Speed read

Cum-ex transactions are complicated equity trading transactions which led to multiple claims for refund of a single payment of dividend withholding tax in Germany, Denmark and other European jurisdictions. The amounts refunded are reported to run to many billions of euros. A German court has recently found two former London-based bankers guilty of criminal conduct related to these transactions and required significant amounts to be repaid. The FCA has also confirmed that it is investigating conduct related to these transactions. Although the relevant transactions took place some years ago, the first wave of investigations is only now concluding, and it looks very likely that further investigation and litigation will follow.

On 19 March 2020, a German court found two former London-based bankers guilty in the first criminal trial related to 'cum-ex' trades. They received suspended sentences totalling 34 months, in light of their extensive cooperation with prosecutors. They and a Hamburg-based bank involved with the transactions are also reported to have been required to repay over €190m in illegally obtained earnings.

The case is notable for establishing for the first time that 'cum-ex' transactions were ineffective under prevailing German tax rules, and that this should have been so obvious to participants that the transactions in question amounted to criminal conduct.

What is cum-ex?

It is helpful to start with dividend arbitrage. In essence, dividend arbitrage trades rely on the fact that many jurisdictions, particularly in continental Europe, impose high levels of dividend withholding tax. However, they frequently reduce those rates for some shareholders, for example, pension funds in treaty territories. Dividend arbitrage transactions involve transferring the shares or the dividend rights over the dividend record date from a high-rate taxpayer to a low-rate taxpayer, with the shares usually transferred back shortly thereafter. The low-rate taxpayers will bear varying degrees of risk in relation to movements in the stock – in some cases limited to counterparty credit risk only. In many jurisdictions, the system works through withholding and repayment rather than relief at source. The low rate taxpayer will hold the share over the dividend date and will be able to reclaim a repayment of the withholding tax.

Shares can be transferred, and risks can be allocated, using an array of complex financial instruments including stock loans, forward sales, total return swaps and others. A feature of many of these instruments is that they involve the payment of dividend equivalents, or manufactured dividends, as they are sometimes called. Tax authorities have struggled with the appropriate way to tax manufactured dividends: witness, for example, the recent decision of the Court of Appeal, HMRC v Coal Staff Superannuation Scheme [2019] EWCA Civ 1610, that the UK's former approach to withholding tax on manufactured overseas dividends was contrary to the fundamental freedoms.

Generally speaking, most dividend arbitrage transactions have been accepted as sensible tax planning. Some jurisdictions, however, have challenged whether the temporary holder at the dividend date is actually the beneficial owner of the relevant dividend or have sought to challenge some transactions under anti-abuse rules.

The key feature of a so-called cum-ex transaction is that party A agrees to sell party B a share carrying the right to a dividend payment ('cum-div') but in fact it settles the transaction with stock which no longer carries the dividend entitlement ('ex-div'). B who receives the ex-div stock is therefore made whole by a manufactured dividend from A, which is paid after withholding so is reduced by the withholding amount. In these circumstances, B would rely on prevailing administrative practice to claim a refund of the withholding tax as if it had actually received dividends.

In the simple transaction described above, there is some logic in this: B has incurred the cost of the withholding (passed on to it by A). The difficulty here is that the express link between the actual withholding tax and the refund claim for that withholding tax has been broken – it is not necessarily the case that A owned the actual share, received the dividend and suffered the withholding. This opens up the possibility that the party that owned the actual cum-div share, who may well have had no knowledge of the transaction between A and B, could also have claimed a refund of the withholding tax. Potentially, you have multiple parties claiming a refund for a single payment of withholding tax.

A number of European tax authorities have identified that the amount they received in dividend withholding tax was actually less than they paid out by way of dividend refunds. From the perspective of those tax authorities, the position is very simple: the system is clearly not intended to work that way, and so they must have been the victims of fraud and should be reimbursed.

However, the situation is actually much more complicated. The question is on what basis do you recover, and from whom. In our example, B made the immediate refund claim, but B also suffered the economic cost of the withholding – it was in the same place as if it had held the share and received the dividend – so it is arguably unjust for B to suffer the cost. If the cost is imposed on B, B may think it is appropriate to try and recover from A and so on. Alternatively, the tax authority could allege that A, for example, is the mastermind of the transactions, and presumably had some financial benefit, though this can be very difficult to piece together from the pricing of the various transactions, and so it can recover direct from A, even though A has not made any tax filings in the jurisdiction.

Participants in a single transaction, in addition to parties A and B, are likely to include custodians for both parties. They may have issued the vouchers that allowed the refund claim. Other participants may include market participants who sold or lent the ex-div shares to A, and financial institutions or funds which provided liquidity for the trade or entered into hedging agreements with the various parties. Transactions would typically have been carried out using standard form documents and market financial terms (sometimes only completed to memorialise trades done over the phone), so the transactions are likely to have been fairly light in terms of documentation, and unlikely to have involved external advisers or credit or reputation committee or similar transaction-specific approvals within institutions. In many cases, transactions would have taken place ten or more years ago, and institutions will have had a significant turnover of staff in the interim. As a result, collecting and presenting evidence is a major challenge.

What will happen next?

It has been reported that some of the participants affected by the verdict of the German court are considering an appeal.

Many banks, brokers and professional services firms have been mentioned in press related to the trades. German prosecutors have said they are continuing investigations into over 50 cum-ex trades involving 400 suspects. This will be the first of many times when courts in Germany and across Europe pronounce on these issues. Criminal investigations have been ongoing for a number of years in several jurisdictions and continue to involve substantial amounts of cooperation and exchange of information between authorities. There is also already fairly extensive civil litigation, and there seems a good chance that this will increase. Authorities in many of the jurisdictions that are thought to have been affected have so far appeared to sit on the sidelines but this verdict may prompt them to action.

What are the implications in the UK?

Although the UK tax authorities are not impacted by the transactions (unlike many European jurisdictions, the UK does not impose dividend withholding tax), it is not a coincidence that the individuals in this case were based in London. Many financial institutions had teams engaged in equities trading, including dividend arbitrage, based in London at the relevant time. UK based financial institutions may also have been involved in transactions in ancillary roles, for example acting as custodians or prime brokers and/or providing liquidity or hedging for transactions.

As a result, UK based financial institutions have been a focus for investigating European tax authorities, including various proceedings to obtain information through the English courts. The Danish tax authorities are also seeking to recoup lost tax through the New York courts.

The regulatory dimension

The Financial Conduct Authority has also confirmed that it has been investigating abusive share trading in London's markets that allegedly supported these schemes. According to Mark Steward, the FCA's executive director of enforcement and market oversight, its investigations 'are now very close to their conclusion and decisions about action are imminent.' It is understood that 14 institutions and six individuals are currently under investigation.

Regulators will be looking at all facets of trade in equities around dividend dates. They will be guided by the exchange of information and intelligence from EU counterparts, and they will go back over many years to establish the extent to which firms were involved. They will be particularly interested in unusual trading patterns, trades with no apparent risk, off-market trades, trades with profitability out of line with the risk involved, the role of firms in promoting dividend arbitrage opportunities to clients, whether firms took care to understand the true nature and purpose of the transactions and whether there was adequate management oversight.

In the UK, at any rate, regulatory proceedings may carry more direct risk for institutions than criminal proceedings relating to fraud or tax evasion. The difficulty with ascribing criminal activity to a company, due to the requirement to attribute knowledge of wrongdoing to its controlling mind, typically its board of directors, is well documented and was a significant factor in the introduction of the corporate criminal offence of failure to prevent tax evasion.

Action points for financial institutions

Although the transactions in question mainly took place at least seven years ago, it appears that the fall-out is only just beginning. The overall amount of tax at stake has been estimated in a New York Times article at $60bn.

Key risk areas include: criminal law offences; anti-money laundering requirements; market abuse regulations; business and individual conduct rules and principles; whether adequate systems and controls are in place to counter the risk of furthering financial crime; civil proceedings for the recovery of WHT reclaims or profits derived from transactions relating to them; civil proceedings by current or former employees against institutions which profited from their activities; and reputational damage.

Financial institutions concerned about potential exposure to these risks should conduct internal investigations, so that they have the facts at their fingertips and are fully prepared to cooperate with regulators and others should the need arise.

When performing internal investigations, financial institutions should 'follow the money'. Of particular interest will be identifying where profits were generated and considering whether those were in line with expectations based on transaction risk or volumes. Trading data can be retrieved and interrogated to establish such profit outliers, as well as to identify trades displaying characteristics of arbitrage activity. For example, transfers of shares made prior to dividend due dates, which are quickly reversed; trades made cum-div which are settled ex-div; or share sales or loans that are accompanied by a payment, which could represent the manufactured dividends.

Analysis of structured trading and payment data should be supplemented with sources of unstructured data (such as instant messaging, email and voice recordings) which can be interrogated around the dates correlating to patterns of interest in the structured data. This information can provide the context and intent behind the trading activity, and it can sometimes unveil a proverbial smoking gun.

The control environment should also be considered as part of a robust investigation, including the approvals required to execute such trades and whether this process was commensurate with the risk. The contemporaneous oversight applied to trading portfolios and identification of anomalous activity should also be scrutinised. These actions together will assist an organisation to assess risk exposure and to establish, where appropriate, defensible positions across the firm's book of relevant business.

Originally published May 26, 2020.

The content of this article is intended to provide a general guide to the subject matter. Specialist advice should be sought about your specific circumstances.