The Euro area comprises 13 countries with 13 different government bond markets, 13 different debt agencies or branches of finance ministries, different definitions for primary dealers, different auction procedures, etc…etc. What a market! And I didn’t even mention the different taxations and the nightmare of dealing with withholding taxes across countries when you manage European portfolios. To that extent, the Euro area is worse than Ferguson’s Switzerland. In comparison to the global government bond markets, the Euroland market does not appear to be a simple one. If we as Europeans find it complicated, then what about non-European investors?

My point in this instalment of European Perspectives is that the Euroland government bond market needs more unification, more harmonisation, more standardisation…and quickly. At a time when large international investors are accelerating their diversification out of U.S. Treasuries and when the euro is becoming a growing reserve currency, European governments urgently need to provide the political will necessary to enhance the European bond market. The sovereign debt issuers together with the European Commission should agree to make the Euroland market more transparent and easier to trade.

January is Sales Season
The European financial industry has made a great deal of progress in the last years to break many barriers within the Euroland bond market. This progress includes the allocation of International Securities Identification Numbers (or ISINs, a number to identify each individual securities issue), some standardisation in communication channels (the bond market language has become clearly English, for example) and the organisation and standardisation of the Euroland corporate bond sector. The development of intermediary-sponsored electronic trading platforms has also tremendously facilitated the trading of government bonds in the secondary market for investors like PIMCO. However, the public sector has been slow to make progress in the way it brings its debt to the market. In the primary market for government bonds, we cannot fail to recognise a list of issues, which should be addressed in order to optimise the Euroland bond market.

As a bond investor I am always amazed by the apparent lack of coordination between the sovereign issuers in terms of their calendars. In fact, there is no issuance calendar at the Euroland level. The only way to get it is by trying to collect the information provided by the national debt agencies as they are released. No easy task. Some debt agencies are early in providing the market with an issuance calendar for the year (as soon as December for the next year in the case of Germany), others only give an “indication” of what the issuance calendar might be (Belgium, for example) and others do not have any issuance calendar at all (Austria and Finland, for example). Debt agencies understandably want to keep some flexibility across the year to adapt their issuance to evolving market conditions as well as public finances, but some coordination would certainly make sense for them as well as for investors.

This lack of coordination becomes obvious when we look at the 30-year sector of the market for example. This sector is particularly vulnerable to flows (issuance, large purchases, etc.), hence the notion that it would make sense to distribute the issuance throughout the year instead of different sovereign issuers supplying the market with new 30-year bonds almost at the same time. Unfortunately this is not the case. There are periods of several months without any 30-year issuance followed by a peak around a particular date as Table 1 shows. Between 25 September 2006 (30-year issue from Belgium) and 9 January 2007, the primary market was effectively “closed” for 30-year sovereign nominal bonds in Euroland. Then, on 9 January, the Netherlands tapped their existing 30-year benchmark, on 10 January Austria issued €3 billion of 30-year bonds and then on 11 January it was Italy’s turn to tap its 30-year benchmark for €2 billion. On 25 January, Germany issued a new 30-year benchmark for €6 billion and on 30 January Greece supplied the market with €4 billion of 30-year Greek bonds. As a result, out of about €50 billion of 30-year issuance in the last 12 months, more than a third took place in the first 45 days of this year.

Is there a sales season for 30-year bonds? It almost looks like one, and this in turn impacts the market. The sector consequently tends to underperform during periods of concentrated issuance while 30-year European bonds tend to do well in the absence of supply. Better distribution of issuance along the year would make the market more efficient. This is an issue that PIMCO, together with the Euro Debt Market Association (AMTE), has been raising and which is currently under discussion together with issuers and intermediaries.

No Single Way to Issue a Government Bond
My friends at the debt agencies used to argue that some competition between the countries can be useful and ultimately serves the interests of the end investors. We can see the point in terms of new products, with France having led the way in the inflation-linked bond market a few years ago, for example. But what is the point of having different issuance procedures and primary dealer definitions?  The issuance techniques themselves have room for standardisation in Euroland as the examples of France and Italy show.

In France, bond auctions take place on the first and third Thursdays of each month. Primary dealers announce their bids by 10:50 a.m. on the day of the auction and the results are released a few minutes later. Primary dealers then have the option, until the following Monday, to purchase an additional amount at the average price, with the size depending on their past market share. In France, the system is called “purely competitive” since bidders pay the price they submitted, and the amount potentially issued in addition is 15% if the option is fully filled. 

In Italy, the process is different. When a bond auction takes place (at no particular day of the week), primary dealers have until 11 a.m. to bid, and own an option to buy more until the next day at 3:30 p.m., size also depending on their market share, at the average price. The additional amount issued via the option is 10% in the case of a tap or 25% in the case of a new bond. In Italy, the system is not purely competitive because primary dealers don’t pay the price they bid, but rather the lowest price (marginal price) bid by the group, which allows the issuer to fill his need. 

France counts 22 primary dealers while there are 24 in Italy. In France, dealers have to use the Banque de France IT system to bid while in Italy they need to know and use the Italian IT system of the Banca d’Italia. In terms of settlement, the two clearing and settlement institutions Euroclear and Cedel can generally be used to book transactions in the primary Euroland market, but national settlement systems in France, Italy, Spain and other countries continue to dominate in the government bond market. 

These are only a few examples of the many differences in auction procedures. Now think what the procedures of all the other countries might add to the confusion. All in all, national interests – to keep national techniques and national teams busy taking care of them – represent a barrier to market integration. This barrier was, en passant, removed from the private bond sector a long time ago.

Slow Progress
The contrast between the primary government bond market, run by sovereign entities, and the secondary market for sovereign bonds, run by the private sector, is striking. Long ago, the European financial industry standardised trading techniques with flourishing electronic platforms and concentration of trading activity with dealers tending to centralise their trading in one Euroland centre and with a single language, namely English, becoming the European bond market language. The liquidity, the transparency and the depth of the secondary Euroland sovereign bond market have indeed progressed tremendously since the introduction of the Euro and the following restructuring of the bond market trading practices. Due to that, end investors like PIMCO might even occasionally get better prices for government bonds than primary dealers, thanks to the electronic platforms and the enhanced competition. 

Progress in the primary government bond market has been very slow in comparison. Sovereign issuers have supported the positive trends described above by better managing the curve positioning of their debt and focusing their issuance on large benchmark-sized issues. They have also been more open to dialogue with investors like PIMCO, and have thus increased transparency. Innovation in terms of new bonds (inflation-linked bonds or new 50-year bonds for instance) has helped meet or has indeed sometimes anticipated new demand trends from investors. Some countries no longer require a physical presence through an office in the issuing country to become an official primary dealer. These and other improvements have been valuable, but more are needed to support the European government bond market at a time when investors look to Euroland for an alternative reserve currency.

PIMCO, together with other members of the AMTE, is spearheading efforts to promote more integration, more harmonisation and more efficiency within the Euroland government bond market. In a special project, which aims to propose and lobby for more improvements, my colleague Brian Tomlinson and I have brought together issuers, intermediaries and investors to join forces. This project, we believe, will be for the good of our markets and ultimately of our clients. We will keep you posted.  

Matthieu Louanges, CFA 
Executive Vice President