Search This Blog

Friday, August 22, 2008

What're the Odds?

Did you notice how Shawn Johnson seemed genuinely happy for Nastia Liukin after the latter's balance-beam routine, even though it might have cost Shawn the gold? That was really nice - I thought it was great that Shawn finally got her gold for that difficult and amazing routine (Nastia, another apparently very nice person (name notwithstanding), had hers already).

Still on the Olympics, I was reflecting further on the whole Phelps thing, and the incredible odds against his really having gotten the Amazing Eight, given those all-but-impossible butterfly and relay finishes. I guess in so many cases it's true that the odds that any particular occurrence would actually take place in this world are maybe a billion to one or more, and yet . . . each and every one of them in fact happens. There's nothing like sports to remind us of things like that.

And, speaking of odds, what're the odds that "they" would give away relief under the 409A regulations not expressly set forth in the regulatory language? ({moan/groan} - I know; these segues grow increasingly painful.) Well, there seems to be one place where they've done so. Under what have come to be known as the anti-"toggling" rules (of Section 1.409A-3(c)), you can generally (unless an exception applies) only have one time and form of payment per particular type of cap-A triggering event.

There may be some emerging informal evidence (rely at your own risk) that, showing some flexibility, Treasury personnel are gravitating to what I'll call a permissible "subset" analysis. To wit, let's say a plan has a provision that provides that the service provider gets paid on date X, or earlier in the event of an involuntary termination. Is that (i) an impermissible toggle (as, frankly, I would've thought) because you've now ultimately got two different payout times depending on the details of the separation, or (ii) a permissible payment date applicable to one type of trigger (separation from service, in my example), albeit a subset thereof, and another permissible payment date applicable to a different type of payment trigger (definite time, in my example)?

The idea is apparently that you're not impermissibly toggling within a trigger type, but rather providing for different rules for two different triggers, even if only for a subset of one of the triggers. Note that the analysis could also help, for example, with accelerations for some but not all "disability" distributions. (Note also the express ability to use subset treatment in the CiC context for toggling within that trigger type, at -3(i)(5)(i).)

This approach would confer helpful flexibility. Indeed, I always thought the regulations way overshot the mark on toggling. The anti-toggling rule expressly scares up and prohibits the possibility that you may write a provision that provides for one type of distribution in the case of Monday separations and another in the cases of all other separations. In effect, all distinctions within a trigger are equated with that extreme theoretical possibility. I would have preferred a regulatory approach under which you could have provided for different times and forms of payment for subsets of triggering types, so long as the distinctions being drawn were for bona fide business purposes, and were not intended to serve as a device for avoiding cap-A constraints. Well, that just didn't happen; but now we have the possibility of some relief in this more flexible approach to subsets where two different types of triggers are involved.

So how far can you take this possible largess? Maybe some will be focusing on how to draft the "involuntary" concepts and on other technical matters. But can the whole thing be taken quite a bit further? Consider: "Payment shall be made on December 31, 2010, or, if earlier, on termination of employment occurring on a Tuesday." Ooh. (I've avoided using Monday, since the regulations seem to be quite upset with Mondays.)

On the one hand, maybe you'll say that the provision I suggest effectively amounts to a putative distribution election. But hasn't the cow left the barn as to this type of reasoning? As noted, the regulations have essentially lumped all intra-trigger distinctions together as being potentially abusive, and solved the problem of potential abuse by universally prohibiting toggling (unless an express exception applies). With that backdrop, it would arguably make no sense to try to rank one type of distinction as being worse than another - they're all equally prohibited. And so, if a flexible analysis in the case of two different triggers results in concluding that acceleration for a subset of a trigger is permissible, then, since the regulations effectively view all distinctions within a trigger as comparable, it is arguably the case that there should be no further rating or ranking of various different possible distinctions. Put another way - if using a subset analysis for involuntary terminations works, then the use of a subset analysis for Tuesday terminations should work, too.

I'm not necessarily suggesting that anyone is really going to push the analysis quite so far. Having said that, though, there may be more nuanced attempts to explore the use of subsets in creative ways. Maybe it's somewhat ironic, but the regulations' use of extensive and complex rules to address potential workarounds may wind up encouraging exploration of just the kind of workarounds with which the regulatory draftspeople seem to have become concerned.


Anonymous said...

I think the difference is that, in your example, if Exec gets fired on a Weds., then he still gets the deferred comp on 12/31/10. So, the parties have in effect an election as to payment time. Not permissible.

In the involuntary termination situation, almost universally that distinction determines whether the Exec gets certain comp (yes if w/o cause, no if resigns), and not the timing of the comp.

xtremErisa - said...

I only intend in my original post to be considering the structure in which the plan indeed says you get paid vested deferred compensation on (i) date X, or (ii) if you are terminated without cause, the date of termination, if earlier. While as you suggest this structure is probably substantially less common than one under which the type of termination acts as a vesting event, I don't think I agree that the payment-acceleration structure is "almost universally" not used. Moreover, the payment-acceleration scenario is indeed the type of arrangement I've understood "them" informally to be addressing when discussing the subset issue. I think that the ability to go forward under 409A with the subset approach to accelerated payment in the early-termination situation can be important, if that's the desired structure. I would also suggest that the whole analysis raises the potentially important further questions I referred to towards the end of my original post regarding how far the concept can be taken. (As to a plan under which the manner of termination relates to vesting (rather than to when or how payment is made), such as in the type of plan you posit in your second paragraph, I agree that the subset issue addressed in my post wouldn't arise in that case.)

xtremErisa - said...

It is noted that this issue has been addressed - favorably - in the 2009 JCEB Q&As, at Q&A 18 thereof.

Unknown said...

To delve deeply into odds is an exciting adventure. At first glance odds may seem unenchanting, however its study is a necessity for any one wishing to intellectually advance beyond their childhood. Though odds is a favourite topic of discussion amongst monarchs, presidents and dictators, odds is not given the credit if deserves for inspiring many of the worlds famous painters.