Abu Saleh — TASI Strategy

Candidates

Strategy Rationale

Abu Saleh is a Donchian-channel breakout strategy adapted from classic US trend-following (similar to the original Turtle system) and tuned for the peculiarities of Tadawul: shorter trend cycles, thinner liquidity outside the top 30 names, ±10% daily price limits, T+2 cash settlement, and no fractional shares. The core idea is straightforward — buy when a stock breaks above its N-day high on real volume, size the trade so that a single ATR-wide stop costs exactly X% of equity, and let the position ride until a trailing stop, a target multiple, or a max-holding limit closes it.

Signal strength is a composite rank of ADX (trend intensity), RSI (momentum confirmation), and a 20-day value-traded percentile (liquidity screen). Only the top-ranked candidates fill open position slots — at most 5 concurrent positions, capped at 20% of capital each, to prevent Aramco-sized concentration.

Exit rules are layered: a hard stop at entry minus 2.5 ATR, a trailing stop at peak equity minus 3.5 ATR, a fixed target at 3R (3× the initial risk), and a max-hold cap of 60 bars (~3 calendar months). Whichever exit fires first wins. This keeps the strategy from overstaying weak trends.

Orders are limit only — no market orders into thin Tadawul books. Entries use the prior close as the limit reference; exits are placed at the trigger level. Commission is modelled at 15.5 bps and slippage at 8 bps, both conservative for retail Tadawul.

The TASI universe is curated (TASI 30 or the full SA Main list), not auto-downloaded, so it avoids survivorship bias from newly listed shells or delisted names. The pipeline refreshes price history from yfinance every 60 minutes.

Parameter Tuning — Original vs TASI

US Turtle defaults adjusted for Tadawul market microstructure:

Filter Matrix Current scan — all filters per ticker

Important Caveats

• Backtests overstate live results. Slippage, partial fills, broker latency, dividend and corporate-action handling, and survivorship bias all eat into real performance. Treat simulated numbers as an upper bound.
• The TASI universe is curated, not the full index. The default TASI 30 list covers the largest and most liquid names. Extending to SA Main adds smaller-cap names with higher bid-ask spreads.
• No fundamental filter is applied. There is no earnings quality screen, no Shariah compliance filter, and no sector exclusion. Add your own overlays before trading with real capital.
• This is research output, not financial advice. Run in paper mode for at least a full quarter before risking real money. Past performance of any model does not guarantee future results.
• yfinance data quality varies for Saudi names. Some .SR tickers return empty or stale data; the pipeline prints warnings and skips them. Ramadan / Eid session lengths slightly inflate intraday ATR estimates.

Configure Backtest