Government bond auction results are an important short term driver of the Fixed Income markets as they provides insights on the demand for securities by market participants. Strong auctions are good for sentiment and could indicate real buying to come into the market if any institutions missed out on buying in the auction because they didn’t bid high enough. The question then is how to read bond auction results.
Note: If you are interested in US Treasury auctions, please check our specific reading US Treasury auction recap – analysis and terminology is a bit different. The details below are relevant for all European and UK auctions.
Was the auction result good, bad or just ok?
To determine if the results of the auction are good, bad or just ok we need to analyse the results against the expectations for that auction and previous similar auctions. Many traders look first for bid to cover data (probably wrongly in many cases). In general we think the order of importance is (1) average price (over or under bidding), (2) price tail, (3) bid to cover data and (4) amount of bond sold (where relevant). Each of these factors is discussed in detail below.
InTouch Fixed Income provide these details with their previews of each auction.
Bid to Cover
Many traders see the bid to cover as the headline indicator of whether the auction was strong or weak and the market often reacts directly to this figure, although it’s debatable how useful it really is. It’s popular because it’s easy to understand, it’s widely available (quoted by all newswires) and at face value it is a good proxy for the strength of the auction. However, the cover can in theory be manipulated or inflated by dealers submitting bids significantly below the expected auction price of the bond (optically makes it look like significantly higher demand for the bond). For this reason it may not be the most reliable indicator of auction strength.
Bid to cover is the ratio of the amount of bids in the auction vs the amount of bonds sold. The higher the cover is, generally the better the auction. To put the result in context, traders compare the cover for an auction against covers in previous auctions in the bond, or bonds with a similar maturity. Some compare the cover to the average from the last 3 or 4 auctions. If it’s higher, that would indicate a strong auction while lower would indicate a weaker auction.
A closely related statistic is the Real Bid to Cover. This is a statistic that we calculate for some government bond auctions where a government doesn’t sell all of the new bonds being issued and instead retains some of the bonds for themselves. The most high profile example of an issuer that does this regularly is Germany. This is explained fully in another article – Real Bid to Covers in Government Bond Auctions.
Can the bid to cover value be lower than 1.0? Strictly speaking, no; it’s not possible to sell more bonds than you get bids for. However, there are cases where an issuer may be unable to attract bids for the full amount of bonds that they wanted to issue. In these cases, the bid to cover may be reported as a figure below 1.0 as the calculation may be modified to reflect the number of bonds the issuer intended to sell, rather than what they actually sold. The real bid to cover (where applicable) can also be below 1.0 – again, see our separate article.
Amount of Bonds Sold
Some issuer Governments use auctions with a fixed size (eg US, UK) and the amount of bonds sold will always be in line with that size, unless the auction “fails” and the issuer doesn’t get enough bids to cover the fixed size – a clear sign of a weak auction.
Other issuers are more flexible and offer a size range of bonds for sale (eg France, Spain). In theory, selling an amount of bonds in the mid-point of the range would be an ok auction, while selling in the upper end of the range indicates a stronger auction. However, some of these issuers have, in practice, almost always issued at the top end of the range and traders look through this “positive” indicator and if the issuer ever sells an amount of bonds below the upper end of the range, that is seen an indicator of a very weak auction. Having insight into prior behavior by the issuer is therefore very important.
It is also helpful to cross-reference the size of the auction against recent auctions (per auction bond) to determine how useful the bid to cover data actually is. If, for example, the amount of bonds being sold is much smaller than prior auctions, it would not be surprising to see much higher bid to covers.
An unexpected increase in the size of an auction can lower the bid to cover. The implications for the market is dependent on why the issuer decided to increase the size. If it’s due to increased demand for that particular security and the pricing data is still in-line with normal, this should be considered positive. If on the other hand, the increase of the issue size results in weaker pricing than normal, it is likely the issuer is attempting to ‘stuff’ the market with bonds they may not necessarily want or need which would be a negative.
Probably the single most important indicator of whether the auction was strong or weak. The prices at which bonds were sold in an auction are a much more reliable indicator of how strong or weak the auction was as they can’t be manipulated like the bid to cover data.
Price Overbidding or Underbidding
Did buyers in the auction pay more or less than the prevailing market price? Ie The difference between the average price at which a bond sold at auction, and the market price of that bond just before the auction bidding deadline (the “snap price”).
If the auction average price is above the snap, that’s “overbidding”. If the average price is below the snap, that’s “underbidding”. Big overbidding is positive. Low overbidding or underbidding is negative.
However, context is everything. Some issuers or maturities often see high overbidding in auction, whilst other bonds typically have softer pricing. Traders need to compare the overbidding for an auction against the overbidding the previous times the bond, or a bond with a similar maturity, was auctioned. Ie the higher the overbidding vs prior auctions, the stronger an auction is deemed to be.
Having access to accurate, comprehensive data is very important to make these comparisons. InTouch Fixed Income has a multi-year database covering all European, UK & US issuance which we leverage in analyzing auctions for our clients.
Some countries include an over allotment option (“green shoe”) when they issue bonds, giving the buyer the option of buying more bonds for a few days after the auction at the auction price. This is effectively a “free” call option for buyer in the auction and makes the auction bond more attractive. As such, these green shoes can result in significant overbidding during the auction, especially where the market is expected to be volatile in the green shoe period after the auction due to important economic data releases, central banks meetings, etc.
The Price Tail
Did all the buyers in an auction pay a similar price, or did some get the bonds on the cheap? The price tail is the difference between the average price paid in an auction and the cut-off price (the lowest price at which a bond was bought in the auction). A big price tail is seen as a negative. Why? Because the marginal buyer of the bond was only willing to pay a much lower price than the average.
The price tail gives us more information and context to judge the overbidding. An auction with high overbidding but an even bigger tail could be interpreted as weak. Again, it’s good to compare to prior auctions.
Some auctions are run as “Dutch” auctions (everyone pays the same price which is equivalent to the ‘clearing’ price). In these cases, there is no price tail. The biggest example of this are US Treasury auctions. In Dutch auctions there is no penalty to bidding much higher than the secondary market price going into the auction. This means the average price is likely distorted higher compared to an auction whereby everyone pays their bid price.
Secondary Market Price
Checking how the auction bond has performed in the market before the auction, both in recent days and immediately before the auction, gives more context.
If the bond was weak into the bidding deadline (as is typical) then dealers have built in a good concession and high overbidding and covers may be expected. Conversely, low overbidding or covers would indicate that even the lower price on the bond has not attracted buyers in the auction and suggests a particularly weak auction.
If the bond rallied into the bidding deadline (ie no concession) then even moderate overbidding could be seen as very positive as buyers were willing to pay above the already high market price.
A rough rule of thumb how to read bond auction results is to assess (1) average price (over or under bidding), (2) price tail, (3) bid to cover data and (4) amount of bond sold (where relevant), in that order of importance, and comparing each one to prior auctions in the same term. The market sometimes changes which metric it is most focused on but that is a good rough guide.
Authors: Michael Colman, Robin Belec