How to Sell Cash-Secured Puts Systematically (Without Guessing)

The discretionary trap

Most premium sellers begin the same way. They read a primer, pick a stock they know, open the chain, find a strike that feels safe, and sell it. If the trade works, that counts as confirmation. If it does not, it counts as a learning experience. A few months later there is a profit curve drifting sideways, a handful of large drawdowns, and no clear rule explaining why those particular strikes were chosen on those particular days.

The decisions felt reasoned at the time. They do not survive being written down. Asked to state a rule that produced last Tuesday's trade and would also produce next Tuesday's, the discretionary seller usually cannot. The system lives in their head, where it is unstable, and it drifts with mood, market noise and whatever they read that morning.

A systematic approach replaces that with rules producing the same decision from the same data, regardless of who is at the keyboard. It does not promise better outcomes on any single trade. It promises a process you can measure and a reason for every decision.

What systematic actually means

Three properties. The inputs are defined: which universe, which data, which time of day. The decisions are rule-based, so that given the inputs the next action follows without a remaining "what do I think about this one" step. And the outcomes are recorded, so every position has a known rationale and a journaled result that can be audited later.

Miss any of the three and the process is not systematic. It is a habit wearing a checklist.

The benefit is not that systematic always wins. It is that systematic gives you something to improve. When a rule fails you can change the rule. When discretion fails, all you can change is how you feel about it.

Universe

Which underlyings are eligible? Liquid US equities and ETFs with active option markets is a typical starting filter. Inside that, sellers commonly screen further by market capitalisation, average option volume, sector, or the presence of an earnings catalyst.

What matters is that the universe is defined in advance. A trader willing to sell a put on anything will eventually sell a put on something they regret. Narrow the field before looking at strikes.

Timing

When is the trade entered? Some processes run daily at a fixed time, once the morning auction has settled. Some run on event triggers such as a known earnings date. Some run weekly on a fixed day.

Whichever it is, the rule needs to be repeatable. Trading when you happen to have time is not a timing rule.

Strike

Where most discretionary sellers leak their edge. Two common axes.

By delta, picking a target and selecting the strike closest to it. Delta responds to volatility, so the strike adjusts when the market reprices risk.

By percentage out of the money, picking a distance from spot. Simpler, and blind to volatility, so it stays put whether conditions are calm or violent.

Each has trade-offs, and most systematic processes commit to one with an explicit rule for when the other overrides it.

A statistical floor matters more than either. Where has the underlying actually moved in comparable conditions? Choosing a strike 8% out means little if the underlying has moved through that level repeatedly in recent comparable windows. A defensible strike sits below demonstrated downside, not just below an arbitrary percentage line.

This is what ShieldIQ supplies, and it is the reason a strike passing every other test can still be rejected.

Holding

The fourth decision is the one most sellers assume rather than make. After entry, do you manage the position, or do you let it run to expiry?

Both are defensible. What is not defensible is doing one while measuring yourself as though you were doing the other. A book closed at a profit threshold produces different statistics from a book held to expiration, and comparing them tells you nothing.

Everything scored here assumes the second. Every contract Strike and Patrol publish is evaluated on the basis that it runs to expiration, and outcomes are recorded on that basis: above the strike is a win, below it is a loss, nothing in between. That places the full weight on entry, because no later intervention is assumed to rescue a bad one.

It also changes what a good strike means. If there is no exit to fall back on, the distance between the strike and a defensible floor has to be right on day one. That is why safety dominates the score rather than sitting alongside other considerations.

If you manage positions actively, the scoring still works as an input. Your realised statistics will diverge from the published ones, and that is expected rather than a fault in either.

Why repeatability beats optimisation

Newer systematic sellers spend enormous effort fine-tuning entry rules and very little on whether the rules they already have are being followed. That is backwards.

Most of the edge in systematic premium harvesting comes from doing the same thing every day. Selecting the same kind of strike. Applying the same universe filter, including on the days when something outside it looks unusually inviting. Optimisation comes later, against a journal of trades all generated by the same process. Without that journal, optimisation is discretion with better branding.

How StratosIQ operationalises this

The platform removes the moment-to-moment load from each decision while leaving the decisions themselves with you.

Universe is handled upstream. Strike scans contracts on underlyings with earnings inside a defined window. Patrol scans the broader market daily. Both apply liquidity and volatility filters before anything reaches scoring.

Timing is handled by the publication schedule. Output arrives at the same time each day, so there is no screening loop to stitch together.

Strike selection is where the engine does the work. Each contract carries a score in which safety, the room between the strike and a defensible floor, carries the largest weight. Return on the capital committed carries most of the remainder, with smaller contributions for execution quality and contract geometry. The output is not "sell this one". It is a ranked, scored set you can apply your own threshold to.

The floor comes from ShieldIQ, reported as Fortified, Secure, Tight or Exposed. Contracts failing it do not publish. Patrol also applies eligibility limits on expiration and strike distance; anything outside them is rejected before scoring begins.

Since March 2026, the highest-scored contract published each day has finished out of the money 93.3% of the time, across 208 contracts. The full record is on the performance page.

Sizing stays with you. So does any decision to manage a position after entry, since both should apply across your whole book rather than come from a screening tool.

Where sellers usually start

Moving from discretionary to systematic rarely begins with a better model. It begins with writing down, in a couple of sentences each, what the rules for universe, timing, strike and holding actually are. Most people discover at that point that one or two of them do not exist yet.

Sophistication comes afterwards, once there is a process worth refining. What separates the two approaches is not the quality of the rules. It is whether they are followed on the days when following them is uncomfortable.

Worth reading next: how an individual trade is constructed, and three contracts that look good on screen.