When to Close a Cash-Secured Put: 22 Years of SPY Data

When to close a cash-secured put is one of the most common questions new put sellers ask. Close early and you leave premium on the table. Hold to the end and one bad month can wipe out a year of wins.

So we tested the same entry rules on SPY: puts near 17 delta, about 45 days to expiration, with three different exit rules. Because only one position could be open at a time, each exit led to a different sequence of trades.

Closing at 50% of the credit won by $5,699.25 over holding to expiration. Most of that gap came from one trade in February 2020.

Why isn’t win rate the whole story?

Win rate isn’t the whole story because one large loss can erase many small wins. The dollars tell you what a strategy earned.

A cash-secured put is a trade where you sell a put option and set aside enough cash to buy 100 shares at the strike price. You collect a credit up front. The credit becomes profit only when the trade is closed or expires.

Here’s a simple example. Ten trades: nine earn $250 each and one loses $2,500. That’s a 90% win rate and a $250 net loss before costs.

That’s why this test tracks dollars, not just winners. New to short puts? Start with our rules-first guide to short put options.

How did we test the 17/45 Rule exits?

We tested Sqilled’s 17/45 Rule: sell the SPY put nearest 17 delta, about 45 days to expiration (DTE). Then we changed only the exit.

Delta estimates how much an option’s price changes for a $1 move in SPY, all else equal. Implied volatility (IV) reflects the size of price movement priced into the option. This run included all IV environments, with no IV Rank filter.

One fully secured SPY put at a time, no overlapping positions. Three exits:

These are profit targets, not stop-losses. They close winners early. They don’t cut losers. Every losing trade in all three runs stayed open until expiration.

Which exit made the most money?

The 50% exit made the most money: $15,880.75 gross, $4,169.45 more than the 75% exit and $5,699.25 more than holding to expiration.

Source: Sqilled Strategy Backtester, run bf6772290f25. One SPY cash-secured put per trade, 17 delta, 45 DTE, all IV environments, January 7, 2005 to October 16, 2026. Gross option P&L before commissions and slippage.

Across 222 SPY cash-secured puts from 2005 to 2026, Sqilled’s backtest found that closing at 50% of the credit earned 56% more gross profit than holding every put to expiration.

Figure 1. Cumulative gross option P&L as trades closed. This is not a daily account equity curve.

Chained return compounds each trade’s profit against its cash collateral. Holding to expiration scores slightly higher there with fewer trades, but earned the fewest dollars.

Why does February 2020 matter so much?

February 2020 matters because one trade created an $8,014 gap between the exits, more than the 50% rule’s entire lead over holding.

On February 4, 2020, all three runs sold the same SPY $310 put expiring March 20 for a $212 credit.

• The 50% rule closed it eight days later for a $106 gain.
• The other two held it into the March selloff and lost $7,908 at expiration.

Exit timing also changed which trades came next. In April 2022, the 50% run skipped a new put on April 5 because its March 1 position was still open. The hold-to-expiration run took that April trade and lost $2,369.

So the 50% rule didn’t win on one trade alone. It won on a different sequence of trades. For a longer SPY comparison that includes covered calls, see our Wheel Strategy backtest.

What did the losing trades look like?

The losing trades were few but large. The 50% run had 8 losers out of 222 trades, and they cost $5,511.50 combined.

In this sample, the average loss was nearly seven times the average win. That’s why position size and the treatment of losing trades matter as much as win rate.

Figure 2. Annual gross option P&L by close year. 2026 covers January to October 16.

The backtest covers 262 calendar months. Trades closed in 212 of them: 204 profitable months and 8 losing months. The other 50 months had no closed trade.

The October 2008 loss of $2,373 was about 20.1% of that position’s $11,800 collateral. The backtest’s maximum drawdown was about 22.5%, measured on closed trades. A real account marked to market daily could have experienced deeper drawdowns while positions were open; those drawdowns were not measured here.

How much money do you need to sell a SPY put?

You need enough cash to buy 100 SPY shares at the option strike. In the September 2026 backtest trade, the put had a $720 strike, so one fully secured contract required $72,000 in collateral. That is a strike-based obligation, not a claim about SPY’s market price that day.

The last trade in the 50% run opened September 1, 2026. It sold a SPY $720 put for a $450 credit and closed September 21 for a $225 gain. Fully securing it took about $72,000.

• A $25,000 account couldn’t fully secure this particular $720-strike SPY put.
• A $100,000 account would commit about 72% of its cash to this one position.
• The credit is the most you can make. The most you can lose is the strike times 100, minus the credit.

Smaller accounts often apply the same rules to lower-priced stocks or ETFs. For the mechanics, see the Options Industry Council’s cash-secured put guide.

Before you sell a put, ask yourself

1. Could I afford to own 100 shares at this strike?
2. How much of my account does one contract tie up?
3. What’s my exit rule before I enter: 50%, 75% or expiration?
4. What’s my plan if the trade goes against me?

Methodology and limits

• Sample: 222 trades (50% exit), 189 (75% exit) and 130 (held to expiration). SPY only, one contract at a time, no overlap, all IV environments, January 7, 2005 to October 16, 2026. Run ID bf6772290f25.
• Gross results: before commissions and slippage. With commissions, the 50% result is $15,603.20, still before slippage and bid-ask costs.
• Assignment: losses are booked at expiration. The test doesn’t hold assigned shares or sell covered calls afterward, so it is not a wheel strategy backtest.
• Return on cash: the 54.3% figure chains simulated trade-by-trade returns against strike-based cash collateral. Its mathematical annualized equivalent over the backtest window is about 2.0%, not a realized whole-account return. Interest potentially earned on reserved cash was not modeled.
• Data: the trade logs reconcile with the monthly exports, but historical option quotes and modeled fills were not independently validated against a second source. Earlier Thinkorswim spot checks identified unresolved pricing or delta differences. January and February 2005 expiry cycles lacked qualifying data. A historical-data correction changed the 50% run from 217 trades to 222.
• Not tested: IV Rank filters, stop-losses, rolling, other tickers, and early assignment.

All backtest results are hypothetical.

FAQs

The takeaway

Selling puts can stack up a lot of small wins. The exit rule decides how many of those wins you keep. In 22 years of SPY data, taking profits at 50% beat holding to expiration by $5,699.25, largely by getting out before one bad month.