Daily RSI Below 30 on US Stocks｜Results by Volatility Bucket
ENTRIQ backtest of RSI-below-30 buys on daily charts of 100 US stocks (5–56 years each): average P/L, average loss and MAE vary clearly by ATR-based volatility.
These are the results of backtesting 100 stocks in ENTRIQ (5 to 56 years of data depending on the stock) (Dataset v2.0). The overall picture of buying when RSI drops below 30 is covered in Buying When RSI Drops Below 30｜Backtest of 3 Exits and Risks. This article digs deeper into the results by volatility bucket.
How the volatility buckets are defined
Even with the same "RSI below 30" condition, the range of price moves after entry is completely different for a stock that moves quietly and one that moves violently. In this backtest, ATR(14) divided by the close (ATR(14)/close, a relative measure that divides ATR by the share price) was used as the measure of volatility.
The buckets are defined as follows.
- High volatility: ATR(14)/close is in the top 20% over the stock's full history
- Low volatility: ATR(14)/close is in the bottom 20% over the stock's full history
- Normal: The remaining middle 60%
Rather than comparing absolute ATR levels between stocks, the buckets are based on the relative position within the same stock. The high-volatility bucket therefore means "periods when the stock is moving more than usual for that stock," and the low-volatility bucket means "periods when the stock is quieter than usual for that stock."
Test conditions
- Universe: 100 US stocks selected from the current S&P 500 constituents, taking the largest by market cap with up to 15 stocks per sector
- Timeframe: Daily (split-adjusted)
- RSI setting: Period 14, Wilder's smoothing
- Entry: Buy at the open on the next trading day after RSI(14) crosses below 30. New signals in the same stock while a position is open are skipped
The exits were compared across 3 patterns.
- Pattern A (RSI back to 50): Exit at the open on the next trading day after RSI returns to 50 or above (if it doesn't, exit at the close on the 60th trading day). No stop loss
- Pattern B (10 trading days): Exit at the close on the 10th trading day after entry. No stop loss
- Pattern C (5% stop loss): Stop loss at -5%. If the stop isn't hit, exit at the close on the 10th trading day. If the stock gaps down at the open, the fill is at that open price, so a single trade's loss can exceed -5%
About survivorship bias: This test uses the current S&P 500 constituents. Stocks that were delisted or removed from the index in the past are not included, so the results may be skewed in a somewhat more favorable direction.
Results by volatility
Pattern A (RSI back to 50, no stop loss)
| Bucket | N | Share of trades closed in profit | Average P/L | Average loss | Median MAE | Largest loss |
|---|---|---|---|---|---|---|
| High volatility | 1,764 | 70.7% | +3.33% | -12.00% | -7.33% | -93.3% |
| Normal | 3,030 | 73.1% | +1.84% | -7.12% | -4.05% | -64.7% |
| Low volatility | 524 | 71.6% | +0.55% | -6.62% | -2.60% | -50.1% |
Note: Past sample (N=1,764, N=3,030, N=524); not indicative of future results.
Looking at pattern A by volatility, average P/L was largest in high volatility, then normal, then low volatility, with the high-volatility bucket showing the largest average P/L. On the other hand, the share of trades closed in profit ranked normal > low volatility > high volatility, making high volatility the bucket with the lowest share.
On the loss side, the average loss, largest loss and median MAE were all the most negative in high volatility, then normal, then low volatility. In the high-volatility bucket, positions are held without a stop loss until RSI returns to 50, so they stay open through adverse moves after entry, and trades whose losses grew large pull the average loss down (more negative). The low-volatility bucket had the median MAE closest to zero, but profit factor was highest in the normal bucket and lowest in low volatility.
Pattern B (10 trading days, no stop loss)
| Bucket | N | Share of trades closed in profit | Average P/L | Average loss | Median MAE | Largest loss |
|---|---|---|---|---|---|---|
| High volatility | 2,541 | 57.4% | +1.91% | -7.68% | -5.63% | -72.5% |
| Normal | 3,828 | 56.9% | +0.86% | -4.62% | -3.13% | -44.3% |
| Low volatility | 660 | 54.7% | +0.38% | -3.31% | -2.15% | -22.1% |
Note: Past sample (N=2,541, N=3,828, N=660); not indicative of future results.
In pattern B, the share of trades closed in profit, average P/L, profit factor, average loss, median MAE and largest loss all ranked high volatility > normal > low volatility (for the loss metrics, from most negative to least negative), showing a consistent tendency for the range of P/L to be wider the higher the volatility. With exits at a fixed 10 trading days, larger returns were observed in the high-volatility bucket, while the average loss was also larger (more negative).
Pattern C (5% stop loss, 10 trading days)
| Bucket | N | Share of trades closed in profit | Average P/L | Average loss | Median MAE | Largest loss |
|---|---|---|---|---|---|---|
| High volatility | 3,143 | 37.9% | +0.54% | -4.92% | -5.00% | -37.9% |
| Normal | 4,058 | 50.9% | +0.58% | -4.08% | -3.14% | -17.6% |
| Low volatility | 664 | 52.7% | +0.31% | -3.18% | -2.16% | -16.4% |
Note: Past sample (N=3,143, N=4,058, N=664); not indicative of future results.
What stands out in pattern C is the change in the ranking of the share of trades closed in profit. In pattern C, which uses a stop loss, the share of trades closed in profit was highest in low volatility, then normal, then high volatility, with the high-volatility bucket the lowest. This is the exact reverse of the ranking in pattern B and also differs from the ranking in pattern A.
On the other hand, average P/L ranked normal > high volatility > low volatility, so the rankings for the share of trades closed in profit and for average P/L differ. Profit factor ranked normal > low volatility > high volatility, with the normal bucket the highest. The average loss, largest loss and median MAE were the most negative in high volatility, then normal, then low volatility, a tendency common to all 3 patterns.
Considering stop width and volatility together
The results by volatility for pattern C (fixed -5% stop loss) help illustrate the relationship between stop width and the size of price moves.
In the low-volatility bucket, the median MAE (-2.16%) was smaller than the stop width, and many trades reached the 10th trading day without hitting the stop. In the high-volatility bucket, the median MAE (-5.00%) was close to the stop width, and a tendency for a relatively larger share of trades to hit the stop was observed.
Note: Past sample (N=664, N=3,143); not indicative of future results.
This backtest uses a uniform fixed stop width of -5%. Results could change if the stop width were linked to ATR, but that is outside the scope of this article. To find out what stop width suits your own approach, it is important to test it on your own stocks, timeframe and holding period.
Comparing the 3 patterns (where each bucket stands)
Looking across the 3 patterns, the following tendencies are observed for each volatility bucket.
High volatility: Average P/L was largest in pattern A, then B, then C. However, the average loss and median MAE were also largest (most negative) in A, and in the pattern that keeps holding without a stop loss, losses during adverse moves were larger. The share of trades closed in profit in pattern C was the lowest of the 3 buckets, with many periods observed in which the -5% stop loss was easily reached.
Normal: In pattern C, profit factor was the highest of the 3 buckets, and the combination of the -5% stop loss and the 10-trading-day exit showed relatively stable results.
Low volatility: Average P/L was the smallest of the 3 buckets in every pattern, and the median MAE was also closest to zero. The share of trades closed in profit in pattern C was the highest of the 3 buckets.
Limitations and caveats
- This backtest uses the current constituents of the S&P 500, so it is subject to survivorship bias (stocks that were delisted or removed from the index in the past are not included)
- The volatility buckets are based on the relative position over each stock's full history. The same stock can fall into different buckets at different times
- ATR(14)/close is used as the measure of volatility, but other measures (such as historical volatility or implied volatility) could produce different bucket results
- Trading costs and slippage are not included
- This backtest covers US stocks only (FX, commodities and crypto assets are not included)
Summary
Splitting buys below RSI 30 by volatility bucket, the following tendencies were observed.
- The average loss, largest loss and median MAE were the most negative in high volatility, then normal, then low volatility, in all 3 patterns
- Average P/L was largest in high volatility in patterns A and B, with a tendency for the range of P/L to be wider
- In pattern C (fixed -5% stop loss), the share of trades closed in profit ranked low volatility > normal > high volatility, a different ranking from patterns A and B, which suggests that the relationship between the fixed stop width and actual volatility may be affecting the results
These observations provide a basis for examining stop width in relation to volatility. However, "what stop width is appropriate" changes with the test conditions, stocks and timeframe, so rather than applying the figures in this article as they are, the first step is to check under your own conditions.
Try the same conditions on your own stocks and timeframe with ENTRIQ's Chart Replay, and build up your results with Strategy Tags to accumulate your own testing data.
Frequently asked questions (FAQ)
Q. Do I have to calculate ATR/close myself? A. ENTRIQ's Chart Replay lets you check each stock's price moves as they are. You need to calculate ATR/close manually, but when reviewing the chart, you can also ask whether the stock is moving more than usual.
Q. Should I avoid the high-volatility bucket? A. This backtest organizes historical observations and does not recommend any particular bucket. The high-volatility bucket has a larger average P/L but also a wider range of losses, and which to prioritize depends on your own rules.
Q. Can volatility buckets be combined with the market regime (trend/range)? A. They can be combined in a backtest, but the finer the buckets, the smaller the sample size and the lower the reliability of the results. The minimum sample size used as a guideline in this backtest is 30 trades.
Q. Are there results for stop widths other than -5%? A. This article only reports results for a fixed -5% stop as pattern C. Results for other stop widths are outside the scope of this article. Consider testing under your own conditions with ENTRIQ.
Q. Are these results the same for Japanese stocks or FX? A. This backtest covers US stocks (S&P 500 constituents) only. Results may differ for Japanese stocks, FX and other instruments.
Related articles
- Complete Guide to RSI｜Reading, Settings, Entries and Testing
- Daily RSI Below 20 on US Stocks｜Comparing 3 Exits
- Daily RSI 9, 14, 21 on US Stocks｜What Changes in Backtests
- Daily RSI Backtest on US Stocks by Trend｜How Results Vary
This article is provided for informational purposes to examine and organize historical data and does not recommend any specific trade. The figures shown are past samples and do not indicate future results. All investment decisions are your own responsibility.
Reproduce this validation yourself
With ENTRIQ's chart replay, you can trade through past charts using the same rules.
Try the demo — no sign-upNo sign-up or card needed. Runs right in your browser. Start 14-day free trial →
Validate your method on past charts
The same validation in this article can be reproduced with your own method. On past charts, do the classics really work?
- ✓ Replay past charts
- ✓ AI feedback on your trades
- ✓ Validate methods with statistics
No sign-up or card needed. Runs right in your browser.