By Jeffrey P. Snider,
The Federal Reserve is running another large-scale test on its Term Deposit Facility (TDF), the account by which it intends to establish a European-style floor to its also intended interest rate corridor. Judging by the fact that nobody has paid any attention to this test or the one that preceded it, except for those that were fooled into thinking the Fed was actually tightening "reserves", it is fair to say that there has been no disruptions to "market" function at this point. That is certainly a welcome development, as the latest weekly H.4.1 has the TDF back over $400 billion again, since the state of "dollar" liquidity does not at this moment need any push into greater imbalance or dearth.
However, the other side of that tranquility is that there is no indication whatsoever that the TDF is having its intended effect. While the FOMC doesn't want to attain massive disruption in funding markets they do want to see at least some minor shuffling. Locking up "reserves" into the TDF is supposed to reduce the quantity of funding available elsewhere, as that is the mechanism (purportedly) of control whereby the FOMC will establish that floor.
The primary funding conduit affected by the TDF "should" be repo, yet there is no hint, again, that repo markets have seen any diminishment. GCF repo rates continue along as if the TDF did not exist.
The current test is for 21-day tenders at a 3 bp spread over IOER (25 bps), with the 0.75% early withdrawal penalty in effect this time. The institutional max was set at $20 billion. The effective rate of 28 bps is at least 10 bps above the current GCF repo mix, meaning that if the TDF was intended as even a quasi-hard floor, it hasn't come close. Again, there "should" be at least a spike up to or around 28 bps in some GCF rates, if the TDF was having the effect of "drawing off" excess liquidity by locking it up outside the "reserve" account.
And so the only tangible effect has been that, where "reserve" balances shrink dollar for dollar with any tendered into the TDF. So, along with being an accounting phenomenon of "absorbing" reserves, any balances tendered into the TDF seem to be coming only from the "reserve" account; thus banks are only trading one form of Fed "delivered" free money for another.
There may be some point at which the TDF acts as intended, either in terms of the interest rate offered or the size of the accepted overall tender, but even at $400 billion they are clearly not even close yet. That seems to prove, yet again, an axiom of central bank "control" in that it has to be excessive to be effective and thus a much higher risk of being disruptive, therefore eluding control in the first place.
You can make that case more easily outside of the still counterfactual case of the TDF by examining collateral flow. There has been little doubt as to QE's impact on collateral availability in terms of fails, especially since June 2014, but that is also apparent in dealer stock. Primary dealer inventory of bills and coupons form the central axis of liquid collateral, as there are various versions of collateral conduits.
These "silos" or market stores of collateral largely developed in the late 1990's and early 2000's as a financial means to escape heavy and persistent repression of risk prices. This is another factor that traces itself back to the relative "expensiveness" of bank "capital" as interest rates are squeezed as a matter of intentional policy (for the demand side). As the FOMC gained full control over the short end, financial institutions looked further and further afield to find ways to boost profitability.
Owing to the explosion in wholesale finance (or shadow finance) there was a direct need for liquid collateral sources. Pension funds and insurance companies (like AIG who went way, way too far) acted as major sources of collateral as they held trillions in securities and their managers were keenly aware of the growing suppression of nominal profitability. The early securities "lending" business was a boon to financialism profits, because there were relatively fat spreads to be made on lending the "choicest" collateral - which included, to all our collective detriments, any kind of MBS.
While those silos were good sources of mostly OFR (off-the-run) almost backup securities, the primary dealers were the anchors in the much more fluid, liquid and important OTRs (on-the-run). We can clearly see the buildup in the housing bubble, backed by the huge expansion in wholesale liquidity, just by looking at dealer inventories of coupons.
Dealers exhibited a massive net short position in coupon UST all the way until the crisis hit. Very much like the contrary behavior of the "dollar" that I described earlier today, the dealer positions here were not at all related to interest rate movements or economic growth potential. The short position indicated that dealers were extending liquid collateral into the marketplace. Thus the opposite, where dealers held less net short or even net long, meant dealers removing and "hoarding" collateral.
With that generalized view in mind, the track of dealer inventory follows spasms of illiquidity almost perfectly. The scale of the change after the panic was immense, as dealers were net short (supplying) $178 billion in coupons in the middle of July 2007, but wound up net long (hoarding) $33 billion by early April 2009 when it all finally relented (from FAS 157 changing mark-to-market). By the time of QE4, dealer's net long went as far as $100 billion, for a total systemic swing of almost $300 billion. That's a huge amount of liquid collateral that went from easily available to beyond hard to find.
Equally representative, in the post-crisis era the tendency of dealers to extend liquid collateral, or "dehoard", back into the market follows the familiar pattern of QEs that we see in other liquidity functions. What we don't see is the compression in spreads owing to QE that squeezes so much profit out of the securities "lending" business, that was itself an outgrowth of prior repression. That is why, despite the huge expansiveness of QEs in successive size, the return of collateral diminishes each time.
As you can plainly see, the current rise in dealer holdings (hoarding) is just a fraction of what took place in the prior episodes - the very kind of asymmetry that marks systemic de-capacity. The incongruence with the "dollar" is astounding, as liquidity is but a shell of itself owing to distortions upon distortions upon distortions, all with the central aim of rendering "control" to maintain order.
Of course, repo silos and available liquid collateral are not the sole explanation for this asymmetry in "dollar" tightening. However, the diminishment in liquidity capacity is certainly a major factor in amplifying any changes in "dollar" behavior, which is the entire point of liquidity in the first place - to absorb changes in selling or funding so that wider market disruption never occurs. So the overall context since June 2014, again, is that liquidity capacity is impaired while "dollar" bank balance sheet supply contracts with little left to absorb that "shock." The result is the "dollar" at a more-than-decade "high", which signals nothing good about pretty much anything.