There’s a version of FIRE that sounds almost too clean. You hold your index funds, you sell options on a basket of quality stocks, and every month a small income arrives — without touching your principal, without selling anything, without depending on the market going up. It’s financial independence with a side income, and the logic is seductive enough that entire communities have been built around it. Forums, YouTube channels, Discord servers, paid courses. The promise is consistent enough that people treat it as settled.
We wanted to know if it actually worked. Not in theory — the theory is fine and we’ll come to it — but in practice, with real stocks, real market cycles, and the actual parameters a systematic trader would use. So we built the system from scratch: rules engine, risk management, position sizing, parameter optimisation, and a full backtest against six years of daily price data across 20 carefully chosen companies.
Then we looked at the numbers without rounding them in the wrong direction.
The results are not what the options-income community usually publishes. They’re also not a reason to stop. That tension is what this post is about.
The received wisdom
The case for options income as a FIRE strategy is straightforward and, to be fair, not entirely wrong. When you sell a put option, you collect a premium in exchange for agreeing to buy a stock at a lower price if it falls below a certain level. If the stock stays flat or rises, you keep the premium and repeat. If you get assigned — if the stock falls and you end up owning it — you sell covered calls to collect more premium while you wait for a recovery. The whole cycle is called the wheel strategy, and it’s been practiced by professional traders and retail investors alike.
The argument is compelling on paper. Implied volatility — the market’s expectation of future swings — tends to be systematically overpriced relative to what actually happens. Academics call this the volatility risk premium, and it’s one of the most consistently documented anomalies in options markets. Selling premium is, in aggregate, a positive-expected-value business. Studies going back to Bondarenko (2003) and the work of Robert Whaley on the CBOE’s BuyWrite Index confirm that selling covered calls has historically generated modest outperformance over holding stocks alone, especially on a risk-adjusted basis.
So the theory holds. The question is what happens when you put real parameters on it, apply real stocks, and test it through a real market cycle — including 2020, 2022, and everything in between.
What the backtest actually showed
We built a simulation engine using six years of historical price data (2019–2024) across 20 carefully selected stocks: a mix of technology, energy, consumer staples, healthcare, financials, industrials, and one REIT. The strategy used protected puts — each short put paired with a long put below it to cap the maximum loss — with the exit rules calibrated through a grid search across 180 parameter combinations.
The best parameters across the entire period:
| Parameter | Optimal value |
|---|---|
| OTM distance | 8% below current price |
| Profit target | Close at 50% of premium collected |
| Stop loss | Close if spread value reaches 2.5× premium collected |
| Protection width | 1.0 standard deviation |
With those parameters, here is what 770 trade cycles produced:
| Metric | Result |
|---|---|
| Win rate | 84% |
| Average win | +$35 per contract |
| Average loss | -$120 per contract |
| Portfolio return | +0.2% per year on $314,000 deployed¹ |
| Average monthly income | $72 gross (before bid-ask spread costs)² |
| Max drawdown | -$2,013 (0.6% of capital) — synthetic IV model, full 2019–2024³ |
| Ruin protection triggered | Never |
¹ $314,000 = sum of cash collateral required for one put spread per stock across all 20 positions at 8% OTM. ² Gross premium collected at mid price. Net after estimated bid-ask spread costs: approximately $57/month. ³ The -$2,013 max drawdown is the peak-to-trough decline in cumulative portfolio P&L over the full six-year period, measured monthly. The same drawdown figure appears within 2022 specifically — the intra-year peak-to-trough in 2022 was also -$2,013, which is why 2022 drove the full-period maximum. The full-year 2022 P&L was -$947 (losses partially recovered in the second half).
The 84% win rate is real and consistent with what the volatility risk premium research predicts. The Sharpe ratio of 0.66 reflects the strategy’s risk-adjusted profile — positive, but modest. The maximum drawdown of just 0.6% of deployed capital is remarkably contained, a direct result of the protection legs doing their job in stressed markets.
But the return. $72 a month. On $314,000.
That is not a typo.
Why the numbers are so low — and why the IV model matters more than you’d think
The gap between the theory and the result comes down to a structural limitation in how we priced the options. Because we don’t have historical options chains — that data costs hundreds of dollars a month from providers like CBOE DataShop or OptionMetrics — we estimated option prices using Black-Scholes with implied volatility derived from realised price swings. This almost certainly understates real market IV by 30–50%. The actual volatility risk premium is larger than our model captures.
We ran the full backtest twice: once with that synthetic IV estimate, and once using the VIX — the market’s actual fear gauge — calibrated per-stock. The results don’t just differ in magnitude. For 2022, they tell opposite stories.
| Period | Synthetic IV (annual P&L) | VIX-based IV (annual P&L) |
|---|---|---|
| Full 2019–2024 | $+683/yr | $+568/yr |
| Early period 2019–2021 | $+998/yr | $+470/yr |
| Late period 2022–2024 | $-394/yr | $-135/yr |
| 2022 alone | -$947 | +$639 |
| 2023 alone | -$345 | +$1,024 |
| 2024 alone | +$1,178 | -$829 |
All rows use the same parameters optimised on the full 2019–2024 period (8% OTM / 2.5× stop / 50% profit target). This is a period split, not a walk-forward: every figure reflects hindsight-optimal parameters applied to a date window, not out-of-sample validation. The individual calendar-year rows are approximate — 14 trades spanning December/January boundaries are attributed differently when the continuous simulation is sliced into yearly windows, which is why summing 2022+2023+2024 does not exactly equal the Late period aggregate. The combined Late period figure (-$394/yr) is the more reliable measure. For a genuine out-of-sample comparison: running the strategy on 2022 data in isolation — starting fresh on 1 January with no carry-over trades — produced -$3,321 under the synthetic IV model, versus -$947 in the continuous run. The ~$2,400 gap is a direct measure of how much the boundary-trade effect and carry-over positions from 2021 cushioned the 2022 result in the continuous run.
The 2022 result is the most important number in this table. In 2022, the VIX peaked above 35 at multiple points as the Fed began its fastest rate-hiking cycle in forty years. Our synthetic IV model — built from backward-looking realised volatility — was still catching up when the market was already pricing in serious fear. Under the synthetic model, the portfolio lost -$947 for the full year 2022. Under the VIX-based model, which captured the fear spike immediately, 2022 was profitable at +$639.
This is exactly what the volatility risk premium theory predicts. When fear is elevated, the premium available to sellers is highest. A model that uses actual market fear data earns more in exactly the months when an equity portfolio is bleeding. A model that uses backward-looking estimates misses the spike.
Neither model is definitive — both are approximations of what real options chains would show. But the 2022 divergence illustrates why IV estimation methodology matters as much as strategy design. We’ll keep both models in the framework and let paper trading reveal which is closer to reality.
For comparison, the same $314,000 in bonds at 4.5% would generate $1,178 a month regardless of which IV model you use.
The per-stock picture — who actually pulls weight
The aggregate conceals something important. Not every stock in the basket performs equally, and the spread tells you something useful.
Strong performers (2019–2024):
| Stock | Total P&L | Win rate |
|---|---|---|
| AMZN | +$828 | 82% |
| XOM | +$472 | 91% |
| ABBV | +$553 | 92% |
| JPM | +$391 | 87% |
| PEP | +$360 | 94% |
| PG | +$301 | 96% |
Underperformers:
| Stock | Total P&L | Win rate | Note |
|---|---|---|---|
| LOW | -$317 | 68% | Housing cycle headwind |
| DDOG | -$177 | 80% | High IV but volatile swings |
| V | -$94 | 78% | Low IV, thin premiums |
| KO | -$68 | 86% | Very low IV — barely worth running |
XOM’s $472 total P&L over six years with a 91% win rate is exactly what this strategy is supposed to look like. Energy stocks carry elevated implied volatility even in calm periods, which means richer premiums for the same OTM distance. Defensive consumer staples like KO generate almost nothing — the IV is simply too low to produce meaningful income at safe distances.
(A small aside worth noting: we screened every stock in this basket against recent earnings reports before including it. Nike — which appears on most “best wheel stocks” lists — was excluded because its direct sales revenue had been declining for five consecutive quarters as of Q3 FY2026. The wheel strategy requires conviction that you’d hold the stock if assigned. Not every high-IV stock deserves your collateral.)
Where the strategy genuinely earns its place
Here is the honest version of why you might still run this, despite the modest return numbers.
Bonds generate more income per dollar of capital. The S&P 500 generates more total return. Neither of those facts makes the wheel strategy worthless — they just clarify what it’s actually for.
The primary value isn’t the absolute income. It’s counter-cyclicality and sequence protection. When markets fall sharply, implied volatility spikes. The portfolio’s worst single month across the full six-year period was -$1,479, recorded in the synthetic IV model during the 2022 bear market. That loss was still a fraction of what an unhedged equity position would have suffered in the same period. More importantly, the IV spike that drove that loss also meant the following month’s entry collected fatter premiums than any month in 2021. The strategy pays more precisely when equity portfolios are hurting most.
The wheel doesn’t replace your VUSA position. It makes your VUSA position easier to hold.
For a FIRE plan built around not selling index funds during downturns, that dynamic has real value. If your monthly expenses are covered by options income, bonds, and dividends combined — even in a year when equities fall 30% — you never become a forced seller. This is consistent with what Bengen’s 1994 framework identified as the primary driver of retirement ruin: being forced to sell depressed assets early in retirement, not the average return over time.
What this actually means
The wheel strategy, run systematically on a diversified basket of quality stocks, is a modestly positive business. It won’t fund your retirement on its own. It probably won’t beat bonds in a normal rate environment. What it will do — if you build and run it with discipline — is generate uncorrelated monthly cash flow that improves the resilience of a FIRE plan built primarily on equity compounding.
The right question isn’t “can options income replace my salary?” It’s “does this improve my probability of never being forced to sell VUSA at the bottom?” The backtest suggests yes, with a caveat: the improvement is real but modest, and only if you hold through the losing months without abandoning the system. The worst month in the simulation (-$1,479) was followed by four consecutive winning months. The investors who quit at the bottom missed the recovery.
The period split adds another caveat worth naming directly. The parameters we optimised on the full 2019–2024 period — including 2022 itself — still produced negative results in the late-period split (-$394/yr on the synthetic model). These are hindsight-optimal parameters that saw every market regime during calibration, and they still lost money in 2022–2024. That is worth sitting with. It does not mean the strategy is broken; it means the numbers you see in a backtest reflect the environment that data was drawn from. Fixed parameters will always have regime-dependent years. The right response isn’t to retune after every bad year — that’s overfitting in a different disguise — but to accept that any systematic strategy will underperform in regimes it wasn’t designed for.
The protection leg is what makes this acceptable. Even in 2022, even with the synthetic IV model showing a full-year loss of -$947, the intra-year peak-to-trough drawdown reached only -$2,013 — the same figure as the full six-year maximum drawdown, and 0.6% of deployed capital. The strategy lost money that year. It didn’t blow up.
There’s also a specific situation where the risk/reward flips decisively in your favour: when you already own a stock and want to sell it at a target price. Selling a covered call at your exit target means you collect premium while waiting, and if the stock reaches your price, you sell it exactly where you wanted to. The wheel doesn’t add risk here — you were going to sell anyway. It just pays you for the waiting period. For anyone holding a concentrated position they want to exit gradually, this is genuinely excellent risk/reward. Not the systematic basket approach. That specific application.
The honest summary: as a standalone income strategy, the wheel underperforms bonds on the same capital. As a sequence-of-returns buffer layered on top of a bond allocation and an equity compounding engine, it earns its place. The framing matters more than the headline return number.
If you’re building toward financial independence, use our FIRE Calculator to model what your withdrawal rate looks like when you layer a modest options income stream on top of an existing portfolio. The interaction between variable options income and a bond floor — especially during the first five years of retirement when sequence risk is highest — changes the survival probability more than the income amount alone would suggest.
For the genuinely curious: the full backtest engine, parameter optimizer, and all 770 simulated trade logs are part of an open framework we’ll be documenting over the coming months. The code is deterministic, the data is from public sources, and the results are reproducible.
This article is for general educational purposes and does not constitute financial advice. Options trading involves significant risk of loss and is not appropriate for all investors. The backtest results presented here use synthetic implied volatility derived from historical price data, which likely understates actual market premiums — results with real options chains may differ materially in either direction. Past simulation results are not indicative of future performance. For advice tailored to your circumstances, consult a qualified financial adviser.