Wednesday, August 11, 2021
Author: Ed Blount
The German Federal Court of Justice's decision two weeks ago to prosecute as criminals anyone who abused dividend arbitrage trades anytime over the previous 25 years is bad news for everyone in the securities lending community. The German tax authorities' new determination to conduct sweeps of securities loans that span dividend record dates should in particular sound the alarm for institutional securities lenders, especially if it presages a new trend among regulators.
Germany is not the only country looking to probe securities finance markets to increase tax penalties and help defray its covid costs. However, penalties won't be easy to collect because the standard method of 'auditing for purpose' in securities finance, i.e., by sampling record-date loans on high-yielding stocks, is more likely to result in increased compliance costs and distractions for legitimate lenders, than in actual disgorgement of unjust riches from the tax criminals.
The vast majority of record-date loans have a legitimate economic purpose, e.g., hedging, scrip arbitrage, et al, but it's next to impossible to prove. The lending chain enters a tunnel when the lender's agent satisfies the prime broker's needs with a new stock loan. So how does this question get answered:
Is the ultimate borrower, the one at the edge of the trade, using the shares for benign or for criminal purposes?
That's hard to answer because edge borrowers are hard to find. If the edge borrower's intention and purpose are important in an audit, then cross-border lenders will be hard-pressed to prove the tax legitimacy of their loans. Consider that the German supreme court ruled that lending chains created for illegal purposes will not be considered to have transferred beneficial ownership for tax purposes.
Title reversion is going to be a big problem with these loans.
A few years ago, dividend arbitrage trades were considered an acceptable evolution of the original, non-resident / tax accommodation loan set up by global custodians. The foreigner-to-native loan and back was intended to save foreign non-profits the trouble of waiting for repayment of treaty-exempt dividend taxes that had been withheld at the sub-custodian level. Fewer reclaims also meant less of a processing burden for harried government workers. However, that accommodation loan morphed over time into one leg of a arbitrage trade that played one issuer's national tax policy against another's.
Shrewd but compliant from a legal standpoint, as everyone agreed. But eventually, the sharpies pushed the trade too far and then came the abuses. Multiple borrows for a complex set of trades resulted in postings of phantom shares with multiple reclaims being filed for taxes that had never been withheld.
Complexity is no excuse with the EU's auditors.
The German Finance Ministry takes the position that every suspicious structure is potentially invalid and abusive. Last September's ESMA report listed several of those structures, all booked across record date. So, lenders must ask, does the court's ruling mean that foreign non-profits with German dividend-paying stocks should unwind all loans before the record-date? Seems the safest route, unless lenders and borrowers can work together to validate those loans for tax purposes.
The broker's cooperation will be a necessity for validating legitimate loans.
In securities finance, the borrowers' brokers must abide by the U.S. Federal Reserve's Regulation T credit rules. That means all loans must pass a purpose test. But the broker's purpose could involve customer segregation rules, delivery fails, or the short-sale settlement needs of their hedge funds and in-house traders. Without help from brokers, lenders will be unable to answer for the use of their lent shares if subpoenaed in a criminal tax audit. Even the brokers may be unable to explain all their customers' trades.
That's another big problem.
Germany's new guidance on cum-ex and cum-cum trades and the Ministry's interpretation sets up a complicated calculus for how dividend tax claims are handled based on whether beneficial ownership in the underlying shares is transferred from the lender to the borrower.
How can the lender (or its custodian) know whether the title has indeed transferred, if the tax exemption criteria are based on the borrower's behavior? Even worse for borrowers, there is virtually no statute of limitations.
And that's still another problem for lenders.
If the borrower's WHT reclaims are disallowed years later because beneficial ownership was never transferred, i.e., title reversion on the loan, global custodians and accountants for U.S. lenders can suddenly find out that their client's manufactured dividends are deemed to be direct payments from foreign issuers. Now correct the accounting.
That's a lot of problems, as we predicted when ESMA published their cum-ex report last October. But there is a better way to prove that the purpose of the loan is benign: use shared ledger technologies to map the loan end-to-end. Show that the borrower had a legitimate purpose for the shares.
Just my opinion ...