| The Short Answer I tested two strategies on SPY for 21 years using real options data: buying 50-delta calls at 45 DTE, and selling 17-delta puts at 45 DTE. Same entry timing. Same exit targets (25%, 50%, hold to expiry). 212 trades each. Long Calls (50Δ, hold): 49.5% win rate, $122 average per contract, 20.6% return on capital. Short Puts (17Δ, hold): 90.6% win rate, $105 average per contract, 0.40% return on capital. Same dollar P&L. 41 percentage point gap in win rate. 44x difference in capital required. The IV filter (entries above vs below 18% ATM IV) flips the optimal exit on long calls and meaningfully improves short puts. Full breakdown below. |
Most retail option traders start by buying calls. It feels intuitive. You think the market is going up, you buy a call, you cap your loss at the debit and your upside is unlimited. The math works on paper.
But after a few years of doing it, you notice your account is not growing the way the trades make it sound like it should. You are right on direction more often than not, you have your big winners, and somehow the cumulative number is still flat or red.
I have heard this story from hundreds of traders. The question is whether the data actually backs up the alternative people keep talking about: that selling premium does better than buying it, especially for retail accounts.
So I tested it. Twenty one years of monthly SPY options data, every cycle from December 2005 through March 2026. Two strategies, run side by side. One contract per trade. No stops. No selective entries. Just the rules, applied month after month, for two decades.
What follows is what the data actually shows.
Part 1: The Setup
The point of a backtest like this is that the comparison has to be honest. You cannot run buying calls in one set of conditions and selling puts in another. Same dataset, same entry day, same exit options. The only thing that varies is which side of the trade you are on.
What I tested
Strategy A: Long Call
- Buy a 50-delta SPY call at 45 days to expiration
- Hold until one of three exits: +25% on the debit, +50% on the debit, or expiration
- No stop loss
- Wait for the next monthly cycle to re-enter
- One contract per trade
Strategy B: Short Put
- Sell a 17-delta SPY put at 45 days to expiration (cash secured)
- Hold until one of three exits: 25% of credit captured, 50% of credit captured, or expiration
- No stop loss
- Wait for the next monthly cycle to re-enter
- One contract per trade
I also added an IV filter. At each entry, I recorded the implied volatility of the 50-delta call (a clean proxy for at-the-money IV on SPY). Trades where ATM IV was below 18% went into the “low IV” bucket. Trades at or above 18% went into “high IV”. This is the same regime split I use in the [AAPL backtest](https://sqilled.co/backtest-analysis-and-findings/) and the same threshold most volatility-aware retail systems use.
What buying power means here
This is the part that throws people off when they first see backtest comparisons between long premium and short premium. They are not the same trade in terms of capital.
For the long call, buying power is the debit you pay times 100. Across 212 trades the average was about $592 per contract. That is what you risk and that is what is tied up.
For the short put on a cash-secured basis (the way 99% of retail accounts run them, especially on margin-restricted accounts and IRAs), buying power is the strike price times 100. Average across 212 trades: about $26,118 per contract. That is what your broker holds aside in case of assignment.
Short puts use roughly 44 times more capital per trade than long calls. Both numbers are in the data and you cannot ignore that gap when you compare them, which is why I report both raw dollar P&L and return on capital throughout.
Part 2: The Headline Numbers
Here is the chart that tells you what you need to know in about six seconds. Win rate on the left. Total dollar P&L per contract on the right. All six strategy variants on the same axis.
Win rate and total P&L per contract across all 212 trades, December 2005 to March 2026.
What jumps out
The win rate gap is the first thing. Short puts win 90.6% of the time when held to expiration. Long calls win 49.5% of the time. That is essentially a coin flip on whether you make money buying premium, versus 9 out of 10 winners selling it.
But the total dollar P&L is closer than you would think. Long call hold-to-expiry made $25,795 over 21 years per contract. Short put hold-to-expiry made $22,246. Long calls actually edge out short puts on raw dollar P&L because their winners are huge. The average winning long call paid $738, while the average winning short put paid $173.
So if dollar P&L is roughly equal, what actually separates them? Two things: variance and capital.
The variance story
Standard deviation per trade tells you how rough the ride is. For long calls held to expiry, it is $818. For short puts held to expiry, it is $320. Roughly 2.5 times more swing per trade on long calls.
In practice, this means the long call account has months where you are up $3,000 on a single trade and months where you are down $1,400. The short put account is collecting $50 to $200 most months with the occasional drawdown when SPY actually moves against you. Over 21 years the totals are similar. The way you live through them is not.
The capital story
Now run the same numbers on a return-on-capital basis. Long call hold-to-expiry: 20.6% ROC over 21 years. Short put hold-to-expiry: 0.40% ROC.
That is not a typo. The long call returns 50 times the percentage on capital because it uses 44 times less capital per trade. If you have a $50,000 account and you are buying one $592 call per month, you are using barely 1% of buying power. If you are selling one $26,000 cash-secured put per month, you are using more than half.
This is why naive comparisons of “% return” between long premium and short premium are misleading. The strategies are not competing for the same dollar.
| Where this all fits together The setups, exits, and IV regimes here are pieces of a larger system I walk through in my free masterclass: how to choose between buying premium and selling it based on your account size, your goals, and what the market is actually paying you on any given day. If you want the full framework, you can register here: https://sqilled.co/free-masterclass/ |
Part 3: Year by Year
Aggregates hide the journey. Here is what each year looked like for both strategies, hold-to-expiry, per contract.
Annual P&L per contract, hold-to-expiry, both strategies. The 2022 selloff is the biggest single-year drawdown in the dataset for long calls.
A few things worth pointing out
Short puts (the navy bars) are positive in 16 of 21 years. They lose money in five years: 2008, 2018, 2020, and 2022 are the meaningful ones. Even in those years, the worst annual P&L was negative $995 per contract (2022). Compare that to the long call worst year, which was negative $7,763 in 2022.
Long calls (the orange bars) are wildly variable. The best year is 2023 at $8,083 per contract. The worst is 2022 at negative $7,763. The best year is a 16x multiple of the worst year in absolute terms. For short puts, the best year (2021 at $3,540) is only 3.5x the worst year (2022 at negative $995).
This is what consistency actually means in practice. Not the absence of losing trades, but the absence of catastrophic drawdown years that make you stop following the strategy.
What happened in 2008?
Worth addressing directly because it comes up. In 2008 there is only one trade in the dataset (the data coverage starts late and the 2008 cycle that fit my filter was October entry, November expiration). The short put on that entry lost $315. The long call lost $568. Both lost money. In 2009 the recovery year, short puts made $459 across 8 trades. Long calls were essentially flat at negative $78. The short put strategy adapted faster because high IV (which is exactly what 2008 produced) is what short puts feed on.
I wrote about this dynamic at length in [the wheel strategy backtest](https://sqilled.co/wheel-strategy-backtest-spy/), where the same pattern shows up: high IV regimes are not the enemy of the put seller, they are the friend.
Part 4: The IV Filter Does Real Work
I split the 212 trades into two buckets. Trades entered when ATM IV was below 18%: 155 trades (73% of the sample). Trades entered when ATM IV was at or above 18%: 57 trades (27%). Then I re-ran the same six strategy variants inside each bucket.
The result is one of the more interesting findings in this study.
Return on capital, low IV vs high IV regime, for the best long call and short put exit variants.
Long calls reverse based on IV regime
In low IV (< 18% ATM IV), the best long call result is hold to expiry: 23.2% ROC. The 25% target only delivers 5.8% ROC. When premium is cheap, you need the right tail to pay for the losers, and capping winners at +25% destroys the math.
In high IV (≥ 18% ATM IV), it flips. The 25% target wins at 17.8% ROC versus hold-to-expiry at 15.5%. When premium is fat, time decay accelerates and the option is more likely to give back its early gains. Taking profits early protects them.
This is exactly what experienced premium traders mean when they say “exit rules depend on the regime.” It is not opinion. It is what the data shows when you actually segment the trades.
Short puts get unambiguously better in high IV
For short puts, high IV improves every metric. The hold-to-expiry ROC jumps from 0.31% in low IV to 0.70% in high IV. The 50% target jumps from 0.24% to 0.62%. Same 90% win rate in both regimes.
The reason is straightforward. When IV is elevated, you collect more credit per put for the same delta. That extra credit is both more profit when you win and more cushion when you lose. The cushion piece matters more than people realize.
The tail risk finding worth talking about
I pulled the five worst short put losses in the entire 21-year dataset. Here is what I found.
| Date Entered | ATM IV | IV Regime | P&L |
| Apr 2022 | 17.7% | Low | -$2,390 |
| Feb 2020 | 14.0% | Low | -$2,322 |
| Nov 2018 | 17.7% | Low | -$1,409 |
| Jan 2022 | 13.8% | Low | -$915 |
| Dec 2015 | 13.3% | Low | -$867 |
All five worst short put losses across 21 years came from low IV regime entries. Worst high-IV loss was -$514.
Every single one of the five worst short put losses came from a low IV entry. The April 2022 trade entered at 17.7% IV (right under the threshold) and got destroyed when SPY sold off into June. The February 2020 trade entered at 14% IV right before COVID and got run over.
In the high IV regime across 57 trades, the worst loss was negative $514. That is it. The fat premium served as a real cushion against vol expansion.
This is not just a “more premium is good” story. It is a structural finding. Selling puts when IV is already elevated means a) you get paid more, and b) when the move you were worried about actually happens, your loss is smaller because the credit absorbed it. The asymmetry runs in your favor twice. If you want a deeper read on this, [my SPY implied vs realized volatility post](https://sqilled.co/spy-implied-vs-realized-volatility/) shows why selling premium when implied is high tends to converge favorably.
Part 5: What This Means For You
The data does not say “always sell puts.” It does not say “never buy calls.” It says different things to different traders depending on what you are optimizing for and how much capital you have.
If you have a small account
Small accounts cannot afford to tie up $26,000 per trade in a cash-secured short put. A $5,000 account literally cannot run the cash-secured short put strategy on SPY at all. Two options exist: you trade short puts on a smaller underlying (which is fine, the same dynamics apply), or you accept the higher variance and trade long calls.
If you are going to buy long calls with a small account, the data is clear: take them in high IV with a 25% target, and hold them to expiration in low IV. Do not run a single fixed exit rule across both regimes. The optimal exit literally inverts. If you want a full breakdown of the long call setup, I wrote [a detailed long call options strategy guide](https://sqilled.co/long-call-options-strategy-guide/) that walks through the entry mechanics in more depth.
If you have an income mandate
If your goal is to generate consistent monthly cash flow on capital you already have, short puts win. The 90% win rate, the low variance, and the ability to scale by adding contracts on accounts that can support multiple cash-secured positions, all of it favors the income trader.
The trade-off you are accepting is opportunity cost. That same $26,000 per put could be in SPY shares earning closer to 12% per year on capital appreciation. The short put earns about 0.40% per cycle on that capital, or roughly 4-5% annualized. Whether that trade-off is worth it depends entirely on whether you value cash flow predictability over total return.
There are also defined-risk versions of selling premium that use less capital. I covered one of them in [the sleep-at-night iron condor post](https://sqilled.co/the-sleep-at-night-iron-condor-wider-slower-safer/) which trades some of the win rate for dramatically lower capital usage.
If you want to use both
This is what I actually do. The two strategies are not mutually exclusive. Short puts in high IV regimes for the income side. Selective long calls in either regime for asymmetric upside when conviction is high. The IV filter helps you decide which lever to pull.
The framework that ties this together (when to buy, when to sell, how to size, when to step aside) is exactly what I cover in the free masterclass. One hour, completely free, walks through the same data and the decision rules I run.
FAQs
Why 50 delta on the long call and 17 delta on the short put?
These are the most common entries for each strategy among retail traders, and they make the comparison representative of what people actually trade. 50-delta calls are at-the-money, which is the standard “directional bet” entry for buying premium. 17 delta on a short put is the standard high-probability entry that targets approximately 83% probability of profit at expiration.
Why 45 days to expiration?
It is the sweet spot for monthly options. Short enough that theta decay is meaningful, long enough that you are not getting whipsawed by daily noise. The same 45 DTE entry is used in [my AAPL backtest](https://sqilled.co/backtest-analysis-and-findings/) and [the wheel strategy backtest](https://sqilled.co/wheel-strategy-backtest-spy/). Standardizing on it makes the comparisons across studies clean.
Are these results net of commissions?
No. Add roughly $1.30 per contract round-trip for typical retail commission rates. Across 212 trades that is $276 per strategy variant. It does not change the rank order of results but it does compress the long call ROC by about 1% and the short put ROC by about 1%.
What about slippage?
I priced both entries and exits at mid-price (the average of bid and ask). Real fills will typically be slightly worse, especially on illiquid strikes. SPY is the most liquid options chain in the world, so the gap is usually a penny or two. On 100 multipliers that adds up over 212 trades, but again, the rank order does not change.
Why did you cap the long call upside at 25% and 50% rather than 100% or 200%?
I tested the standard exit targets that retail traders actually use. A 100% target is fine for biotech catalyst plays, but on monthly SPY options it almost never hits inside 45 days unless there is a major move. The hold-to-expiry variant captures the right tail naturally because it is willing to ride the option to its full intrinsic value at expiration.
Could you have done better with a stop loss?
On the long call side, a stop at 50% of the debit would cap losses at about $300 per losing trade instead of $400+. But it would also cut off some recoveries. I tested this informally and the net effect was small. On the short put side, a 2x credit stop would have prevented the worst three losses (-$2,390, -$2,322, -$1,409) but it would also have triggered on recoveries that ended up profitable. The user explicitly chose no stops to keep the comparison clean.
What was the worst single-trade loss?
Long call: negative $1,387 on the May 2022 entry. The entire debit was lost when SPY dropped through the strike and stayed there. Short put: negative $2,390 on the April 2022 entry. SPY dropped from $452 at entry to about $377 at expiration, and the 17-delta put was deep in the money at exit.
Why does the short put have a bigger max loss than the long call?
Because the short put has unbounded loss exposure (the strike can go to zero), while the long call has bounded loss at the debit. This is the classic asymmetry of selling premium. You collect a known credit and accept an unknown maximum loss. In practice for cash-secured puts, the max loss is the strike times 100 minus the credit, which is a large but finite number.
What if I roll instead of taking the loss?
Rolling is a separate strategy with its own backtest. The wheel strategy backtest I published handles assignment by transitioning into covered calls instead of taking the put loss, and the results are different. For this study I wanted a clean apples-to-apples on entry/exit/expiration, so no rolling.
Why is the data only 212 trades over 21 years if there are 12 monthly cycles per year?
Roughly 252 monthly expirations exist in the dataset, but the entry filter (find a contract closest to 45 DTE with a 40-50 day window) plus quality checks (delta within 10 points of target, non-zero pricing on entry) trims out a few cycles per year, especially in the early years when monthly options coverage was sparser. 212 is the clean sample.
Where did the SPY options data come from?
Historical end-of-day options chains, the same data source I use across all my [research at sqilled.co](https://sqilled.co/). Every entry, exit, and IV figure in this article is computed from real bid/ask data, not from a model.
Is this enough sample size to be statistically significant?
For directional strategies on broad indices, 212 trades is a solid sample. The win rate confidence interval at 90.6% across 212 trades is roughly ±4 percentage points at 95% confidence. The high-IV bucket (57 trades) is smaller and the conclusions there should be treated with appropriate humility, but the direction of the results is consistent enough across both regimes that the qualitative findings hold up.
What’s Next
Backtests like this answer a narrow question well: which of these two strategies has worked historically? But applying the answer to your account is a different question. It depends on your capital, your goals, and what the market is paying you right now.
I run a free 90-minute masterclass that walks through the framework I use to decide between buying premium and selling premium on any given week. Same data you saw above, but turned into a decision system you can apply Monday morning.
| Free Masterclass: How to Read the Market and Pick the Right Side Live, 90 minutes, completely free. Walks through the entry framework, the IV filter, the position sizing rules, and the management decisions for both buying premium and selling it. Open to anyone who is past the first 6 months of trading and wants a real system instead of guessing. Reserve your seat: https://sqilled.co/free-masterclass/ |
See you in there.
Addy
