Return on Capital vs Annualised Return: Comparing Contracts Fairly

Two contracts, no way to choose

One pays $120 on $5,000 committed for three weeks. Another pays $190 on $6,500 for eight weeks.

Which is better? The raw premiums say the second. The percentages say the first is 2.4% and the second is 2.9%, which still says the second. Both readings ignore that the second ties up more money for nearly three times as long.

Until the two are expressed on the same basis, comparing them is guesswork, and guesswork applied daily across a book compounds into a systematically worse selection than the seller believes they are making.

Return on capital

Start with the honest, unembellished number.

Return on capital = premium received / capital committed

For a cash-secured put, capital committed is the strike multiplied by 100 — the money set aside to honour the obligation.

• Contract A: $120 / $5,000 = 2.4%
• Contract B: $190 / $6,500 = 2.9%

This is a real figure and it has one flaw: it says nothing about time. A 2.9% return is better than 2.4% only if the periods match, and here they do not.

Use the bid, not the mid, when you calculate it. The mid is a price nobody is obliged to give you, and on a wide contract the difference is material — see the bid-ask spread.

Annualising

To compare across durations, scale to a common period:

Annualised return = (premium / capital) × (365 / days held)

• Contract A: 2.4% × (365 / 21) = 41.7%
• Contract B: 2.9% × (365 / 56) = 19.0%

The ranking inverts. Contract A, which looked worse on the raw figure, works your capital more than twice as hard. That reversal is the whole reason the calculation is worth doing.

Two notes on mechanics. Use calendar days, not trading days — capital is committed over weekends and time value decays through them. And this is a simple scaling, not compounding; compounding assumes you can immediately redeploy at the same rate, which is a claim about the future rather than a property of the contract.

Three ways the annualised figure misleads

It is the right comparison tool and a poor headline number. Three reasons, and all three are routinely ignored.

It assumes continuous redeployment that will not happen. A 41.7% annualised figure means "if I could do this exact trade continuously for a year". You cannot. Contracts of that quality are not always available, capital sits idle between positions, and after an assignment it is not available at all. The annualised figure describes one contract's efficiency, not a portfolio's return.

It flatters short durations enormously. The multiplier is inversely proportional to days held, so very short contracts produce spectacular annualised numbers from trivial premiums. A contract paying 0.4% over three days annualises to nearly 49%. It also carries sharp gamma risk and pays the spread frequently — time decay covers why faster is not better.

It is a gross figure. It contains no assignment losses, no spread, no commissions. A book of contracts each annualising at 40% does not return 40%; it returns that minus the cost of the ones that went wrong, and those losses are far larger than any individual premium. Why a 90% win rate can still lose money works through the arithmetic.

Capital committed is not always the strike

For a genuinely cash-secured put the denominator is straightforward: strike × 100. That is the money set aside and unavailable.

Two adjustments make the figure more honest.

Subtract the premium. You received $120 up front, so the net capital at risk is $5,000 − $120 = $4,880. It is a small refinement and it is the more accurate denominator.

Margin changes the number and not the risk. On a margin account the broker may require considerably less than the full strike. That inflates the calculated return dramatically — and the obligation is unchanged. The contract can still put 100 shares in your account at the strike. A return figure computed on a margin requirement is measuring the broker's collateral rules, not the position's economics, and comparing it to a cash-secured figure is comparing two different things.

For consistency, compute return on the full obligation. It is the conservative choice and it is the only one that stays comparable across accounts.

The comparison worth making

Annualised return ranks contracts against each other. It does not tell you whether to take any of them, and that is the more important question.

Two further comparisons make it useful:

Against the return on undeployed cash. The cash sits in your account regardless. If short-term rates are meaningful, the contract needs to beat that baseline to justify accepting the obligation at all — and in a higher-rate environment that hurdle is not negligible.

Against the downside it accepts. The 41.7% figure is only interesting alongside what happens when the contract fails. A high annualised return on a strike with a thin buffer is not a better trade than a lower one with a robust buffer; it is a different point on a risk curve. Ranking on return alone systematically selects the riskiest contracts available, because that is precisely where the market pays most.

How StratosIQ handles it

Return is one scored input among several and it is not the dominant one. Safety carries the largest weight — the distance between the strike and a defensible floor, measured by ShieldIQ against how the underlying has actually behaved rather than against a model-implied distance, and reported as Fortified, Secure, Tight or Exposed.

That ordering is a deliberate inversion of what the market offers. Premium is always richest where risk is highest, so an engine ranking on return would reliably publish the contracts most likely to fail. A contract failing the safety floor does not publish, however it annualises.

Because every contract is scored on a hold-to-expiry basis, the duration used in any return calculation is the actual contract duration rather than an assumed early close.

The full published record, contract by contract and including every loss, is on the performance page.

Worth reading next

Reading the options chain covers where the inputs these calculations use actually come from.