**Complex Contracts and Combinations: Compounding Complications**. Before going into a discussion of some of policy considerations behind Treasury’s treatment of complex contracts, we should discuss some additional complications that result when paired with the concept of combined contracts.

As discussed previously, in certain cases, two or more derivatives that would not otherwise be treated as subject to withholding under the new regulations can be treated as a single instrument and tested as such under the combination rules. However, before jumping into issues created by the combination rules and their interaction with complex contracts, one question should be asked: did Treasury intend for combined contracts to be tested under the substantial equivalence test?

A strict textual reading is slightly ambiguous. While the combination rules generally refer to testing combined instruments as single instruments without excluding the potential application of the substantial equivalence test, the ordering rule requires combining transactions in a manner that results in the most transactions with a delta of 0.80 or higher with respect to the underlying security. This ordering rule sort of confuses things since complex contracts are not treated as transactions with a delta of 0.80 or higher; rather, they are simply treated as section 871(m) transactions. Consequently, any combination that fails solely as a result of the application of the substantial equivalence test would theoretically lead to 0 transactions with a delta of 0.80 or higher.

It’s difficult to believe that Treasury meant to preclude combinations that resulted in complex transactions from the substantial equivalence test, however, as this would leave a fairly large hole in the regulations and, by all appearances, Treasury is trying to cover up every hole they possibly can. Therefore, for purposes of the following discussion, we will assume that combinations can be tested under the substantial equivalence test and that the ordering rule language will be clarified either to support this or to render much of the following irrelevant.

The foregoing ambiguity aside, the combination rule creates two similar issues: (1) combinations of two or more simple contracts to create a complex contract and (2) combinations of complex contracts with other derivatives.

I raised the first issue in an earlier post. In that post, I mentioned how, in certain instances, the regulations require contracts entered into “in connection with” each other to be treated as a single contract. Some combinations will be relatively straightforward. For example, a put and call with the same maturity and the same strike entered at the same time.

However, this gets exceedingly complicated once one enters into puts and calls with overlapping strike prices and with different maturity dates. While the financial engineers that deal with exotic products may be able to implement the substantial equivalence test, it seems very unlikely that those who deal with fairly plain vanilla options will be ready to deal with the math required in the regulations. Of course, one could simply build a system that can run this data, but the math gets really complicated, really fast. For example, let’s say you have four options entered into in connection with each other: A, B, C, and D. This creates the following possible combinations:

- A,B
- A,C
- A,D
- B,C
- B,D
- C,D
- A,B,C
- A,C,D
- B,C,D
- A,B,D
- A,B,C,D

Each of these could potentially fail the substantial equivalence test and, therefore, must be tested. And this is just 4 options! For 10 options, there are over 1,000 combinations. For 20 options, there are over 1,000,000. Fun times!

*Perhaps *the ordering rule could reduce some of these calculations assuming the test can be performed in an iterative fashion rather than merely by brute force. In any event, even if the calculation can be performed there are further ambiguities because it is possible that combinations with the same number of contracts (such as (7)-(10) above) could all potentially fail the substantial equivalence test. In that case, the ordering rule doesn’t make clear which one would be regarded as the “bad” combination.

Perhaps the ordering rule could be re-stated to provide that the combination that results in the highest “initial hedge” should be chosen as the offender, but even in that case, there could be a tie! While it may make little substantive difference, it may create some practical difficulties as withholding agents must identify and track the correct individual lots that are subject to withholding.

Lastly, these calculations could get exceedingly complicated if the derivative references multiple shares of stock and the derivative doesn’t otherwise qualify for the index exception. These calculations could spawn multiple complex contracts from multiple different combinations.

The second issue relating to combinations and complex contracts is simply a variation on the first, but it just kicks the difficulty up a notch. This difficulty results when the complex contract is combined with a simple contract(s) or other complex contract(s). In those cases, the math becomes exceeding complex since a complex contract is not simply the result of a combination, but, rather, one of the addends. Perhaps more importantly, because complex contracts could take the form of publicly traded instruments, it’s possible that an investor is holding the instrument in a non-issuer brokerage account. While the issuer would likely know the math underlying its own issuance, a non-issuing broker may have no clue how the math works for the issuer’s complex contract. The non-issuing broker would need to reverse engineer the math and the assumptions underlying the instrument in order to perform the calculations in order to * begin *the combination calculus, and it’s unlikely the non-issuing broker’s math will match the issuer’s math.

Of course, a non-issuing broker may rely on certain presumptions that protect them from having to do the math relating to combinations involving complex contracts (e.g., no “knowledge” of in connection with), but the investor does not have these protections and, in any event, the investor may simply inform the broker of the “in connection with” combinations in order to force the obligation onto the withholding agent.

Speaking of investors, I don’t really expect them to do any of the math behind complex contracts (not that they could anyway). I expect issuers will (except in rare instances) design instruments that will not, by themselves, fail the substantial equivalence test. Such issuances will likely contain assurances that the instrument is not subject to section 871(m) but the disclosures may also contain some pretty chilling caveats about the potential for combinations and, consequently, the potential for withholding to apply in such instances.

**Does this make sense?** Given the foregoing as well as the complications discussed in the last post, does it make sense for Treasury to finalize this complicated test? In my personal opinion, I find strange Treasury’s willingness to use concepts such as “standard deviation” and “probability”. It seems to me that they are looking for a fairly precise tool that can equate the treatment of simple contracts and complex contracts; tools that would be familiar to financial engineers and not to tax lawyers, tax accountants, and certainly not layman. In fact, some of the ambiguity in the regulations suggests Treasury’s own unfamiliarity with the concepts that they are using.

And, of course, even if we assume that Treasury has a good understanding of these statistical/financial concepts, what of the internal/external tax advisors? I could see many understanding this on an intuitive level but not on a real mechanical level. There couldn’t be more than a handful that understand the real math that’s going on and I’d even be surprised if it was a handful. Both internal/external advisors may be limited to running smell tests on these types of calculations and may be unable to consider whether they truly comply with the regulations or not.

And what of an IRS audit of this issue? I have no clue what that will look like or how the IRS could possibly wrap their heads around the formulas and assumptions that financial engineers will be using. Ultimately, it may devolve into a review of the process surrounding the calculation but I’d be surprised if the IRS could satisfactorily audit these calculations without having a math expert review sampled calculations and challenge the financial engineers underlying data. I can’t imagine the IRS would have the resources to audit this properly.

In any event, it appears that Treasury is aware of the above issues as they’ve requested comments regarding this test, including its administrability.

The securities industry also seems aware of this issue as they commented on these regulations in a letter sent to Treasury on October 27, 2015 of last year:

*… the securities industry must determine how to apply the novel and complex substantial equivalence test to a broad range of financial instruments. While the concept underlying this test originated from within the securities industry, many market participants are unfamiliar with it, and the securities industry in general is not sure how it applies to many instruments. The government is still seeking comments on the test, and further guidance may be issued with respect to the test that cannot be anticipated as of the early effective date.*

While I don’t doubt the sincerity of the comment, I do find it amusing as the securities industry does admit that the test originated from within the securities industry. It’s like asking for a puppy on Christmas, getting it, and then complaining you don’t know how to take care of it.

__Other Ways to Approach Complex Contracts__*. *It seems that Treasury is trying to deal with complex contracts by coming up with an equivalence test that makes a “fairly” tight fit with the delta 0.80 concept. However, Treasury could potentially take two different approaches that are less complex but don’t provide as tight a fit by either narrowing or broadening the application of the withholding rules as they would apply to complex contracts. In other words, try not be so precise!

*Narrowing the rules. *The narrowest rule would be to simply rule out the application of Section 871(m) to complex contracts. This would obviously eliminate the need for a substantial equivalence test and greatly simplify combinations. The underlying notion here is that investors don’t enter into complex contracts in order to avoid withholding because the premiums associated with exotic instruments (such as high bid-ask spreads) make it too inefficient for that goal.

However, this would likely require some form of backstop since avoiding the withholding rules may become all too simple. An investor would simply have to make their instrument (or combinations) slightly more complicated in order for it to fall within any newly minted “complex contract” exception.

Perhaps one such backstop could be an intent-based test such as the application of the anti-abuse rule. While it can be difficult to divine the intent of the taxpayer, there may be certain obvious cases where even a dolt like me can see that an instrument is virtually replicating the stock or a high delta derivative and the intent is obvious on the face of it. If the case is not obvious, one could perhaps live with such complicated instruments escaping withholding as they are not likely done with a tax avoidance purpose in the first place.

*Broadening the rules*. The opposite approach would be to broaden the rules as they apply to complex contracts. There will obviously be some crying from the issuers and investors. But who buys these complicated instruments anyway? A bunch of fat cats who can afford them? And why are we protecting them anyway? Am I right, folks?

What form could these rules take if they were broader? Well, you could simply *flip* around one of the comments made to Treasury for dealing with complex contracts. The Treasury preamble noted, “one comment suggested that the delta should be calculated using the highest possible number of shares that could be referenced by the derivative at maturity.” Treasury could simply flip that around and calculate delta using the *least* possible number of shares that could be referenced by the derivative at maturity. I don’t know if this test could be performed on every possible complex contract but maybe it’s a good place to start? Of course,this should broaden the number of complex contracts that will be section 871(m) transactions but if you’re in the “who cares” camp, then you don’t care.

Of course, perhaps the easiest way to broaden the treatment of complex contracts is to simply make them all subject to withholding at the relevant delta on the underlying dividend date regardless of their economics on the issuance date. This obviously treats complex contracts much more severely than their simple counterparts. But maybe the real question is why aren’t all instruments subject to withholding at the relevant delta on the dividend date. Treasury seems to be interested in the true economics of these instruments and wouldn’t this actually be the true economics of the implied dividend? If this were the rule, these regulations could be much easier to implement as both the combination rule and the special rules for complex contracts would be unnecessary.

*Re-think Dividend Withholding.* Of course, the last point above may lead to the broader question of why are we imposing withholding tax on dividends in the first place? I’ll discuss this in a later post.

Pingback: Section 871(m) and Complex Contracts, Part I | Tax Pro Super Happy Fun Time - A Tax Blog

Pingback: US Securities Market Coalition Commentary on Section 871(m) and Complex Contracts | Tax Pro Super Happy Fun Time - A Tax Blog

Pingback: US Source Taxation of Dividends | Tax Pro Super Happy Fun Time - A Tax Blog