We have previously speculated extensively on the recent appearance of direct bidders as a key participant in Treasury auctions. What is known is that the direct take down during the last 2-3 months has at least doubled for most coupon auctions up to and including the 7 year. What is not known is/are the identities of the bidder(s). We have provided some observations on the topic previously (here and here) although our preliminary conclusions are based on circumstantial evidence at best. Additionally, we have highlighted that even as direct bidding take downs have increased, bid-to-cover ratios have reached near record highs, which in itself is also paradoxical and the only immediate explanation is that this is simply a confirmation of Say's law, as this phenomenon certainly does not fit the normal supply/demand pattern.
Our obsession with the direct bidder conundrum is easily explainable as this is a new and very critical presence in the treasury bidding process. The last thing primary dealers need is a source of volatility in the auction process, which could potentially have destabilizing consequences on this most critical cog in funding America's future record deficits. Today, Daniel Kruger at Bloomberg picks up on the topic and thrusts it front and center into the public spotlight, his analysis further confirming our concerns.
Three of the nine primary dealers that met with Treasury officials ahead of today’s announcement of the government’s quarterly financing plans said they’re concerned about the increase in so-called direct bids, according to people involved in the discussions. The Treasury said it would sell $81 billion of 3-, 10- and 30-year securities next week.
Depending on the validity of Kruger's source, this observation should be quite troubling, as it implies the identity of the primary bidder is not even known to the PDs, traditionally entities who are all to aware of all key players in the auction market, be they counterparties or not.
Kruger goes on to highlight the increasingly prevalent role of direct bidders:
Direct bidders accounted for 10 percent or more of the total in 12 of 42 fixed-rate auctions since July, compared with only 6 times from 2004 to 2008, according to Treasury data. The Wall Street firms say the increase in bids sent directly to the Treasury by investors including banks, large money managers and hedge funds may raise borrowing costs for the Treasury and taxpayers if dealers bid less aggressively because of higher volatility at the sales.
“It doesn’t crash our market, but it becomes an interference, and it will become more costly to place the debt,” said James Caron, head of U.S. interest-rate strategy in New York at Morgan Stanley, one of the 18 primary dealers that also act as counterparties to the Federal Reserve when it sets monetary policy.
This observation in itself is odd, as the bid to cover for indirect bids has continued to be in its historic range of about 1.3x. All else being equal, there should be no reason why direct take down should be spiking the way it is (granted, we don't have the full dutch allocation thru the high allotment so we are speculating).
Unlike PDs, Washington does not seem to bothered by the recent change in allocations:
“This is a good thing from the standpoint that it breeds competition and helps us achieve our goal of financing the government with the lowest cost over time,” Matthew Rutherford, deputy assistant secretary for federal finance, said today in Washington. “We do not encourage or discourage customers from bidding either directly or indirectly. Ultimately we think this is a business decision that they need to make on their own.”
Another alternative implied by Kruger, is that the direct bidders are purposefully avoiding direct bidders. If so, why?
“The Fed and Treasury have a lot of expectations and requirements for primary dealers and you don’t want direct bidding to interfere with their ability to underwrite Treasury debt on a consistent basis,” Bitsberger said. “On the other hand, anyone should have the right to bid directly without going through a primary dealer.”
Here is an example of how primary dealers stand to suffer material and immediate losses if an auction is not immediately beneficial to them:
Primary dealers betting on a jump in direct bids at the Treasury’s $13 billion sale of 30-year bonds on Jan. 14 were whipsawed when demand wasn’t as high as anticipated. The sale was preceded by a record direct bid for three-year notes and an above-average 10.8 percent direct bid for 10-year securities the prior two days.
The 30-year bonds, which traded at a yield of 4.68 percent just before the bidding deadline close, were sold at a yield of 4.64 percent, with dealers putting in bids of $1.84 per dollar sold, compared with the 2009 average of $1.63. Dealers, who won 54 percent of the offering, or $7 billion, typically sell the securities in a short-sale before the auction and then buy them back at the auction at a lower price. Direct bidders won 4.9 percent, down from 6.9 percent at the prior bond sale.
The bids by the dealers led to the four-basis-point difference in the yield before and after the auction, which may have lost them about $44.8 million, according to Thomas di Galoma, head of Treasury trading at Guggenheim Securities Inc., a New York-based brokerage for institutional investors.
Confirming our previous observations about the increasing strength of auctions throughout the year, Kruger notes the following:
The average ratio of bids to debt sold at 10- and 2-year note auctions is the highest since at least 1993, when the government began reporting the data. On average, the Treasury received $2.625 in bids for each dollar of debt sold at 10-year sales compared with an average of $2.413 from 2004 through 2008. For two-year debt, bidders asked for $2.937 for every $1 sold last year, compared with an average $2.391 from 2004 to 2008, Treasury data shows.
So as it stands, courtesy of Bloomberg we have yet more datapoints which however merely add to the mystery to both the identity and the recent surge in direct bidder interest. Once released, we will analyze December TIC data for hedge fund interest (Carribean, UK banking centers) and juxtapose it with the increase in Direct take downs during the November and December period. We are fairly confident the data will provide no statistical correlations whatsoever, bringing us back once again to our hypothesis #1.