A rules-based cBot for cTrader that buys short-term panic in a long-term uptrend and sells the bounce.

Built and tested specifically for US100 (Nasdaq 100 CFD). This bot is not a general-purpose, multi-instrument system and its logic and parameters have not been validated on other symbols.

The Core Idea

Since the mid-1990s, quantitative research on US equity indices has documented a persistent short-term pattern: when an index closes near the low of its daily range, or when a very short-term momentum indicator reaches an extreme oversold reading, the following days tend to see a partial rebound. This is the mean-reversion edge popularized by Larry Connors' RSI(2) research and by Internal Bar Strength (IBS) studies on index ETFs and futures.

IbsMeanReversionBot automates this exact logic on the daily chart. It does not predict the market or use machine learning. It applies two well-documented, transparent statistical signals, with strict risk controls wrapped around them.

• IBS (Internal Bar Strength) = (Close − Low) / (High − Low). A value below 0.2 means the day closed near its low, a sign of short-term panic.
• RSI(2) below 10 flags an extreme, very short-term oversold condition (the Connors RSI2 approach).

The bot buys when these conditions trigger inside a broader uptrend and exits once the panic resolves, rather than holding a losing thesis indefinitely.

How It Trades

Entry (next daily candle open):

• Signal mode is configurable: IBS only, RSI(2) only, both required (stricter, fewer trades), or either (more trades).
• Optional trend filter: only buy panics when price is above its 200-day SMA, to avoid fighting a genuine downtrend.
• Long only. Backtesting across market cycles shows the mirror-image short strategy (selling euphoria) is weak and unstable on equity indices, so this bot doesn't attempt it.

Exit (first condition met):

• IBS crosses back above the exit threshold (the bounce has played out), or
• price closes above the previous day's high (a classic Connors-style exit), or
• a time-based stop closes the trade after a set number of days if the bounce never materializes, or
• a catastrophe stop (a wide multiple of ATR(14)) closes the trade if the market keeps falling. This is not a tight scalping stop, it exists purely to cap tail risk. Mean-reversion strategies are known to be destroyed by tight stops, so this one stays deliberately wide.

Built-in risk controls:

• Position size cap: prevents oversized positions during unusually calm markets, when ATR-based sizing would otherwise produce outsized exposure.
• Volatility filter (optional): blocks new entries when the current ATR(14) spikes far above its recent average, avoiding the trap of "buying the panic" right as a genuine crash begins rather than a normal pullback.
• Single position at a time: no averaging down, no stacking.
• Only one instrument and one position open per run, keeping the logic simple and auditable.

Instrument & Setup

• Platform: cTrader (cAlgo)
• Instrument: US100 (Nasdaq 100 CFD) only
• Timeframe: Daily (D1)
• Trade frequency: roughly 30 to 40 trades per year
• Average holding period: ~3 days
• Position direction: Long only

Because trades are held for one to several nights, overnight swap/financing charges apply and are already reflected in every backtest figure below. If you trade this on a different broker, check that broker's swap conditions on index CFDs before assuming identical results, since swap markup varies significantly between brokers and can meaningfully affect net returns on a strategy that holds positions overnight.

Backtest Methodology

All results below were produced using cTrader's "Accurate" tick-data simulation mode, not the faster bar-interpolation modes. This reconstructs price movement from granular historical tick data and automatically simulates realistic spread, swap and fill behavior for every trade, rather than approximating them. It is the closest available approximation of real trading conditions within a backtest, though it still cannot fully replicate live execution: slippage, requotes and liquidity gaps in real trading can still cause results to differ.

Backtest Results (US100, Jan 2019 to Jun 2026)

• Starting balance: 10,000 (base currency)
• Net profit: +37,438
• Profit factor: 1.61
• Total trades: 266
• Winning trades: 187
• Losing trades: 79
• Win rate: ~70%
• Max drawdown: ~24.7% (at the tested risk level)

These figures come from a single broker's historical data on the US100 instrument only, using a specific parameter set arrived at through optimization. Read the disclaimer below before drawing conclusions from these numbers.

Year-by-Year Breakdown

• 2019: 28 trades, net +1,332
• 2020: 42 trades, net +6,486
• 2021: 35 trades, net +3,975
• 2022: 43 trades, net +214
• 2023: 38 trades, net +6,715
• 2024: 36 trades, net +3,012
• 2025: 29 trades, net +8,806
• 2026 (through June): 15 trades, net +6,898

No year in the tested period was net negative, though 2022 was close to flat. This does not guarantee the same will hold going forward, particularly in a sustained bear market that the tested period did not fully include.

Recommended Starting Settings

• Signal mode: IBS only
• IBS buy threshold: 0.2 / IBS exit threshold: 0.8
• Time stop: 7 days
• Catastrophe stop: 3x ATR(14)
• Volume cap: set relative to your account size, not left uncapped
• Volatility filter: enabled, threshold around 3x the 50-day average ATR
• Risk per trade: start at 1 to 2%, not the 10% used to produce the headline backtest figures above. A 10% per-trade risk setting maximizes backtest returns but also maximizes drawdown and the impact of a losing streak. It is a stress-test setting, not a recommended live setting.

Who This Is For

Traders who want a fully mechanical, non-discretionary system with a transparent, published rationale, not a black box. It suits someone comfortable holding CFD positions overnight (and paying the associated swap), trading the US100 index specifically and who can tolerate a double-digit percentage drawdown in exchange for a documented statistical edge.

Who This Isn't For

Anyone looking for a high-frequency scalper, a guaranteed-return product, a multi-instrument system, or a strategy that avoids drawdowns entirely. Mean reversion systems, by design, have long stretches where a position sits in the red before working out. Reacting to that discomfort by intervening manually will damage the strategy's own logic.

Important Disclaimers

Past performance does not guarantee future results. The backtest above reflects one instrument, one broker's historical swap and spread conditions and one time period that was, for much of its length, a rising market for US equity indices. A sustained bear market or a change in swap conditions could produce materially different results.

This strategy has been tuned using optimization on historical data. Every parameter shown above was chosen, in part, by testing multiple values and selecting what performed best historically. That process improves the fit to the past, it does not guarantee an equivalent edge going forward. Buyers are strongly encouraged to forward-test on a demo account before committing real capital.

This bot is validated for US100 only. It has not been tested or optimized for other indices, forex pairs, or asset classes. Running it on a different instrument without independent testing is not recommended.

This is not financial advice. IbsMeanReversionBot is a tool that automates a documented trading rule set, it does not manage risk on your behalf beyond what is explicitly configured and it is not a substitute for your own judgment about position sizing, capital allocation, or your personal risk tolerance. CFD trading carries a high level of risk and may not be suitable for all investors. You can lose more than your initial deposit.