Quantitative Laboratory
I Backtested the CNN Fear & Greed Index Across 9,000+ Trades and Here's What the Data Says
The common wisdom of “buy fear and sell greed” is oversimplified, and sometimes dead wrong.
What Is the CNN Fear & Greed Index?
Most investors who have any exposure to financial social or mainstream media have at some point come across the index. It is posted on the CNN website here. The index was begun in 2011.
The US Fear and Greed Index (CNN F/G) is constructed from seven indicators of stock market performance. They are measured on a scale of 0 to 100 and an equally weighted average of the seven indicators is calculated. The seven indicators are:
- Momentum
- The S&P 500 index compared to its 125-day moving average.
- Stock Price Strength
- How many stocks have hit 52-week highs and 52-week lows on the New York Stock Exchange.
- Stock Price Breadth
- The volume of shares traded in stocks where prices are rising compared to the volume of shares traded in stocks where prices are declining.
- Put/Call Ratio
- How many call options traders have purchased vs. how many put options they have purchased.
- Demand for Junk Bonds
- The spread between the yields on junk bonds and investment grade bonds.
- Market Volatility
- The CBOE Volatility Index, or VIX, measures the U.S. stock market's 30-day expected volatility based on S&P 500 options. The CNN Fear and Greed Index uses this, based on a 50-day moving average.
- Demand for Safety
- The difference in return on stocks vs. the return on US Treasury bonds.
Backtest Methodology
I used the DTN IQFeed for the SPY data used in the analysis. The Wealth-Lab software was used to import the CNN Fear & Greed Index data, and backtesting was performed on this platform.
I tested entries across all decile ranges (0-10, 10-20, 70-80 etc.) with hold periods from 1 to 21 bars (1, 2, 3, 4, 5, 10, 15 and 21 specifically), testing the whole range of CNN F/G data back to 2011.
Has buy fear and sell greed been an effective strategy in these shorter holding periods of under 1 month?
Actually, sentiment regimes matter more than sentiment extremes, and hold period must match the regime.
Terminology
First off I need to define some terminology used in the results spreadsheet and as I unwrap what the testing indicated.
- Bars Held
- The holding period, defined from entering the next day after the specific Fear & Greed regime is in place and holding X periods.
- Avg Profit Pct
- The net average profit per trade when analyzing all trades, including winners and losers.
- Profit Factor
- A ratio that compares the total profits generated by winning trades to the total losses incurred by losing trades. A reading greater than 1 means the strategy is profitable. For example, a profit factor of 1.5 indicates we earn $1.50 in profits for every $1.00 in losses.
- Win Rate
- The percent of trades that finished positive.
- Position Count
- The number of trades triggered.
- Profit Per Bar
- Assumes each trade is given a $5,000 fixed position size, and divides the sum of the trade profit by the number of bars held.
- Exposure
- What percentage of time is spent in the market. For example, 1.41 indicates being in the market 1.41% of the time.
The Findings: SPY Returns by Fear & Greed Range and Hold Period
The methodology was straightforward: when the CNN Fear & Greed Index closed within a specific decile range (0-10, 10-20, 20-30, etc.), I entered SPY at the following day's open and held for a predetermined period of 1, 2, 3, 4, 5, 10, 15, or 21 trading days.
This allowed me to answer questions like: does buying SPY when Fear & Greed reads 50-60 and holding for 5 days outperform buying at 70-80 with the same 5 day hold? Is entering at 20-30 and holding for 21 days (approximately one month) more profitable than entering at 50-60 for the same duration? By testing every combination of sentiment range and hold period, I could identify which regimes favor which timeframes.
Below is a graphic which summarizes the findings. On the left scale we see the number of bars held, and the bottom scale shows the CNN F/G range.
The Sweet Spot: Moderate Fear (20-40 Range)
This is where the magic happens. The 20-40 Fear & Greed range with 15-21 bar holds consistently produces the best risk-adjusted returns in the entire study.
| F&G range, 21-bar hold | Avg profit | Win rate | Profit factor | Trades |
|---|---|---|---|---|
| 20-30 | +1.45% | 70.69% | 2.00 | 58 |
| 30-40 | +1.68% | 71.79% | 2.64 | 78 |
These aren't just good numbers, they're exceptional for a single broad “sentiment” measure. A 2.64 profit factor means your total winnings are 2.64 times your total losses. Combined with a 71.79% win rate, you're stacking the odds heavily in your favor.
Why does this range work so well? It captures the inflection point where fear has subsided enough for buyers to step in, but sentiment hasn't reached euphoria where gains are already priced in. The market is transitioning from capitulation to cautious optimism, and that transition is extremely profitable over a 3 to 4 week timeframe.
The 10-20 range also shows solid performance at longer holds (+0.88% at 21 bars, 62.16% win rate), but the 20-40 zone is the clear champion for mean reversion plays.
The Sample Size Advantage
One crucial detail: the 20-40 range has some of the largest sample sizes in the study. The 30-40 range alone generated 78 trades at the 21-bar hold period, and 285 trades at the 1-bar period. This isn't a lucky streak, it's a statistical edge.
The Danger Zone: 60-70 Range (Comfortable Greed)
If the 20-40 range is the sweet spot, the 60-70 range is where profits go to die. Look at these dismal numbers:
−0.01%Avg profit, 2-bar hold
49.82%Win rate (sub-50%)
0.98Profit factor
277Trades
You're more likely to lose than win, and even when you win, the gains barely cover your losses. The 1 bar hold at this range produces just +0.01%, essentially breakeven despite 371 trades worth of data.
This “comfortable greed” zone represents market indecision. Sentiment is positive enough that contrarian mean reversion doesn't work, but not strong enough to create genuine momentum. The result? Choppy, range-bound action that grinds through your capital without producing meaningful edge.
Even extending to 21 bars only improves performance to +0.81% average (versus +1.68% in the 30-40 range). The opportunity cost of trading this zone is significant.
The Momentum Zone: 70-100 Range (Growing Greed)
This is where the data gets really interesting and challenges conventional “sell into greed” wisdom.
The 70-80 range produces explosive short term results. At a 5 bar hold: +0.42% average profit, a 69.75% win rate, a 2.32 profit factor, and profit per bar of 4.22, the highest efficiency in this range.
But the real story emerges at longer holds across the entire 70-100 spectrum:
| F&G range, 21-bar hold | Avg profit | Win rate | Profit factor |
|---|---|---|---|
| 70-80 | +0.62% | 71.70% | 1.76 |
| 80-90 | +1.11% | 72.00% | 3.52 |
| 90-100 | +1.35% | 77.78% | 4.77 |
That 4.77 profit factor at extreme greed is remarkable: when you do lose, the total losses across all losing trades are much smaller than the total gains across all winning trades. The 80-90 range at 4 bar holds shows an even more extreme 4.00 profit factor, suggesting that short term momentum bursts can be captured efficiently.
The Extreme Greed Paradox
The 90-100 range deserves special attention. At 15-bar holds, it produces an average profit of +1.21%, a 90% win rate, and a 6.75 profit factor, the highest in the entire dataset.
This is extraordinary. A 90% win rate with winners nearly 7x larger than losers. However, and this is critical, the sample size is only 10 trades. The 21 bar hold at 90-100 has just 9 trades.
These exceptional numbers could represent:
Genuine blow-off top momentum that continues longer than expected.
Statistical noise from small samples.
I'm inclined to think it's a combination of these, with genuine momentum playing the largest role. When markets reach extreme greed, the speculative fervor often persists for weeks before exhausting itself. But with such limited data, treating this as a primary strategy would be premature.
Hold Period Optimization: Patience Pays in the Right Regime
One of the clearest patterns across the entire dataset: longer holds generally outperform shorter holds, but the magnitude varies dramatically by sentiment range.
At extreme fear (0-10), extending from 1 to 5 bars transforms a losing strategy into a marginal winner. But in the sweet spot (20-40), extending to 21 bars produces gains that are 8-30x larger than 1-bar holds.
Here's the profit progression for the 30-40 range:
+0.02%1 bar
+0.35%5 bars
+0.74%10 bars
+1.12%15 bars
+1.68%21 bars
The trend is clear and consistent. Patient capital gets rewarded, especially in the right sentiment regimes.
Two Fear & Greed Trading Strategies From the Data
Based on this data, here are two concrete approaches you can implement, on SPY or on stocks that tend to be correlated with it:
Strategy 1: The Patient Contrarian
Highest confidence- Entry: CNN F&G reads 20-40
- Hold: 15-21 bars (3-4 weeks)
- Expected: 1.12-1.68% per trade, 70-72% win rate, 2.0-2.6 profit factor
- Best for: Swing traders comfortable with month long positions
- Why it works: Captures recovery after fear subsides, before euphoria
Strategy 2: The Momentum Rider
Moderate confidence- Entry: CNN F&G reads 70-80
- Hold: 5-10 bars (1-2 weeks)
- Expected: 0.42-0.48% per trade, 69% win rate, 2.0-2.3 profit factor
- Best for: Active traders wanting to ride positive sentiment waves
- Why it works: Captures short-term momentum bursts when greed is building
What About the Middle? The 30-60 Range
The 30-60 range produces good results across most timeframes, especially when holding longer than 10 days. This “neutral to slightly bullish” sentiment zone becomes increasingly profitable with patience.
| F&G range | Hold | Avg profit | Win rate | Profit factor | Trades |
|---|---|---|---|---|---|
| 30-40 | 10 bars | +0.74% | 67.54% | 1.74 | 114 |
| 30-40 | 15 bars | +1.12% | 69.47% | 2.17 | 95 |
| 30-40 | 21 bars | +1.68% | 71.79% | 2.64 | 78 |
| 40-50 | 10 bars | +0.61% | 68.46% | 1.71 | 130 |
| 40-50 | 15 bars | +0.59% | 67.92% | 1.49 | 106 |
| 40-50 | 21 bars | +1.23% | 70.65% | 2.26 | 92 |
| 50-60 | 10 bars | +0.22% | 59.56% | 1.24 | 136 |
| 50-60 | 15 bars | +0.65% | 64.91% | 1.75 | 114 |
| 50-60 | 21 bars | +0.76% | 68.04% | 1.69 | 97 |
Notice the consistent trend: as you extend your hold period to 15-21 days, win rates climb into the high 60s and low 70s, while profit factors improve significantly. The 30-40 range clearly leads this pack, but even the 50-60 “neutral” zone, which looks mediocre at short holds, transforms into a respectable performer at 21 bars.
What's particularly compelling is the sample size advantage. The 30-40 range alone generated over 100 trades at the 10-bar hold, and the 40-50 range produced 130 trades. These aren't fluky results; they're statistically robust edges. The market is telling us that when sentiment is neither fearful nor greedy, giving positions 2 to 4 weeks to develop produces consistent, reliable gains.
This zone represents the “steady grind higher” market environment, where there's no panic to fade and no euphoria to ride, but gradual accumulation and positive price action reward patient holders.
The Bigger Picture: Sentiment Regime Matters More Than You Think
Perhaps the most important lesson from this entire analysis: market sentiment regimes require different strategies.
Extreme fear requires patience and the willingness to endure continued volatility. Moderate fear offers the best risk-adjusted returns for patient capital. Comfortable greed (60-70) is a trap. Growing greed (70-80) creates genuine momentum that can be captured. And extreme greed, at least in this dataset, shows surprising persistence.
The old adage “buy fear, sell greed” is too simplistic. The data shows you should:
- Buy moderate fear (20-40) and hold for weeks
- Buy growing greed (70-80) and hold for days or weeks
- Avoid extreme fear with short holds
- Avoid comfortable greed (60-70) with any short term hold
It's not about timing the emotional extremes. It's about matching your holding period to the sentiment regime and understanding when mean reversion versus momentum strategies will work.
In part two of exploring this dataset I will explore the implementation of real world strategies inside the various regimes we explored today. We'll also look to extend the holding period beyond 21 days to see if we find any edges there.
Have a great night!
Raw Research is Not a Strategy
A sentiment reading is one input. The Trading Time Machine systems turn market context into complete, mechanical rules: entries, exits, position sizing and risk, tested across decades of SPY data.
Explore the Systems Free SPY ForecastOriginally published on Substack: I Backtested the CNN Fear & Greed Index Across 9,000+ Trades and Here's What the Data Says