Study 9: Which Options Strategies Actually Work — A 20-Year, 404-Ticker Regime Study
We backtested 12 options strategies across all S&P 500 companies from 2005 to 2025, tagging every trade with four regime dimensions: VIX bucket, market trend, rate environment, and NBER recession. Calendar spread is the only strategy that works in every regime. Here is exactly when to run each one.
The Question
Most options education teaches you what a strategy is. Very little of it tells you when each strategy actually makes money — and when it bleeds.
That distinction matters more than people realize. A short put strategy with an 80% win rate over 20 years still loses money in a flat, low-VIX market. An iron condor with a 72% win rate has a negative Sharpe. Calendar spread — the quiet, unsexy trade — has the highest risk-adjusted return of any strategy we tested, in every single market regime.
This study ran 1,082,172 synthetic options trades across 404 S&P 500 tickers from 2005 to 2025. Every trade is tagged with four regime dimensions: VIX bucket, market trend, rate environment, and NBER recession status. The goal is a decision rule — not “should I trade options,” but “given today’s regime, which structure gives me the best risk-adjusted edge?”
Study Design
Strategies Tested
All 12 strategies use standard options mechanics with synthetic pricing — Black-Scholes implied volatility estimated from realized volatility with a 1.20 volatility risk premium (reflecting the typical spread between realized and implied vol in equity markets). Entries are placed approximately 30 days before each monthly expiration date. Commission is $0.65 per contract per leg, round-trip.
| Strategy | Structure | DTE | Key Parameter |
|---|---|---|---|
| Covered call | Long stock + short OTM call | 30 | 0.30 delta call |
| Protective put | Long stock + long OTM put | 30 | 0.30 delta put |
| Collar | Long stock + short call + long put | 30 | 0.30 delta both legs |
| Short put | Naked short OTM put | 30 | 0.30 delta |
| Bull call spread | Long ATM call + short OTM call | 30 | 0.50 / 0.30 delta |
| Bear put spread | Long ATM put + short OTM put | 30 | 0.50 / 0.30 delta |
| Long straddle | Long ATM call + long ATM put | 30 | 0.50 delta both |
| Short strangle | Short OTM call + short OTM put | 30 | 0.25 delta both |
| Iron condor | Short strangle + long wings | 30 | 0.25 / 0.10 delta |
| Iron butterfly | Short ATM straddle + long wings | 30 | 0.50 / 0.25 delta |
| Calendar spread | Short front-month + long back-month at same strike | 30/60 | ATM |
| Wheel | Short put → covered call cycle | 30 | 0.30 delta |
Regime Classification
Four dimensions are applied independently to every trade:
VIX Bucket — measured at trade entry:
- LOW: VIX < 15
- MID: 15 ≤ VIX < 25
- HIGH: 25 ≤ VIX < 35
- CRISIS: VIX ≥ 35
Market Trend — SPX position relative to its 200-day moving average:
- BULL: price > 2% above MA200 with positive slope
- BEAR: price > 2% below MA200 with negative slope
- SIDEWAYS: everything else
Rate Regime — 3-month T-bill 6-month change:
- RISING: +50 bps or more
- FALLING: −50 bps or more
- FLAT: within ±50 bps
Recession — NBER recession dates (2007–2009 financial crisis, Feb–Apr 2020 COVID recession).
Overall Results
The 20-year headline tells a clear story. High win rates are often meaningless. The strategies with the highest win rates — wheel (73%), short put (80%), iron condor (72%), short strangle (75%) — cluster at the bottom half of the Sharpe table or go negative entirely.
| Strategy | Sharpe | Win Rate | Mean Return | Total P&L |
|---|---|---|---|---|
| Calendar spread | 1.17 | 56.1% | 35.8% | $3,719,030 |
| Collar | 0.67 | 59.2% | 1.1% | $5,929,314 |
| Covered call | 0.54 | 65.4% | 1.3% | $6,443,813 |
| Protective put | 0.51 | 50.2% | 1.3% | $6,692,527 |
| Bull call spread | 0.51 | 45.8% | 19.5% | $1,784,017 |
| Wheel | 0.25 | 73.1% | 0.5% | $1,679,130 |
| Short put | 0.21 | 80.1% | 0.3% | $927,982 |
| Long straddle | 0.21 | 42.7% | 8.5% | $50,591 |
| Short strangle | −0.20 | 75.3% | −0.6% | −$1,722,462 |
| Iron condor | −0.36 | 72.4% | −5.4% | −$847,191 |
| Iron butterfly | −0.41 | 31.3% | −20.8% | −$1,020,431 |
| Bear put spread | −0.85 | 33.4% | −25.7% | −$3,417,412 |
Two things stand out immediately.
The win rate trap: iron condor wins 72% of trades over 20 years and still loses money. This is the defining feature of premium-selling in low-volatility environments — you collect small credits consistently until a volatility event arrives and wipes out months of gains in a single expiration.
The bear put spread collapse: a strategy that should theoretically protect in downturns has the worst Sharpe of the study at −0.85. The reason is simple — 20 years of S&P 500 history is structurally bullish. Bear put spreads expire worthless in uptrending markets. The very few months where they pay off (CRISIS VIX regime) don’t compensate for the continuous theta bleed during the other 92% of months.
VIX Regime: The Strongest Signal
VIX is the most actionable regime dimension. It explains more variance in strategy performance than market trend, rate environment, or recession status combined.
Sharpe by Strategy × VIX Bucket
| Strategy | LOW (<15) | MID (15–25) | HIGH (25–35) | CRISIS (>35) |
|---|---|---|---|---|
| Calendar spread | 1.22 | 1.11 | 1.16 | 1.51 |
| Short put | −0.03 | 0.02 | 1.76 | 0.75 |
| Covered call | 0.36 | 0.35 | 1.69 | 0.71 |
| Collar | 0.66 | 0.54 | 1.31 | 0.57 |
| Wheel | 0.03 | 0.07 | 1.32 | 0.58 |
| Bull call spread | 0.48 | 0.42 | 1.08 | −0.08 |
| Long straddle | 0.44 | 0.29 | −0.01 | −0.60 |
| Iron condor | −0.56 | −0.40 | 0.04 | 0.22 |
| Short strangle | −0.42 | −0.28 | 0.24 | 0.51 |
| Bear put spread | −0.79 | −0.68 | −1.86 | −0.88 |
| Iron butterfly | −0.47 | −0.43 | −0.33 | −0.42 |
| Protective put | 0.60 | 0.43 | 0.91 | 0.02 |
The key observations:
Calendar spread is the only strategy above Sharpe 1.0 in every VIX bucket. It benefits from elevated VIX (higher time premium on the near-month leg) without the catastrophic risk of naked premium selling when realized volatility spikes. In CRISIS regimes it hits its best Sharpe at 1.51, because volatility term structure steepens and the long back-month position benefits.
Short put and covered call only work when VIX is high. In LOW VIX environments, both strategies hover near zero Sharpe (−0.03 and 0.36 respectively). The premium you collect simply isn’t enough to compensate for the times you get put stock or your upside gets capped. When VIX crosses 25, both strategies cross Sharpe 1.5+. This is not a coincidence — you’re collecting materially more premium for the same downside risk.
Bear put spread is actively dangerous in HIGH VIX. Its worst Sharpe (−1.86) occurs precisely when you might think you need protection most. High VIX already prices in significant downside, meaning your put spread is expensive. If the market reverts (which it usually does after volatility spikes), you’ve paid a premium for a move that didn’t materialize.
Long straddle only works in LOW VIX. This makes intuitive sense: you’re buying both sides when implied vol is cheap, and volatility tends to mean-revert higher from low levels. When VIX is already elevated, straddles become expensive and the expected move is already priced in.
Market Trend: Smaller But Real Effect
| Strategy | BEAR | BULL | SIDEWAYS |
|---|---|---|---|
| Calendar spread | 1.33 | 1.22 | 0.91 |
| Covered call | 0.77 | 0.53 | 0.34 |
| Short put | 0.57 | 0.15 | 0.02 |
| Collar | 0.77 | 0.73 | 0.49 |
| Wheel | 0.64 | 0.18 | 0.00 |
| Iron condor | 0.09 | −0.43 | −0.55 |
| Bear put spread | −0.97 | −0.85 | −0.77 |
| Long straddle | −0.18 | 0.25 | 0.34 |
The counterintuitive finding here: covered call and short put do better in BEAR markets than BULL markets. This seems backwards until you remember that bear markets correlate strongly with elevated VIX. It is the VIX effect, not the trend direction itself, that explains the improvement. When VIX is normalized within trend buckets, the trend effect weakens considerably.
Calendar spread is the only strategy that improves in sideways markets relative to its SIDEWAYS peers — it remains above 0.90 Sharpe regardless of trend. Every income strategy deteriorates in sideways chop because VIX tends to be compressed and premiums are thin.
Rate Regime: The Weakest Dimension
| Strategy | FALLING | FLAT | RISING |
|---|---|---|---|
| Calendar spread | 1.39 | 1.14 | 1.12 |
| Collar | 0.31 | 0.85 | 0.48 |
| Covered call | 0.33 | 0.64 | 0.40 |
| Protective put | 0.02 | 0.69 | 0.34 |
| Short put | 0.28 | 0.22 | 0.13 |
| Bear put spread | −0.34 | −1.03 | −0.53 |
Rate regime has the smallest effect of any dimension tested, but a consistent pattern emerges: FLAT rate environments are the sweet spot for income strategies. When rates are stable, equity volatility tends to be lower and more predictable, which benefits covered calls, collars, and protective puts through more orderly markets.
Calendar spread again defies the regime — it outperforms in FALLING rate environments (1.39 Sharpe), likely because falling rates tend to coincide with volatile markets where term structure opportunities are richest.
Recession: The Sharpest Reversal
| Strategy | Expansion | Recession |
|---|---|---|
| Calendar spread | 1.14 | 1.54 |
| Long straddle | 0.17 | 0.61 |
| Bear put spread | −0.95 | 0.28 |
| Bull call spread | 0.61 | −0.75 |
| Covered call | 0.76 | −0.39 |
| Short put | 0.43 | −0.49 |
| Wheel | 0.43 | −0.43 |
| Collar | 0.81 | −0.06 |
Recessions are regime inverters. Every income strategy that works in expansion — covered call, short put, wheel — goes negative in recession. The reason: realized volatility spikes, the market gaps down repeatedly, and your sold premium is overwhelmed by assignment risk and adverse moves.
Calendar spread not only survives recessions, it thrives. Its recession Sharpe of 1.54 is its highest across any regime dimension in the study. This is because recessions dramatically steepen the volatility term structure — near-month vol spikes while longer-dated vol rises more slowly — which is exactly the environment the calendar spread is designed to harvest.
Long straddle also improves, going from 0.17 to 0.61 Sharpe in recessions. You’re paying elevated implied volatility but realized volatility more than compensates.
Bear put spread finally produces positive Sharpe (0.28) in recessions — the only regime where it works — but the improvement is modest and arrives too infrequently to make it a viable core position.
The Decision Framework
Taking all four dimensions together, here is the regime playbook.
Run Always
Calendar spread — Sharpe above 1.0 in every VIX regime, both expansion and recession, all rate environments. This is the only strategy with genuine regime independence. If you run one options structure, run this one.
Run When VIX > 25
Short put / covered call / wheel — These strategies flip from near-zero to Sharpe 1.5–1.8 when VIX crosses 25. The mechanism is straightforward: elevated implied volatility means you’re collecting more premium per unit of risk. Below VIX 25, the premium barely covers commissions and adverse-selection risk.
Bull call spread — Crosses Sharpe 1.0 in HIGH VIX environments (1.08). Use when you want directional exposure on a stock that has sold off sharply and IV is elevated.
Run in Recessions Only
Long straddle — Justified only when VIX is low (< 15) or during recessions. In both cases, you’re buying optionality before a volatility event that is not yet priced in. In HIGH VIX environments, the straddle is already expensive and mean-reversion of IV works against you.
Avoid
Iron condor / iron butterfly — Positive Sharpe in CRISIS regime only (0.22 / −0.42). The rest of the time, these structures are a slow bleed. The high win rate is a psychological trap — you collect small credits for months and give it all back in a single bad expiration.
Short strangle — Similar to iron condor but with unlimited tail risk. The CRISIS Sharpe of 0.51 is the only regime where it works, but that is precisely the regime where naked short volatility can produce catastrophic losses.
Bear put spread — Works only in recessions (Sharpe 0.28) and requires being in a bear market during that recession. The cost of running it during the other 95% of history is severe.
What This Means for Portfolio Construction
The clearest application of this research is a regime-switched options overlay on top of an existing equity portfolio:
-
Core position, always: calendar spreads on your highest-conviction names, rolled monthly. Aim for positions where the near-month premium covers at least 1.5% of notional.
-
VIX trigger overlay: when VIX crosses 25, add covered call writes on the existing long positions or short puts on names you want to own at lower prices. When VIX falls below 20, take those structures off.
-
Recession hedge: maintain a small long straddle position on SPY during any month where the Fed is cutting rates and the yield curve has already inverted. This position is a drag in expansion but pays meaningfully in the scenarios where the rest of the portfolio is stressed.
-
Hard rules: no iron condors or iron butterflies unless you are specifically modeling the CRISIS regime and have a clear view on when to exit. No bear put spreads outside of confirmed recession dates.
The research does not tell you which stocks to run these on — that is the subject of the per-asset profiling work in Studies 7 and 8. The regime study tells you which structure to use once you’ve identified the trade.
Methodology Notes
All returns are synthetic. Options pricing uses Black-Scholes with realized 30-day historical volatility × 1.20 as the implied volatility input. This vol premium estimate is conservative — academic literature puts the average equity VRP between 1.10 and 1.40 depending on the period. Results would improve modestly with a higher VRP and deteriorate at 1.0 (no premium). Strike selection uses exact delta targeting via the Black-Scholes formula. Commissions ($0.65 per contract, round-trip per leg) are included. Slippage is not modeled — real execution would add friction, particularly for the multi-leg structures.
The 421 failed jobs (8% of the total queue) are primarily tickers with insufficient price history before 2008 or delisted companies where the data download failed. These are excluded from aggregates.
All results reflect historical simulation and should not be construed as a guarantee of future performance. Options involve significant risk of loss.