Research Memo
What a zero-cost structure actually costs, and why the tax treatment usually decides whether it worked
Framework memo. Illustrative throughout; no client facts are described. Constructive sale and straddle treatment are legal determinations. Nothing here is tax or legal advice. Premium mechanics are developed further in the covered call memo.
An option overlay on a single position is presented as risk management, and it is risk management. It is also a financing transaction, and the two functions are present in every version of the structure in different proportions. A presentation that describes only the first is incomplete in the same way that describing a covered call as an income strategy is incomplete: not wrong about what it does, wrong about what it is.
A collar exchanges future upside for present certainty. The exchange rate on that trade is the entire analysis, and most presentations show the payoff diagram without ever pricing what was surrendered.
The second half of the analysis is tax, and it is where these structures most often fail. A collar that looks sound before tax can suspend a holding period, defer losses, and strip preferential treatment from dividends the client was receiving. Those costs do not appear on the payoff diagram either.
1 · Zero cost means the premiums net to roughly zero, not that the structure was free. The put was paid for with upside. Where the underlying carries meaningful skew, the client surrenders more potential gain than the downside they removed. Whether that exchange is worth making is a judgment about the client, but it should be stated in those terms rather than described as costless. (§3)
2 · Coverage ratio is the first design decision, not the strike. How much of the position does the client want fully exposed? Write only against the remainder. Setting strikes before answering that question produces a structure sized by the option chain rather than by the client. (§4.1)
3 · A collar written too tightly is a sale. Under §1259, a position that substantially eliminates both risk of loss and opportunity for gain can be treated as constructively sold, realizing the gain the structure existed to defer. The band has to retain meaningful exposure in both directions, and how much is a legal determination rather than a modeling one. (§5)
4 · The straddle rules can cost more than the structure saves. A collar generally creates a straddle. Losses on one leg are deferred, carrying charges are capitalized, and the holding period on the underlying can be suspended. Where the shares were not already long-term, the overlay can convert a long-term gain into a short-term one. (§6)
5 · Hedging can disqualify dividends from preferential treatment. Qualified dividend treatment requires a holding period that does not count days on which the holder has diminished risk of loss. A collar is such a position. For a client holding a high-yielding stock, the overlay can quietly convert preferentially taxed dividends into ordinary income for as long as it is in place. (§6.3)
A collar is long stock, a purchased put below spot, and a written call above it. The payoff is bounded: losses stop at the put strike, gains stop at the call strike.
Described that way it is a risk structure, and that description is accurate. But consider what has happened economically. The client has given up a claim on future appreciation in exchange for present certainty, and the counterparty has taken the other side. That is an exchange of value across time, which is what financing is.
The proportions vary by structure, and seeing them on a spectrum is more useful than treating each as a separate instrument.
A covered call is mostly a short volatility position with a small financing character: cash now, upside surrendered later. A collar sits in the middle, exchanging upside for downside protection with the two premiums roughly offsetting in the standard construction. A prepaid variable forward, treated in §7, is mostly financing with the hedge attached: the client receives most of the position's value up front and commits to deliver shares later.
Same two ingredients throughout. What changes is the mix, and the mix determines which risks the client has actually taken on.
The standard construction sets the call strike so that the premium received for the call approximately offsets the premium paid for the put. Cash moves in both directions and nets to roughly zero, and the structure is described as costless.
The premiums netted. That is the only sense in which it was free.
Equity index options exhibit persistent negative skew: out-of-the-money puts trade at higher implied volatility than out-of-the-money calls at comparable distance from spot. The pattern is textbook, described in Hull among others, and it became a structural feature of index option markets after 1987. The Options Industry Council attributes it in part to hedging activity itself, since investors who are long assets buy protective puts and write upside calls, creating persistent demand on one wing and supply on the other.
Where that skew is present, the consequence for a collar is direct. The put being purchased is richly priced relative to the call being written, so funding one with the other requires writing the call closer to spot than the put. The protected downside and the surrendered upside are not symmetric, and the asymmetry runs against the client.
Where skew is steep, a structure protecting against a decline of a given size will cap appreciation at something less than that size. That is not an execution failure. It is the price of the insurance, paid in upside rather than in cash.
The skew described above is largely an index phenomenon, and a collar on a concentrated position is written on a single name.
Bakshi, Kapadia and Madan find that option-implied distributions for individual equities are substantially less negatively skewed than the index distribution. Bollen and Whaley find that the two markets are driven by different order flow: index options are especially sensitive to put buying pressure, while single-stock option markets can be more affected by call buying pressure, which pushes the surface in the opposite direction.
So the asymmetry cannot be assumed on a single name. On some underlyings it is mild, and on a name with heavy call demand the exchange can be neutral or favorable. The practical instruction is to measure the surface on the specific underlying rather than apply an index intuition to it, and to quote the client the actual upside being surrendered rather than a general argument about skew.
This is the same shape as the finding in the covered call memo, and for the same structural reason. The premia that are reliably available in broad baskets are weak or absent in individual names, because the systematic demand that creates them operates at the index level.
Stated properly, the question put to the client is whether they would accept a known cap on appreciation in exchange for a known floor under the position. For a client whose concentration genuinely threatens their plan, the answer is frequently yes and the asymmetry is acceptable.
For a client who is simply uncomfortable, it may not be. The honest version of the conversation includes the number: here is the appreciation you are giving up, over this period, to remove this much downside. A client who sees that trade and still wants it has made a decision. A client shown only the payoff diagram has not.
Before any strike is chosen, decide how much of the position the client wants to keep fully exposed. Collar only the remainder.
This single decision does more to shape the outcome than strike selection does, and it is frequently skipped because the option chain presents strikes and not ratios. A client who still believes in the underlying and wants protection against a catastrophic outcome is describing a partial collar. A client who wants the position's risk substantially removed is describing something closer to a sale, and should be shown that comparison.
The unencumbered shares serve a second purpose. They preserve flexibility: they can be gifted, sold on the client's schedule, or used to fund a buyback if the written calls move against the plan.
Longer-dated structures cost less per unit of time and require less maintenance, but they commit the client for longer and are harder to unwind cleanly. Shorter structures preserve flexibility and accumulate transaction costs, which compound against a position where the entire economics are a few volatility points.
Where the overlay exists to manage a position through a defined period, a planned unwind, a vesting schedule, a lockup expiry, the tenor should match that period rather than being chosen from what is liquid.
Single-name option markets vary enormously in depth. A structure that cannot be adjusted or unwound without material slippage is a commitment rather than a position, and that should be understood before it is entered rather than discovered when it needs to change.
Section 1259 treats certain transactions in appreciated financial positions as constructive sales, triggering immediate recognition of gain. The provision exists precisely to prevent a taxpayer from eliminating economic exposure while deferring the tax.
A collar that removes substantially all risk of loss and substantially all opportunity for gain is the central case the provision contemplates. A collar with a wide band generally is not. Where the line sits between those two is not a question an investment professional should answer alone.
Two practical points follow.
The structure has to be designed with this constraint in mind from the outset, not checked afterward. A collar tightened to improve the economics can cross into recognition and realize the entire embedded gain, which is a far worse outcome than the structure was ever going to improve on.
And the determination should come from tax counsel on the specific terms. This is the single technical point in this memo most likely to be gotten wrong by someone reasoning from the economics, because economically a tight collar looks like exactly what the client wanted.
For a position held in a taxable account, the tax consequences of the overlay frequently exceed its economic benefit. They are also entirely predictable in advance, which makes them the cheapest part of the analysis to get right.
Offsetting positions in the same property generally constitute a straddle under §1092. A protective put against a long stock position qualifies. The consequences are a set of restrictions rather than a single rule: losses on one leg are deferred to the extent of unrecognized gain on the other, and certain interest and carrying charges must be capitalized rather than deducted.
The practical effect is that the client cannot harvest a loss on the hedge while the underlying carries an offsetting gain, which removes one of the benefits an overlay might otherwise have provided.
Entering a straddle can suspend the holding period of the underlying while the offsetting position is open. Where the shares were already long-term before the collar was established, this is largely moot and the exposure is minimal.
Where they were not, it matters a great deal. A position approaching its one-year mark can have the clock stopped by the overlay, and a sale that would have been long-term becomes short-term at ordinary rates. For a recently vested or recently acquired position, this should be checked before anything is written.
This is the consequence most often missed, and it applies to a structure that is otherwise working.
Preferential treatment of qualified dividends requires the shareholder to satisfy a holding period around the ex-dividend date, and days on which the holder has diminished risk of loss do not count toward it. A collar is a diminished-risk position.
So a client who collars a dividend-paying stock may find that dividends received while the structure is in place are taxed as ordinary income rather than at the preferential rate. For a high-yielding position held by a client in the top bracket, that recurring cost can be material relative to whatever the overlay saved, and it compounds for as long as the hedge remains.
The point is not that collars should be avoided on dividend payers. It is that the comparison should include this cost, and it almost never does.
Worth knowing rather than reaching for. A prepaid variable forward combines a collar with a loan: the client contracts to deliver a variable number of shares at a future date, typically one to five years out, and receives cash up front amounting to roughly three-quarters to nine-tenths of current value. The number of shares delivered at settlement depends on the price then, bounded by a floor and a cap. The shares are pledged as collateral in the meantime.
It sits at the financing end of the spectrum described in §2. The client has received most of the position's value and committed to a future delivery, which is a very different posture from holding shares with options written around them.
The tax treatment is settled in outline and sensitive in the detail. Revenue Ruling 2003-7 holds that a conventional prepaid variable forward is neither a current sale nor a constructive sale. In Anschutz Co. v. Commissioner, decided by the Tax Court in 2010 and affirmed by the Tenth Circuit in 2011, the taxpayer's forwards were accompanied by share-lending agreements permitting the counterparty to borrow and sell the pledged shares, with the taxpayer receiving a fee. The courts examined the arrangement as a whole and held that a current sale had occurred, recognizing gain on several hundred million dollars of proceeds at inception.
The distinction that survives is narrow and worth holding onto: the conventional structure stands, and attaching a share loan to it does not.
These were more widely used before that decision than after it. For most clients the combination of complexity, counterparty exposure, cost embedded in terms that are difficult to compare, and tax sensitivity makes a plain collar or a staged sale the better answer. The structure is included here because it is the clearest illustration of the memo's argument: it is the same two ingredients as a collar, mixed differently, and the difference in mix changes what the client has actually done.
A collar bounds the outcome on a position. It does not reduce concentration, it does not realize the gain, and it does not move the client toward a diversified portfolio. The shares are still there, the embedded gain is still there, and the problem that prompted the conversation is still there when the structure expires.
What it buys is time, and time is genuinely valuable where there is something to do with it. An overlay running alongside a staged sale protects the position while the sale proceeds. An overlay running alongside nothing is a client paying in upside to postpone a decision.
Where the objective is diversification rather than protection, the overlay is at best a complement to a disposition and should be presented that way. The concentrated positions framework memo treats the alternatives and the conditions under which each applies.
This memo prices the structure, not the underlying. Whether the position should be hedged at all depends on a view about the security and about the client's capacity to bear the risk, neither of which is addressed here.
Skew is not constant. The exchange rate described in §3 moves with market conditions, and a collar priced in a calm market differs from the same structure priced during a decline. The direction of that difference generally disfavors the client at the moment they most want the protection.
Execution quality matters more than it appears. Two legs, each with a spread, on a single-name option chain that may be thin. The quoted economics and the achieved economics can differ by more than the structure's entire expected benefit.
Restricted stock introduces constraints this memo does not cover. Insider status, trading windows, and employer hedging policies frequently prohibit these structures outright, and that should be confirmed before any analysis is performed.
Framework memo prepared for illustrative purposes. Not tax or legal advice. Constructive sale treatment under §1259, straddle treatment under §1092, holding period consequences, and qualified dividend eligibility are legal determinations that depend on the specific terms of a structure and should be confirmed with qualified counsel before implementation.
CFA® and Chartered Financial Analyst® are registered trademarks owned by CFA Institute.