Vertical spreads, from the arithmetic up

What a vertical spread is

A vertical spread is a two-leg options position defined by four constraints. Break any one of them and you no longer own a vertical, you own something else with a different risk profile.

• Same underlying. Both legs reference the same stock, ETF, or index.
• Same expiry. Different expiries make it a calendar or diagonal, and those carry open-ended volatility and term-structure risk that a vertical does not.
• Same option type. Both calls or both puts. Mixing types gives a combo, a collar, or a synthetic, not a vertical.
• Two different strikes, one long and one short. One contract bought, one sold, usually in equal quantity. Unequal quantity is a ratio spread, which reintroduces undefined risk on the excess short leg.

It is called vertical because on a traditional option chain the strikes run down the page while expiries run across it. The two legs sit in the same column, one above the other.

The distance between the strikes is the width, and width is the single most important number in the position. Because equity options carry a 100-share multiplier, a $5 wide spread is a $500 object. Every max profit and max loss figure in this guide resolves to some split of that $500.

The four verticals

There are exactly four, formed by two directions crossed with two option types. Two are paid for up front (debit), two are sold up front (credit).

Notice the symmetry. A bull call spread and a bull put spread express the same directional view at the same strikes with nearly the same payoff shape. The difference is which side of the cash flow you sit on, how much capital the broker holds, and which leg carries the assignment risk. Direction and cash flow are independent choices.

A useful shorthand: in a debit spread the leg you buy is closer to the money than the leg you sell. In a credit spread the leg you sell is closer to the money than the leg you buy. That one sentence determines the sign of theta, the sign of vega, and where assignment shows up.

Debit spread arithmetic, worked

Take a stock trading at $100.00. A 100/105 bull call spread: buy the $100 call at $4.20, sell the $105 call at $2.00. Net debit $2.20 per share, which is $220 per contract.

Width = 105.00 − 100.00 = 5.00 → $500 per contract Net debit = 4.20 − 2.00 = 2.20 → $220 paid Max profit = width − debit = 5.00 − 2.20 = 2.80 → $280 Max loss = debit = $220 Breakeven = lower strike + debit = 100.00 + 2.20 = $102.20

The $220 leaves the account the moment the order fills, and it is the entire risk. Nothing the stock does can cost more than that, which is the defining feature of the structure. The $280 upside is reached and then frozen the instant the stock touches $105.00, whether that happens on day one or at the closing bell on expiry.

Here is the payoff at expiry across a range of prices. Spread value is the long call's intrinsic value minus the short call's intrinsic value, and profit is that value times 100 minus the $220 paid.

The last three rows are the trade-off in plain sight. A $20 move and a $5 move pay identically. You sold the tail in exchange for a cheaper entry, and if the underlying is the kind of instrument that occasionally travels 20% in a week, you have sold something expensive. This is the same reasoning that makes leveraged instruments behave differently from what a simple directional view suggests.

Credit spread arithmetic, worked

Same stock at $100.00. A 95/90 bull put spread: sell the $95 put at $1.60, buy the $90 put at $0.70. Net credit $0.90 per share, $90 per contract, on the same $5 width.

Width = 95.00 − 90.00 = 5.00 → $500 per contract Net credit = 1.60 − 0.70 = 0.90 → $90 received Max profit = credit = $90 Max loss = width − credit = 5.00 − 0.90 = 4.10 → $410 Breakeven = short strike − credit = 95.00 − 0.90 = $94.10 Return on risk = 90 ÷ 410 = 21.95%

The shape is the point. You win a small amount across a wide band of outcomes and lose a large amount in a narrow one. Risking $410 to make $90 means you need to be right far more often than you are wrong just to break even.

Buying power reduction

Sell a naked $95 put and the broker wants the account to survive the stock going to zero. Cash-secured, that is $9,500 minus the $160 credit, so $9,340 held against one contract. Add the long $90 put and the required capital collapses.

Credit spread BPR = (width × 100) − credit received = (5.00 × 100) − 90 = $500 − $90 = $410 held per contract

The broker holds exactly the max loss, no more, because the long leg makes the worst case knowable. Below $90.00 every additional dollar the stock falls costs the short put a dollar and pays the long put a dollar. The two cancel, permanently, and the loss stops at $410.

That is a large capital effect: $410 held versus $9,340 for the cash-secured naked put. Return on capital for the same $90 goes from 1.71% to 21.95%. The catch is that the higher percentage is computed on a much smaller base, and it is not free money. Both positions carry the same short $95 put and the same probability of being wrong. The long leg buys you a floor, and you paid $70 for it out of a $160 gross credit, which is 44% of the premium spent on insurance.

Buying power reduction is also what limits how many spreads an account can carry at once, which makes it a position sizing constraint rather than an accounting detail. The per-trade sizing arithmetic applies with max loss standing in for the stop distance, and the position sizer handles the division.

Debit versus credit, side by side

Neither column is better. They are two ways of being paid for the same opinion, and the arithmetic in the last four rows is what actually differs.

The 75% win rate that still loses money

Credit spreads are marketed on win rate, and the win rate is genuinely high. It has to be. The break-even win rate for any binary payoff is fixed by the ratio, not by the strategy's reputation.

Break-even win rate = max loss ÷ (max loss + max profit) Credit spread: 410 ÷ (410 + 90) = 410 ÷ 500 = 82.0% Debit spread: 220 ÷ (220 + 280) = 220 ÷ 500 = 44.0%

So a 95/90 credit spread must win more than 82 times in 100 to make a cent. Run it at 75%, a win rate most traders would call excellent, over 100 all-or-nothing trades:

75 wins × $90 = +$6,750 25 losses × $410 = −$10,250 Net = −$3,500 At 82%: (82 × 90) − (18 × 410) = 7,380 − 7,380 = $0

Three quarters of the trade log is green and the account is down $3,500. This is the single most common way retail traders lose money selling premium, and it is not a market failure, it is arithmetic that was fixed at order entry.

Real trades are not purely binary, since many losers get closed before reaching full width and some winners get taken early, which moves the true threshold around. That is precisely why the number worth tracking is expectancy per trade rather than win rate. Win rate on its own is close to meaningless without the average win and average loss beside it, and the expectancy calculator will not let you quote one without the other.

Defined risk is not small risk. The max loss on a credit spread is the number the broker holds, and it happens in full whenever the underlying closes past the long strike. Selling a $5 wide spread for $0.90 twenty times and taking one max loss wipes out roughly 4.5 winners. Positions can also stop being defined without warning: an early assignment on the short leg leaves you holding stock or short stock overnight, with gap exposure the spread was supposed to prevent. Nothing on this page is a recommendation to trade any of it.

When a spread beats a single long option

• Implied volatility is elevated. Buying an outright option into high IV means paying for volatility that usually reverts. Selling a leg against it means you are paying high prices and receiving high prices at the same time, which largely offsets.
• Theta is a problem. The short leg's decay funds part of the long leg's decay, so a debit spread bleeds more slowly than the equivalent naked long, and a credit spread collects.
• Cost. $220 versus $420 for the outright $100 call. Same directional idea, roughly half the capital at risk, and the difference is the tail you gave up.
• The risk needs to be defined. A short vertical has a known worst case. A naked short option does not, and the difference is a margin call.
• The target is a level, not a direction. If the thesis is resistance near $105, the payoff above $105 was never part of the plan, so selling it costs nothing you expected to collect.

When it does not

• The move could be enormous. Earnings, a catalyst, a takeover. Capping upside is exactly wrong when the whole idea is the tail.
• Implied volatility is already low. There is little premium in the short leg, so it barely reduces cost while still capping the gain. You give up a lot for a small discount.
• The chain is illiquid. Two legs means two bid-ask spreads to cross on entry and two more on exit. On a wide chain this cost dwarfs everything else.
• Very short dated. Gamma near expiry makes both legs jump, and the position can swing between near-max-profit and near-max-loss within a single session.
• You will need to manage it under pressure. A single long option has one decision: sell it. A spread has four legs of decision at expiry and an assignment path underneath.

Implied volatility: why vega mostly cancels

A long option is a long volatility position. If implied volatility drops after you buy, the option loses value even if the stock does exactly what you predicted. That is IV crush, and it is why so many correct earnings calls still lose money.

In a vertical the two legs have vega of opposite sign because one is long and one is short. They largely cancel. They do not cancel completely, because vega peaks at the money and falls off in both directions, so the leg closer to the money always has the larger vega.

Long $100 call vega = 0.10 per share per vol point Short $105 call vega = 0.07 per share per vol point Net vertical vega = +0.03 → $3 per contract per vol point 5-point IV drop: Vertical: 5 × $3 = −$15 on a $220 debit → 6.8% Naked call: 5 × $10 = −$50 on a $420 debit → 11.9%

The debit spread's long leg sits closer to the money, so it retains a small positive vega and still dislikes an IV collapse, just far less than the outright. The credit spread's short leg is closer to the money, so its net vega is negative: a volatility spike hurts it before the stock has moved anywhere. Less IV-sensitive is not IV-neutral, and residual vega grows as the strikes get further apart.

Theta runs in opposite directions

Time decay is not a property of options generally. It is a property of a specific position's net short or net long premium.

A debit spread that is out of the money loses value as time passes, because the long leg is closer to the money and therefore decays faster than the short leg you are collecting on. It needs the move, and it needs it before expiry. A credit spread out of the money gains value with time for the mirror-image reason.

There is a clean exception worth knowing: a debit spread that is already fully in the money has positive theta. Its value can only converge upward to the full width as the remaining extrinsic value on both legs drains away, so time becomes an ally at exactly the point most traders assume it is an enemy. This is also why a fully in-the-money debit spread rarely trades at full width before expiry: the market is still pricing the chance it comes back.

Assignment, exercise, and pin risk

US listed equity options are American style, which means the short leg can be assigned on any business day, not just at expiry. Index options are usually European style and cash settled, which removes early assignment entirely. Knowing which one you are holding matters more than most of the Greeks.

Early assignment on short calls, and dividends

A rational holder exercises a call early only when the remaining extrinsic value is worth less than the dividend they would capture by owning the shares. So early call assignment clusters on the day before the ex-dividend date, on in-the-money short calls with little time value left.

If it happens, you are short 100 shares at the short strike, you owe the dividend on those shares, and you still hold the long call. The economics remain capped, but the position is no longer the thing you opened. The usual resolutions are to exercise the long call to cover the short stock, close the whole thing, or buy the shares back and sell the long call for its remaining value.

Early assignment on short puts

Short puts get assigned early when they are deep in the money and extrinsic value has gone to roughly nothing, since the holder gains by taking the cash now. You end up long 100 shares per contract at the short strike, with the cash outlay that implies, and a long put still open beneath it. If the account cannot fund the shares, the broker may liquidate at the open on its own schedule rather than yours.

Pin risk and the single assigned leg

Pin risk is the expiry-day case where the underlying settles almost exactly at the short strike. You do not know whether you will be assigned, and you do not find out until the weekend has passed. If you are assigned on the short leg while the long leg expires out of the money, Monday opens with an unhedged 100-share stock position and full overnight gap exposure. That is the moment a defined-risk trade stops being defined.

The related trap is auto-exercise. The OCC automatically exercises options that finish $0.01 or more in the money unless instructed otherwise. A spread where both legs finish in the money settles to the full width, which is fine, but it does so through two separate exercise or assignment events, each of which may carry a fee and each of which briefly creates a real stock position. Closing the spread before the final bell avoids the entire sequence, and the cost of doing so is usually a few cents of spread.

When only one leg fills

Legging in, entering the two legs as separate orders, is how traders accidentally end up naked. Fill the short call and miss the long call, and you are short an uncovered call with theoretically unlimited risk until the second order fills. Fill the long leg only, and you have simply bought an option at a price you did not intend to pay.

The defence is mechanical: submit the position as a single spread or combination order so it fills as one package or not at all. The same applies on exit. Closing the profitable leg first and leaving the other open is a decision to convert a defined-risk position into an undefined one, usually made under the pleasant feeling of being right.

Liquidity is the real cost

Commissions are visible and small. The bid-ask spread is invisible and large.

Suppose the 100/105 vertical is quoted $2.10 bid, $2.30 ask, with a $2.20 mid. Paying the ask to open and hitting the bid to close gives up $0.10 each way, $0.20 per share, $20 per contract round trip.

Slippage (paying ask, selling bid) = 0.20 × 100 = $20 per contract Commissions at $0.65/leg = 4 legs × 0.65 = $2.60 Slippage ÷ commissions = 20 ÷ 2.60 ≈ 7.7× Slippage ÷ max risk = 20 ÷ 220 = 9.1% of the trade's entire risk

Slippage here is nearly eight times the commission and consumes over 9% of the risk before the stock moves at all. On the credit spread it is worse in relative terms: $20 against a $90 max profit is 22% of the best possible outcome. This is why the option chain liquidity matters more than the strategy's name, and why an idea that tests well on mid prices can be flat or negative on real fills. Any backtest of an options strategy priced at the mid is describing a market you cannot trade in.

Commissions per leg

Options commissions are charged per contract per leg, so a vertical costs twice what a single option costs, both opening and closing. At a common $0.65 per contract:

1 contract: open 2 × 0.65 = $1.30 round trip = $2.60 10 contracts: open 20 × 0.65 = $13.00 round trip = $26.00 $26 against $2,800 max profit (10 debit spreads) = 0.93% $26 against $900 max profit (10 credit spreads) = 2.89%

Now fold commissions into the break-even win rate. The credit spread's win becomes $87.40 and its loss becomes $412.60, so the threshold moves from 82.0% to 82.5%. Half a percentage point sounds trivial until you realise it is roughly one extra winning trade in every two hundred, permanently, for nothing.

Expiring worthless is cheaper than closing, since there is no exit commission, but choosing to hold into expiry to save $1.30 while carrying assignment and pin risk over a weekend is a bad trade in every dimension except the one being measured. Log the fees per trade rather than estimating them, and let the trade journal compute net expectancy from actual fills.

What to record if you study these

1. Width, net debit or credit, and max loss in dollars, at entry, before anything moves.
2. Break-even win rate computed at entry. If you cannot realistically clear it, the structure was wrong regardless of the outcome.
3. The fill price against the mid at the time of entry, so slippage becomes a measurable line item rather than a vague complaint.
4. Implied volatility of both legs at entry and exit, so IV changes can be separated from directional results.
5. Whether the position was closed, expired, or assigned, and whether any leg was ever held alone.
6. Net result after all commissions and fees, in R, so it is comparable to every other trade in the book.

Thirty to fifty logged spreads is roughly where per-structure expectancy starts to mean anything, and even then the confidence interval will be wide. Sizing conclusions from small samples is the failure mode that drawdown arithmetic makes expensive, and testing on a simulator first is what the paper-to-live process exists for. Treat every number on this page as arithmetic, not as an edge: the payoff formulas are exact, the profitability of any particular spread is not.

Educational content, not financial advice. No live profit-and-loss figures are claimed anywhere on this site; backtest and walk-forward results are always labelled as such. Full terms: /terms.html

Tools referenced in this guide

• Trader's risk toolkit — position sizer, expectancy and break-even calculators for turning max loss into a contract count.
• TradeLog — log spreads in R with fills and fees so net expectancy is computed rather than guessed.
• R-multiple & expectancy guide — why win rate without average win and average loss tells you nothing.