Key results
- Total return
- +26.9%
- Buy & hold: +528.3%
- Annual return (CAGR)
- +2.4%
- Buy & hold: +20.2%
- Max drawdown
- -25%
- Buy & hold: -35.7%
- Win rate
- 75.38%
- Profit factor
- 1.55
- Trades
- 65
- Longest losing streak
- 3 trades
- Period
- Oct 3, 2016 to Oct 1, 2026
- 10 years
- Market
- NAS100/USD
- Daily
Equity curve vs buy & hold
Returns year by year
| Year | Return | Trades |
|---|---|---|
| 2017 | +4.8% | 4 |
| 2018 | -1.4% | 8 |
| 2019 | -1.7% | 7 |
| 2020 | -12.2% | 2 |
| 2021 | +7.5% | 9 |
| 2022 | +10.4% | 7 |
| 2023 | +4.6% | 7 |
| 2024 | +7.8% | 8 |
| 2025 | -0.6% | 8 |
| 2026 | +6.9% | 5 |
Calendar years, compounded from monthly results. The first and last years can be partial.
Larry Connors and Cesar Alvarez popularised RSI(2) in Short Term Trading Strategies That Work: buy a sharp two- or three-day pullback inside a long-term uptrend, then sell into the first bounce. The rule set we tested follows the version documented by Quantified Strategies. The Nasdaq 100 is the natural first market to try. It is the most volatile of the large US indices, and the usual argument is that RSI(2) works better where pullbacks are deeper, because the bounce that follows is bigger.
The rules we tested
- Market and period: Nasdaq 100 (NAS100/USD), daily candles, October 2016 to October 2026.
- Long entry: the 2-period RSI is below 5 and the close is above the 200-day simple moving average (SMA).
- Long exit: the close is back above the 5-day SMA, or the close falls below the 200-day SMA.
- Short entry: the mirror image, RSI(2) above 95 with the close below the 200-day SMA.
- Short exit: the close is below the 5-day SMA, or back above the 200-day SMA.
- Stop and target: a wide protective stop-loss far from entry (it was hit only twice) and no take-profit. The moving-average exits do the work.
- Sizing and costs: 100% of equity per trade, 0.02% commission per side, $10,000 starting capital.
What the backtest shows
The win rate is real: 75% of the 65 trades closed in profit and the longest losing streak was only 3 trades. The payoff is the price of that comfort. The average win was 1.58% and the average loss 3.11%, so one loser erases about two winners. With those sizes, the strategy needs roughly 66% winners to break even. It got 75%, which is the margin behind the 1.55 profit factor.
The problem is scale. +26.9% over ten years is 2.4% a year, against 20.2% a year for the index. With six or seven trades a year lasting a few days each, the account is flat most of the time. Commissions are not the issue: 65 round trips at about 0.04% each cost under 3% of the position over the decade.
Read the drawdown in context
The strategy's worst drop was 25.0%, against 35.7% for buy and hold. That looks safer, but the strategy's drawdown is almost as large as its whole ten-year gain, while the index's drawdown was a small fraction of its gain.
Year by year, 2020 is the scar: two trades, both losers, cost 12.2%. That is what a pullback rule looks like when the pullback keeps going. The best year was 2022 (+10.4%, 5 winners out of 7 trades), the year the index fell and spent long stretches below its 200-day average, which is the only place the short rules can trade. On the S&P 500 the same rules made +18.1% over the same decade, so the Nasdaq is where they did best.
Why it works, and why it is not enough
Short-term mean reversion in stock indices is one of the better documented effects: after a few sharp down days inside an uptrend, the following days tend to bounce. The 200-day filter keeps the trade on the side of the trend, and the 5-day average exit banks the bounce early. That design produces a high win rate and small wins, and it leaves the long climbs between signals unclaimed.
The losses are larger than the wins by construction. A losing trade is held until the price reclaims the 5-day average or breaks the 200-day one, and in a fast sell-off that can be far away.
What you could test next
- Loosen the entry to RSI(2) below 10 and check whether the extra trades keep the same quality.
- Exit later, for example on a 10-day SMA, to let the bounce run further.
- Remove the short side and compare: on an index that rose most of the decade, does it add or subtract?
- Add a time exit after a fixed number of candles so a losing dip cannot drag on.
Methodology and assumptions
- Starting capital
- 10,000
- Position size
- 100% of equity
- Commission
- 0.02% per side
- Data window
- Oct 3, 2016 to Oct 1, 2026
- Run date
- October 2, 2026
- No repainting: every signal is computed on closed candles only, so the backtest never acts on a price a trader could not have seen yet.
- Past performance does not predict future results. A backtest is a historical simulation, not a forecast.
Frequently asked questions
What are the Connors RSI(2) rules?
Buy when the 2-period RSI is below 5 while the close is above its 200-day simple moving average. Exit when the close goes back above the 5-day moving average, or when it falls below the 200-day average. The mirror rules go short: RSI(2) above 95 with the close below the 200-day average.
Is the RSI(2) strategy profitable?
In this test, modestly: +26.9% over ten years on the Nasdaq 100, a profit factor of 1.55 and 75% winning trades. It came nowhere near buy and hold (+528.3%) because it is invested only a few days at a time.
Why does a 75% win rate still lose to buy and hold?
Because each win is small (1.58% on average) and each loss is about twice as large (3.11%), and because the strategy is out of the market on most days, including most of the days the index rises.
What timeframe does the RSI(2) strategy use?
Daily candles. The RSI uses 2 periods, the trend filter is the 200-day simple moving average and the exit is the 5-day simple moving average.
Reproduce it, then change it
Every number on this page comes from the Backtrex engine. Rebuild the strategy in the app, then change the market, the timeframe or a parameter and see whether the result holds.
Ready to backtest your strategies?
Create your free account and take the guided tour: a real 10-year backtest in 3 minutes, no coding, no credit card.
Create your account in 30 seconds and run your first backtest. No credit card.
Other backtests
- Strategy
- +37.7%
- Buy & hold
- +527.2%
- Strategy
- +18.1%
- Buy & hold
- +255.4%
- Strategy
- +14.4%
- Buy & hold
- +136.7%
- Strategy
- -4.5%
- Buy & hold
- +0.1%