Wednesday, May 06, 2015

Tax policy colloquium, week 14: Gregg Polsky's "A Compendium of Private Equity Tax Games"

Here is a picture of a place on Route 23 in Northern New Jersey that is reported to have excellent kielbasa, pierogis, etc.  I'm definitely hoping to try it sometime.

Yesterday, at the last session of the 20th annual NYU Tax Policy Colloquium, we instead had "Polsky Smack" - an illuminating smackdown by Gregg Polsky of current or at least recent practices in the private equity realm.

One could call this our third installment, over the years (I hope I am not forgetting one) of significant contributions to the private equity / "2 & 20" debate.

A long time back, Vic Fleischer presented the initial "2 & 20" piece at our colloquium.  This was before he published it, and thus before it became a huge story.  We then, a few years later, had Chris Sanchirico's piece, in which he discussed the counterparty implications and the fact that one has to look at the transaction as a whole.  (BTW, Chris will be co-teaching the NYU Tax Policy Colloquium with me in January-May 2016, when we will once again meet on Tuesdays from 4 to 6 pm.)

What Polsky's paper brings to the party is a detailed account of current practice, which has some eye-opening elements.  Here's a little hypothetical that can help to convey it here.

Suppose a PE hotshot (aka the Manager) gets tax-exempt investors, such as from pension funds and university endowments, to pony up $100 million ($100M) to buy stock in a public company, meant to be turned over for a large profit in fairly short order.  Despite the difficulties of outsmarting the capital markets, this effort actually succeeds.  In just over a year, the stock is sold for $200M.  (Rather a flattering example, don't you think?  I rather doubt that this is par for the course, and hesitate to reinforce genius-PE mythologies, but it makes for nice round numbers.)  The only other flow of funds in the interim, other than between the players in this little drama, is $1M spent out of pocket by the Manager.

They make a standard "2 and 20" deal, under which the Manager will get 2% of the funds under management (i.e. $2M) and 20% of the capital gain (i.e., $20M, as it turns out, given the $100M profit).  But the Manager is also supposed to kick in $1M so he has "skin in the game."  He therefore ends up with a net of $20M (i.e., -1 + 2 + 20 -1).

Economically, we may think of this as labor income.  But it's plausible that he will get to treat the entire $20M as capital gain (CG), taxable at only a 20% rate.  With taxable investors, this would be a detriment to them, but since they are tax-exempt they don't care.

How does he manage this?  The paper identifies 4 strategies, 2 of which it agrees are legal under present law, but 2 of which it says are not under typical deal terms.

Strategy 1: This is the classic CG treatment for the 20% "carry," first identified in academic circles by Fleischer, and reflecting the standard rule of partnership tax law that the entity-level characterization of income (capital gain from selling a capital asset) works through to the partner level.  Despite a recent case on ERISA law, Sun Capital, which arguably supports viewing the PE as in business, and thus as selling a business asset, rather than a capital asset, yielding ordinary income, Polsky believes that this classification is likely to be safe unless Congress enacts a statute changing it. Such an enactment strikes me as unlikely, despite public sentiment which probably (insofar as the public has a view) would strongly support it, unless Congress needs a pay-for to help fund something bigger, in which case it might start to become affirmatively likely.

All this was well-known before.  But what about the +2M, which is ordinary income, and the two -1Ms?

Strategy 2: Through bogus paper-shuffling that lacks any economic substance - as we are told in the paper, based on familiarity with real-world examples - the parties pretend to do a deal in which the manager waives $1M of his $2M in exchange for boosting his capital interest from 20% to 21%.  Hence, with these numbers, he loses $1M in ordinary income fees, but gets an extra $1M of back-end capital gain.  In effect, they purport to change the deal from "2 and 20" to "1 and 21," but highly tailored arrangements are used to make sure that absolutely nothing will actually change.

A point to note here is that, if the parties actually did agree to "1 and 21" instead of "2 and 20," then unambiguously there would be only $1M of ordinary income and $21M of capital gain (ignoring for now the $2M in outlays).  But the parties apparently didn't want to do this upfront, since the Manager likes getting a $2M guarantee.  So they wait until they know the outcome and then purport to change it, without actually doing anything to subject the Manager to the entrepreneurial risk that he had been eager to limit to the 20%.  Thus, suppose the law was being enforced properly, and the parties responded by keeping the deal at "2 and 20."  This would be an example of the Manager's risk aversion operating as a friction that constrained tax optimization,  There is really no other reason for us to care whether he bears more risk or less.

Getting to the paper's elements of "Polsky smack," the paper argues that, under actual deal terms that are standard, and in light of such provisions as IRC section 707(a)(2(A), which deals with disguising payments for services as other partnership transactions, the chance that this strategy actually "works" as reported is so low as to invite comparison between any tax lawyers who give the thumb's up here and those who were marketing corporate tax shelters 15 years ago.  This is not a characterization that the lawyers working on these deals are likely to embrace.  As it happens, I gather that the elite law firms that often do the transaction work here pointedly do not opine, nor are they asked to, that the strategy actually works.  But the IRS hasn't issued express guidance to the contrary, and the audit rate here (so far as we know) is effectively zero.

Strategy 3: OK, so now we're down to $1M ordinary income and $21M CG, leaving aside the offsets.  Suppose the remaining $1M management fee is used to fund the "skin," which is unproblematic (after all, it is still first reported as gross income).  That provides an actual $1M basis for the capital interest that yields the $21M gross return, so we are down to $20M CG, which is fine taking everything else here as given.  But what about the $1M that the Manager spends out-of-pocket?  Since he is spending it in order to earn $1M ordinary income and $20M capital gain, arguably it should be allocated between the two, and perhaps 95.2% to the latter if one adopts a pro rata approach in the absence of anything better.  But because of the so-called INDOPCO regulations (perhaps better termed the anti-INDOPCO regulations, since they involved the Treasury's deciding to retreat comprehensively from a Supreme Court victory on capitalization vs. expensing issues in the eponymous case), the entire amount is allowed to be expensed.  Thus, it reduces the Manager's ordinary income to zero, while his CG remains at $20M.

Here, like Strategy 1, we have something that the paper agrees is legal under current law, although arguably inappropriate on policy grounds.  BTW, note that, if I can decide to spend additional amounts, deductible against ordinary income that is taxed at, say, a 39.6% marginal rate, in order to generate additional CG that will be taxed at only a 20% rate, I may profit after-tax even if each extra dollar that I spend generates less than a dollar of extra CG.

Strategy 4: Here we are back in the realm of Polsky smack, i.e., the calling out of taxpayers and their tax advisers for doing things that may be unsupportable under present law, and thus reliant on the audit lottery for their payoff.  (Although it's not really even a "lottery" if the audit rate is effectively zero.)  The investors can't use the deduction for the $1M gross fee that remains after purporting to convert half of the original $2M into more "carry."  So they push down the deductions to the portfolio company - i.e., they cause it, rather than the partnership with all the investors in it, to be the party that actually pays this amount to the Manager.  There may even be a 1-to-1 offset - i.e., each penny paid by the company reduces by a penny the amount to be paid by the investors through the fund partnership.  This is done in a non-arm's length fashion, and without regard to any services actually offered to the company by the Manager, given that it doesn't matter economically which level pays.  It would be looting of the company if there were minority shareholders outside the partnership, but since there aren't, it is instead (the paper argues) a disguised dividend.

One point of possible interest here: Even if this push-down of the fee deduction has no economic substance whatsoever, it is possible that the Manager is actually doing things for which the company would be willing to pay, in an arm's-length set of arrangements between distinct parties.  In relation to this point, let's return to the basic question: How is it that they bought a $100M company, and just a year later sold it for $200M?  I see 3 basic possibilities, which differ in 3 dimensions.  First, would the company pay for it in an arm's length transaction?  Second, does it create income that the corporate tax reaches?  Third, does it have social value commensurate with its private value?  The first point relates to evaluating Strategy 4 if the parties did it properly from the start, while the second and third relate to my next topic here: evaluating the tax results normatively.

1) Stock-picking - To put it in terms of a polar case although in practice there might be a bit of each scenario, suppose the Manager does nothing whatsoever to the company. He is merely a stock-picker, who believes that it is undervalued and will soon go up.  At the termination date he is proved correct, although of course it might have been just luck rather than skill.

Here the company basically wouldn't pay anything for someone simply deciding to bet that its stock is going to go up.  Its operations and profitability aren't improved in any way.  Also, here the corporate tax doesn't reach the trading gain, which effectively is a betting transaction between winners and losers in the Great Casino on the side, and the private gain greatly exceeds the social gain (since the money is just going from other investors' pockets to those of the lucky ones in this deal).

2) Business strategy - Suppose the Manager does stuff to the company's operations, so that they become more profitable.  The company would pay for this at arm's length, the corporate tax will reach these added profits if it is otherwise operating effectively, and the private gain may at a first approximation equal the social gain,

3) Tax strategy - Suppose the Manager improves the company's tax planning, so that it pays a lot less tax on the same "true" profits as previously.  For example, this might involve levering up the company with lots more debt.  (But given the stock appreciation, apparently the market hadn't already been assuming that this would be done.)  The company would pay for this at arm's length, the corporate tax will not reach the added after-tax profits - it isn't taxed on paying less tax - and the private gain at a first approximation greatly exceeds the social gain.

OK, onto the bottom line: What if anything is wrong with all this in substance?  (Leaving aside the paper's central focus, which is on taxpayers not complying with the law on the books, and the IRS's failing to enforce that law.)  More specifically, is $20M in capital gain, rather than ordinary income, to a high-income Manager who is economically earning labor income as bad as it looks?

On the whole, I would say yes.  But two points that are at least quibbles, and potentially more than that, need to be addressed.

1) What about the counterparties? - If there were taxable, rather than tax-indifferent, counter-parties (i.e., the investors), the net tax benefit from arranging things so that the Manager gets CG instead of ordinary income would be reduced.  Indeed, if the net tax benefit were eliminated, I would say the problem was entirely eliminated, other than as a matter of optics.  But I gather that the investors really are generally tax-exempts.

Might a counter-party analysis still incline one to a more favorable view of the tax results here than otherwise?  This amounts to asking whether the tax-exempts ought to be able, in effect, to sell tax benefits that they can't use, in particular from the CG label for gross income that a transaction produces, and from investor-level deductions that they can't use.  Although this would require a longer discussion than I feel I should include here, my conclusion is No, but it's a fair topic for debate.

2) What about the entity-level corporate tax on the portfolio company? - Insofar as what's going on, beneath the surface of this little story, is that the Manager made the company more profitable via the choice of a new business strategy - and insofar as the corporate tax is actually functioning well enough to reach the resulting increase in corporate income - there really is no problem here.  Indeed, one might even conclude that taxing shareholder-level CG, including that pocketed by the Manager, results in inefficiently over-taxing corporate income relative to other income.  But this is not the case insofar as the $100M CG in this little story reflects either stock-picking, or the Manager's improving tax minimization at the corporate level.

1 comment:

Herry Johnson said...

Selling a business is a complex endeavor. A worst-case scenario is closing the business doors and walking away after giving up on all dreams to get decent cash for years of hard work. In buying a business, you eliminate the risks related to start-ups. However, ready for use an already assembled car can drive you instantly. Here are options for Sell business, buy business, business broker, Florida, business immigration, investment and private Equity.