Ok, so by now I am somewhat known for blog titles that are a play on words. I admit to struggling with Singapore, not for lack of a suitable word play (everyone surely knows what a Singapore Sling is), but more for how to connect that to the subject matter of this post. But do bear with me – I’ll get to it!

While I’m travelling the world obviously continues to go round. In my rather prosaic world of cross border taxation, changes continue to occur that will affect investors in major ways. Contrary to popular belief, the industry is not getting easier or more automated. It seems that for every step forward in automation and standardisation, there are steps back in making things more complex.

I’ve hear, for example, that Germany is looking to implement an anti tax arbitrage strategy of requiring evidence with tax reclaims of holding for at least 45 days post pay date. If true, this would mean new elements to consider. First, claims could not even be started until 45 days after the pay date in order to create the evidence trail. Second, what evidence will be required? Holding on pay date is evidenced with a tax voucher. Simply evidencing a holding with no associated corporate action will require work from a custodian unless a statement is going to be acceptable. Several tax authorities already require evidence of holding prior to pay date – usually six months – to show that the security was held for a bona fide investment purpose. If we join the pre pay date and post pay date requirement, we will end up with a ‘holding block’ of around seven months. While this might be a simple empirical way to demonstrate the applicability of the treaties and deter those seeking to abuse them, it may also have some unintended consequences in terms of inbound investment.

Equally, I read a recent PwC report that notes that the US is starting to reject 1040-NR claims due to difficulties at the IRS in connecting a claim to the actual tax paid at the top of the chain. This actually just highlights some of the inefficiencies of the US reporting system. It’s already extremely complex for a foreign financial firm to even understand these reporting requirements (I am talking here about IRC Chapter 3 not FATCA). There are problems in the US cascade reporting model, even if everyone in the chain does precisely the right thing. When, as often happens, someone in the chain misses out just one component data element on their reporting, the whole system fails. The IRS issues penalty notices to financial firms down the chain even when the failure was further up the chain. The mitigation process for these incorrectly issued penalties usually lasts months and can cause significant stress on employees.

When you delve behind the headlines, it’s clear that, apart from any disconnects in the reporting, the US simply does not want non-qualified intermediaries out there. My logic is simple. The issue is being drawn to our attention with respect to individuals filing long formclaims to the IRS (the 1040-NR is the claim form for individuals, it’s an 1120-F for corporates). The problem being that the 1042-S they are given doesn’t connect up the chain. However, the only individual beneficial owners who would get a 1042-S would receive it from an NQI. A QI does not need to issue a downstream 1042-S to an individual, as it pools its reporting of direct beneficial owners upstream. So, the scenario painted would be almost exclusively for clients of an NQI. If a QI had such clients that were over-withheld for some reason, there are already three methods within the 1441 NRA regulations that allow the QI to refund its client without the need for a 1040-NR, of which the most common would be the ‘collective refund’. So, while NQIs are not explicitly mentioned, they are the clear targets.

The other unmentioned problem with the US long form claims process is the unintended consequence. If the IRS accepted such claims, as they have in the past, it would and should trigger a completely new amended reporting cycle. Consider that the income reported by the whole chain for a given beneficial owner at the bottom of the chain below an NQI was at 30%. Now the BO files a 1040-NR, and let’s say the IRS accepts and pays the claim. This means that that proportion of income previously reported by everyone at 30% is now incorrect and the report should show this as, say, 15%. For the system to work properly, the IRS would have to require that the US withholding agent issue an amended 1042-S, which would then cascade down the system. Clearly this would cause havoc because of its entirely random nature.

Finally, key people in the industry are starting to have to come to grips with US tax regulations 871(m) and 305(c). 871(m) deals with substitute dividends. When read in some detail, the (hopefully unintended) effect could be to make a whole swathe of the bonds industry disappear through an understandable reluctance to make bonds subject to 871(m) eligible in the clearing and settlement systems. In essence, if any part of a bond (particularly for this discussion those that are not directly related to US securities) has an indirect element that even merely refers to an underlying US security – e.g. to calculate its interest rate – then when the referred US security pays a dividend, the holder of the non-US bond will be deemed to have received a substitute US dividend and be subject to US tax. How this would work in practice is difficult to see, once one gets past the seeming ‘over-reach’ involved. If the bond has not yet actually paid anything, but the underlying linked security has, there would have to be an accrual account (more cost) to keep track of the tax. 871(m) provides for such a mechanism – i.e. there’s a taxliability separate from the tax payment, which can be satisfied later. Once a payment is received, even though there will probably be no withholding tax due on the interest payment itself, US tax would have to be deducted to offset the accrual. If no payment is ever received, the holder would presumably have to sell something in order to meet the US tax liability. Not easy or simple issues to resolve, which is perhaps why 871(m) continues to be pushed back, but perhaps indicative of a concerning outreach by governments to extract tax from pretty much anything.

So, while the industry may be making strides towards automation and standardisation, I don’t see the same thing at tax authorities. I need that Singapore Sling!

Image Credit: Khairul Nizam