The Treasury sells the same 10-year note three times. The first auction is the original issuance. Two reopens follow in the same quarter, each adding to the outstanding amount of the same cusip. The 30-year bond runs the same three-auction cycle. The two-, three-, five-, and seven-year notes get a fresh cusip every month and do not reopen. The reopen cycle is the reason the 1:00 ET auction time on a Wednesday in the middle of a quarter is the second-most-watched event of the day at a rates desk, behind only the 8:30 ET data release that set up the auction.
The bid-to-cover ratio is the headline. It is the total dollar amount of bids divided by the dollar amount the Treasury accepted, and it gets the first line of the post-auction recap on every wire. It is also the metric that tells the rates desk the least. A reopen of a 10-year note at $39 billion that draws $90 billion in bids prints a 2.31 bid-to-cover, which sounds strong and rounds to the same number as the prior six auctions in the series. The number that tells the desk something is the tail.
The tail is the read
The tail is the gap between the high yield the Treasury accepted at the auction and the when-issued yield trading in the secondary market at 1:00 ET, the moment bidding closes. A positive tail (the auction stopped at a higher yield than the when-issued market was quoting) means demand was weaker than the market priced. A stop-through (the auction stopped at a lower yield than when-issued, the inverse of a tail) means demand was stronger than the market priced. Both move the curve in the half-hour after the auction, and the size of the move is roughly proportional to the size of the tail in basis points.
The reference range that desks carry: the median 10-year reopen over the last five years has tailed by about 0.2 basis points, which is statistically indistinguishable from on-the-screws. A tail of 1.0 basis points is a soft auction. A tail of 2.0 basis points or more is a weak auction and reprices the 10-year by 4 to 8 basis points by 2:00 ET. A stop-through of 1.0 basis points is a strong auction and reprices the 10-year 3 to 5 basis points the other way. The auctions that move the curve hardest are the ones that surprise relative to the when-issued, not the ones with the biggest bid-to-cover numbers.
The reason the tail dominates the bid-to-cover for the desk read is that the when-issued yield already incorporates the bid-to-cover the market expected. The auction either confirms the when-issued or it does not. The tail is the size of the confirmation gap.
The indirect bid is the second read
The indirect bidder share is the percentage of the auction allocated to indirect bidders, a category that captures foreign central banks, sovereign wealth funds, and large foreign and domestic asset managers that bid through primary dealers rather than through the direct-bid channel. The reference range for 10-year reopens has run between 65% and 75% indirect share since 2022. A print outside that band is the data point that gets attention.
A high indirect share (above 72%) generally means foreign official demand is firm, which historically has been the demand floor for the long end of the US curve. A low indirect share (below 64%) means the demand is coming from the primary dealer takedown channel, which the dealers then need to distribute, and the price discovery during the distribution period puts upward pressure on the yield over the next 24 to 48 hours.
The indirect share is the metric that picks up shifts in foreign demand for US duration. It does not show up cleanly in the bid-to-cover because the bid-to-cover counts the dealer indications-of-interest, which dealers always submit at scale. The indirect share counts what was actually allocated, which is closer to what was actually wanted.
The direct bid and the dealer takedown round out the picture
The direct bidder share captures institutional buyers (mostly US asset managers and hedge funds) that bid directly through the Treasury’s TAAPS auction system rather than through a primary dealer. Direct shares run between 12% and 22% on a 10-year reopen. A direct share at the upper end of that range typically means a specific class of buyer (often a large fixed-income manager rebalancing) showed up at the auction, and the next day’s secondary market often shows continued demand from the same channel.
The primary dealer takedown is the residual: whatever the indirects and the directs did not take, the dealers had to. A dealer takedown above 22% on a 10-year reopen is the signal the auction did not clear cleanly at the when-issued, and the dealers will be working to distribute the excess inventory in the secondary market through the rest of the week. That distribution shows up as upward pressure on the cash 10-year and a steepening of the 5s-30s curve.
Why the reopen mechanic matters
A reopen sells additional supply of an existing cusip rather than introducing a new cusip. The market has already priced the original 10-year auction at the start of the quarter, and the two reopens that follow get graded against the demand the first auction established. A strong original followed by a soft first reopen is the signal the quarter’s demand was front-loaded. A soft original followed by a strong reopen is the signal the market got the quarter’s clearing yield wrong on the first read and the second auction found the right level.
The reopen size is typically smaller than the original. A 10-year quarterly cycle generally runs $42 billion at the original, $39 billion at the first reopen, and $39 billion at the second reopen. A change in the announced reopen size (Treasury announces auction sizes via the quarterly refunding statement) is a separate signal: an increased reopen size means Treasury is funding more at the 10-year point, which is supply-bearish for the long end and the curve.
The single sentence that compresses the auction read: when the tail prints inside one basis point and the indirect share prints inside the 65 to 75 band, the auction confirmed the when-issued and the curve does not move. When either number prints outside, the curve moves in the direction the surprise pointed. Everything else, including the bid-to-cover headline that runs in the wire, is the framing the recap leans on rather than the data the desk actually trades on.