of liquidity extraction via reverse repos is not likely to be effectuated any time soon, especially in light of the failed Fed test on Monday. The "don't fight the Fed" ideology has become everyone's bedtime mantra. And as long as the Fed is willing to throw trillions simply for the sake of redefining triple digits P/E multiples as the new "normal," so be it. Madoff also thought he could run his scheme for ever. Either that, or he simply did not think about the consequences of what happens when the status quo changed. The Federal Reserve is now caught squarely in the same position. And, just like Madoff's ponzi scheme imploded with $50 billion disappearing overnight (and no, there was no greater fool to throw the hot potato to), so the same will happen with the US capital markets one day. The event is a certainty, the timing is still unknown - but keep an eye out on such data points as the OIS-GC spread for some early warnings.