Buy-in damage
10 March 2015
There鈥檚 a dark side to the CSDR that has remained unexplored鈥攗ntil now
Image: Shutterstock
It is a truth universally acknowledged that a central securities depository operating in the EU must be in want of a regulation. In July last year, the European Parliament and Council made this so, passing the Central Securities Depository Regulation (CSDR) into law, and now we are on the cusp of new legislation being added to this, due to come in to force in 2016.
Generally, the added legislation has been welcomed. It aims to improve efficiency and make trading through CSDs safer for all involved. In the midst of the legislative text, however, is a section entitled 鈥榮ettlement discipline鈥�, which is designed to reduce the number of failed trades, and is made up of three sections.
The first involves fairly uncontentious improvements to settlement instructions and trade matching information. The second is a little more controversial, imposing cash penalties on failing counterparties. But it is the final stipulation that is a cause for serious concern, and that is mandatory buy-ins.
There is nothing new about the concept of buy-ins. They already exist as a layer of additional protection for the buyer of securities. If one counterparty fails to deliver as agreed, the buyer has the right to appoint an agent to purchase those securities at the higher market price for guaranteed delivery.
The buyer then purchases the securities for the agreed price, and the seller is obliged to make up the difference.
Currently, this is purely on a discretionary basis, and occurs fairly infrequently. In fact, according to Andy Hill, a director in the market practice and regulatory policy team at the International Capital Market Association (ICMA), and researcher and author of ICMA鈥檚 impact study on CSDR mandatory buy-ins, an average large international bank might see three or four mandatory buy-ins per quarter.
鈥淚t鈥檚 quite an aggressive thing to do, and brings stresses to a relationship,鈥� he says.
鈥淎lso, if you鈥檙e an investor and you want an offer in a relatively illiquid security, that鈥檚 how you get a decent price and how you get exposure to the instrument you鈥檙e trying to buy. This therefore allows the market maker a little bit of flexibility, a little bit of comfort.鈥�
Under CSDR, a trade will be considered 鈥榝ailed鈥� if it has not delivered after seven days for illiquid securities, or after just four days for liquid securities. If it misses these deadlines, then a buy-in would be mandatory, and it wouldn鈥檛 just be conducted by a third-party agent, either.
Hill says: 鈥淵ou will be bought in, but it could be by the counterparty you failed to, or the settlement system you elected, the trading platform, or, in the case of a cleared trade, a central counter party.鈥�
鈥淭his complicates things. It not only puts a responsibility on the settlement system, which has been nominated, but also the trading platform, which has no knowledge of whether the trade fails or not.鈥�
CSDs will be burdened with more responsibility under the new rules, taking on mandatory tasks of collecting information, appointing agents and running consistency tests.
The European Central Securities Depository Association (ECSDA) has openly criticised this, calling the obligations impractical, and pointed out that the move is more likely to increase a CSD鈥檚 risk profile and create new liabilities. Like the ICMA report, it suggests that buy-ins should be handled on a trading level, not on a settlement level.
In its response to the European Securities and Markets Authority (ESMA) on the subject, ECSDA estimated the market-wide effect that mandatory buy-ins would have, putting the approximate number of buy-ins at 1.8 million per year. This would represent securities amounting to about 鈧�2.5 trillion, while the gross amount of late settlement penalties could reach 鈧�2.2 billion.
This would mean 7,500 buy-ins per day, compared to the current three or four per business, per quarter.
鈥淚t鈥檚 un-implementable, un-manageable. It cannot work,鈥� says Hill, but the figures provided in his report are perhaps even more shocking.
ICMA contacted sell-side traders, asking how the inconvenience and lack of flexibility involved in mandatory buy-ins would affect their pricing in real terms, including government, public and corporate bonds, both liquid and illiquid.
The lowest projected price increase was from public illiquid bonds, which could see a rise of 74 percent. The highest was corporate liquid bonds, with expected increases of 155 percent.
Public illiquid, sovereign illiquid and sovereign liquid bonds all predicted price increases of about 120 percent, and corporate illiquid bonds expected to increase costs by 99 percent.
The survey also questioned whether they would continue to offer securities not currently on their books. While no sovereign liquid bond traders said they would cease offers, 25 percent of sovereign illiquid bond traders said that they would. This is compared to 23 percent of public liquid bond traders and 33 percent of those trading public illiquid bonds, to 20 percent of corporate liquid bond traders, and 40 percent of corporate illiquid bond traders.
Calculating the actual costs of these buy-ins is tricky, given the lack of solid data for the European bonds market, and without knowing the scale of the impact on market players鈥攚hich companies will withdraw or scale back offerings, and which clients will pull out after a price hike.
Hill says: 鈥淲e wanted to be a little bit careful about pinning down an exact a number, so we worked out the cost per trillion euro of volume. Based on this, it will cost almost 鈧�1.4 billion per trillion.鈥�
Hill came to an estimated cost using market data from Trax. It assumed the split of secondary volumes across sovereign, public and corporate bonds would remain stable, and that the ratio of liquid to illiquid bonds would be determined by the Markets in Financial Instruments Directive (MiFID) II.
The numbers were estimates, and so the report worked to the lower possibilities. Even so, Trax puts the annual market volume of 2014 at about 鈧�24 trillion, which would have incurred costs of about 鈧�33.6 billion.
Hill pointed out in the report: 鈥淓ven if the true number is a fraction of this, the resulting annual cost will still run in to several billions.鈥�
On top of this, a somewhat accidental knock-on effect of the proposed regulation is the costs to the repo market.
Buyers of repo trades also have the right to affect a buy-in style transaction. Under CSDR, these will become compulsory at both the start and end stages of a repo trade, 鈥渨here practical鈥�.
鈥楶ractical鈥� is expected to mean within 14 days, meaning a 14-day grace period between the initial fail and the buy-in taking effect. Based on the ICMA repo study data from December 2014, Hill made a conservative estimate of what the costs to the repo market would be.
The analysis was based on a current market size of $5.8 trillion, applied only to sovereign bonds, and assumed that all underlying repo assets were liquid. These factors are more likely to under-estimate the total costs than exaggerate them, and, if nothing else, it is based on an original survey that is thought to capture only 80 percent of the market.
Even taking this in to account, the estimated market cost is more than 鈧�3 billion per year.
According to Hill, what is most worrying about these costs is that they hit the end user hardest.
鈥淭hese are not costs to the banking sector, these are costs to investors. This is the additional spread they will have to pay,鈥� he says. 鈥淭he figures also don鈥檛 fully capture the withdrawal of liquidity, and you can鈥檛 really put a price on that.鈥�
Godfried De Vidts, chair of the ICMA鈥檚 European Repo Council, supports the findings of the report, calling for it to be presented to the European Commission and parliaments. He references the MiFID pre-and post-trade regulations, which are expected to cost the commodities sector about 鈧�4 billion, or 鈧�20 per household in Europe.
He says: 鈥淲e are talking much bigger numbers here. Can you really justify to the taxpayer that they should pay more and get a less safe system? Nobody would actively object if the regulation made the system much more robust, but here it just unwinds the work that we have done over the last 20 or 25 years.鈥�
His advice to ESMA is to slow down and wait to see what effect Target2-Securities (T2S) will have on harmonising the market.
鈥淎 lot of fails are settled the next morning, because deadlines are not coordinated in Europe. Sometimes things get stuck in the piping, and it鈥檚 only when you give it a good flush that everything goes through, and that鈥檚 T2S. If CSDR comes in before that, it will be a disaster.鈥�
ECSDA also supported a delay in the implementation of CSDR because of the impact of T2S, suggesting an extension of the phase-in from 18 months to 24 months.
De Vidts goes further though, saying ESMA should consider making amendments to the level-one legislation.
鈥淥ur recommendation would be to wait to implement this until they know what the impact of T2S is going to be, and to see the improvements. Then decide if we need mandatory buy-ins or not, because right now this is a solution looking for a problem.鈥�
鈥淓SMA can change the rules. It has happened before, it can be done.鈥�
Regulations are there to protect the industry and the end user, and one little clause can be enough to lead to damaging effects for consumers who are largely unaware of any change.
鈥淭his is creating unsafe markets, instead of making them safer,鈥� says De Vidts. 鈥淚t鈥檚 a cost to you and me, to the investor or the pension holder.鈥�
More than a mere irritation, regulations must be studied to unearth the real, and often unintended, consequences.
鈥淣obody believed that CSDR would really achieve settlement efficiency,鈥� Hill concludes. 鈥淏ut this is the first attempt to quantify the impact, and it鈥檚 actually quite shocking.鈥�
Generally, the added legislation has been welcomed. It aims to improve efficiency and make trading through CSDs safer for all involved. In the midst of the legislative text, however, is a section entitled 鈥榮ettlement discipline鈥�, which is designed to reduce the number of failed trades, and is made up of three sections.
The first involves fairly uncontentious improvements to settlement instructions and trade matching information. The second is a little more controversial, imposing cash penalties on failing counterparties. But it is the final stipulation that is a cause for serious concern, and that is mandatory buy-ins.
There is nothing new about the concept of buy-ins. They already exist as a layer of additional protection for the buyer of securities. If one counterparty fails to deliver as agreed, the buyer has the right to appoint an agent to purchase those securities at the higher market price for guaranteed delivery.
The buyer then purchases the securities for the agreed price, and the seller is obliged to make up the difference.
Currently, this is purely on a discretionary basis, and occurs fairly infrequently. In fact, according to Andy Hill, a director in the market practice and regulatory policy team at the International Capital Market Association (ICMA), and researcher and author of ICMA鈥檚 impact study on CSDR mandatory buy-ins, an average large international bank might see three or four mandatory buy-ins per quarter.
鈥淚t鈥檚 quite an aggressive thing to do, and brings stresses to a relationship,鈥� he says.
鈥淎lso, if you鈥檙e an investor and you want an offer in a relatively illiquid security, that鈥檚 how you get a decent price and how you get exposure to the instrument you鈥檙e trying to buy. This therefore allows the market maker a little bit of flexibility, a little bit of comfort.鈥�
Under CSDR, a trade will be considered 鈥榝ailed鈥� if it has not delivered after seven days for illiquid securities, or after just four days for liquid securities. If it misses these deadlines, then a buy-in would be mandatory, and it wouldn鈥檛 just be conducted by a third-party agent, either.
Hill says: 鈥淵ou will be bought in, but it could be by the counterparty you failed to, or the settlement system you elected, the trading platform, or, in the case of a cleared trade, a central counter party.鈥�
鈥淭his complicates things. It not only puts a responsibility on the settlement system, which has been nominated, but also the trading platform, which has no knowledge of whether the trade fails or not.鈥�
CSDs will be burdened with more responsibility under the new rules, taking on mandatory tasks of collecting information, appointing agents and running consistency tests.
The European Central Securities Depository Association (ECSDA) has openly criticised this, calling the obligations impractical, and pointed out that the move is more likely to increase a CSD鈥檚 risk profile and create new liabilities. Like the ICMA report, it suggests that buy-ins should be handled on a trading level, not on a settlement level.
In its response to the European Securities and Markets Authority (ESMA) on the subject, ECSDA estimated the market-wide effect that mandatory buy-ins would have, putting the approximate number of buy-ins at 1.8 million per year. This would represent securities amounting to about 鈧�2.5 trillion, while the gross amount of late settlement penalties could reach 鈧�2.2 billion.
This would mean 7,500 buy-ins per day, compared to the current three or four per business, per quarter.
鈥淚t鈥檚 un-implementable, un-manageable. It cannot work,鈥� says Hill, but the figures provided in his report are perhaps even more shocking.
ICMA contacted sell-side traders, asking how the inconvenience and lack of flexibility involved in mandatory buy-ins would affect their pricing in real terms, including government, public and corporate bonds, both liquid and illiquid.
The lowest projected price increase was from public illiquid bonds, which could see a rise of 74 percent. The highest was corporate liquid bonds, with expected increases of 155 percent.
Public illiquid, sovereign illiquid and sovereign liquid bonds all predicted price increases of about 120 percent, and corporate illiquid bonds expected to increase costs by 99 percent.
The survey also questioned whether they would continue to offer securities not currently on their books. While no sovereign liquid bond traders said they would cease offers, 25 percent of sovereign illiquid bond traders said that they would. This is compared to 23 percent of public liquid bond traders and 33 percent of those trading public illiquid bonds, to 20 percent of corporate liquid bond traders, and 40 percent of corporate illiquid bond traders.
Calculating the actual costs of these buy-ins is tricky, given the lack of solid data for the European bonds market, and without knowing the scale of the impact on market players鈥攚hich companies will withdraw or scale back offerings, and which clients will pull out after a price hike.
Hill says: 鈥淲e wanted to be a little bit careful about pinning down an exact a number, so we worked out the cost per trillion euro of volume. Based on this, it will cost almost 鈧�1.4 billion per trillion.鈥�
Hill came to an estimated cost using market data from Trax. It assumed the split of secondary volumes across sovereign, public and corporate bonds would remain stable, and that the ratio of liquid to illiquid bonds would be determined by the Markets in Financial Instruments Directive (MiFID) II.
The numbers were estimates, and so the report worked to the lower possibilities. Even so, Trax puts the annual market volume of 2014 at about 鈧�24 trillion, which would have incurred costs of about 鈧�33.6 billion.
Hill pointed out in the report: 鈥淓ven if the true number is a fraction of this, the resulting annual cost will still run in to several billions.鈥�
On top of this, a somewhat accidental knock-on effect of the proposed regulation is the costs to the repo market.
Buyers of repo trades also have the right to affect a buy-in style transaction. Under CSDR, these will become compulsory at both the start and end stages of a repo trade, 鈥渨here practical鈥�.
鈥楶ractical鈥� is expected to mean within 14 days, meaning a 14-day grace period between the initial fail and the buy-in taking effect. Based on the ICMA repo study data from December 2014, Hill made a conservative estimate of what the costs to the repo market would be.
The analysis was based on a current market size of $5.8 trillion, applied only to sovereign bonds, and assumed that all underlying repo assets were liquid. These factors are more likely to under-estimate the total costs than exaggerate them, and, if nothing else, it is based on an original survey that is thought to capture only 80 percent of the market.
Even taking this in to account, the estimated market cost is more than 鈧�3 billion per year.
According to Hill, what is most worrying about these costs is that they hit the end user hardest.
鈥淭hese are not costs to the banking sector, these are costs to investors. This is the additional spread they will have to pay,鈥� he says. 鈥淭he figures also don鈥檛 fully capture the withdrawal of liquidity, and you can鈥檛 really put a price on that.鈥�
Godfried De Vidts, chair of the ICMA鈥檚 European Repo Council, supports the findings of the report, calling for it to be presented to the European Commission and parliaments. He references the MiFID pre-and post-trade regulations, which are expected to cost the commodities sector about 鈧�4 billion, or 鈧�20 per household in Europe.
He says: 鈥淲e are talking much bigger numbers here. Can you really justify to the taxpayer that they should pay more and get a less safe system? Nobody would actively object if the regulation made the system much more robust, but here it just unwinds the work that we have done over the last 20 or 25 years.鈥�
His advice to ESMA is to slow down and wait to see what effect Target2-Securities (T2S) will have on harmonising the market.
鈥淎 lot of fails are settled the next morning, because deadlines are not coordinated in Europe. Sometimes things get stuck in the piping, and it鈥檚 only when you give it a good flush that everything goes through, and that鈥檚 T2S. If CSDR comes in before that, it will be a disaster.鈥�
ECSDA also supported a delay in the implementation of CSDR because of the impact of T2S, suggesting an extension of the phase-in from 18 months to 24 months.
De Vidts goes further though, saying ESMA should consider making amendments to the level-one legislation.
鈥淥ur recommendation would be to wait to implement this until they know what the impact of T2S is going to be, and to see the improvements. Then decide if we need mandatory buy-ins or not, because right now this is a solution looking for a problem.鈥�
鈥淓SMA can change the rules. It has happened before, it can be done.鈥�
Regulations are there to protect the industry and the end user, and one little clause can be enough to lead to damaging effects for consumers who are largely unaware of any change.
鈥淭his is creating unsafe markets, instead of making them safer,鈥� says De Vidts. 鈥淚t鈥檚 a cost to you and me, to the investor or the pension holder.鈥�
More than a mere irritation, regulations must be studied to unearth the real, and often unintended, consequences.
鈥淣obody believed that CSDR would really achieve settlement efficiency,鈥� Hill concludes. 鈥淏ut this is the first attempt to quantify the impact, and it鈥檚 actually quite shocking.鈥�
NO FEE, NO RISK
100% ON RETURNS If you invest in only one securities finance news source this year, make sure it is your free subscription to Securities 麻豆影视传媒 Times
100% ON RETURNS If you invest in only one securities finance news source this year, make sure it is your free subscription to Securities 麻豆影视传媒 Times
