Methodology
Data, execution and formulas
Everything Monoqant reports can be recomputed by hand from what is on this page.
Synthetic data
Each scenario is 252 daily bars generated from a fixed seed. Log-returns r_t = μ_t + σ_t·z_t with z_t ~ N(0,1); drift and volatility depend on the regime. Open/high/low/volume are derived from the close path with bounded noise.
| Scenario | Model | Seed | Bars |
|---|---|---|---|
| Steady trend | μ = +0.07% per bar, σ = 0.8% | 1105 | 252 |
| Range-bound | μ = 0.06 · ln(100 / price), σ = 1.0% | 2202 | 252 |
| Volatility shock | σ = 0.6% for 140 bars; −8% gap; μ = −0.6% for 15 bars; σ ≈ 2.4% decaying over 60 bars | 3355 | 252 |
| Drawdown & recovery | μ = −0.22% (90 bars), 0 (50 bars), +0.28% (rest) | 4432 | 252 |
Execution model
The signal is evaluated on bar t's close and executed at bar t+1's open. A rule therefore never trades on information from the bar it acts on.
Slippage moves each fill against you; the fee is charged on the traded notional. Position size is a fixed share of current equity.
buy fill = open_{t+1} · (1 + s) sell fill = open_{t+1} · (1 − s) fee = fill · qty · f
- No order book, queue position or latency
- No partial fills or market impact
- No financing, borrow or taxes
- Constant cost assumptions
Metrics
All metrics use a 10,000 sample account and are rounded to two decimals.
net = (E_end − E_0) / E_0 gross = Σ (exit − entry) · qty / E_0 cost share = total costs / gross gains max drawdown = max_t (peak_t − E_t) / peak_t volatility = stdev(E_t / E_{t−1} − 1) · √252 return / vol = net / volatility exposure = bars with position / bars
Rule semantics
Max position size is checked once against the configured size. The drawdown stop is evaluated on every close; when reached, the position is closed at that close and no further entries are allowed.
The single-bar rule logs every bar whose equity change is worse than −limit; it does not stop trading. The frequency rule blocks an entry if the number of entries in the previous 20 bars has reached the limit.
Versions and trail
Saving recomputes results on the server from the inputs and stores a new version with a full snapshot. The trail records the action and the list of changed fields (old → new). Exports are logged too.
AI explanations
Disabled unless the operator configures a provider. When enabled, the model receives the report's numbers and your assumptions — not your email — and returns a short explanation labelled with the model name. It is a second reading, not a verdict.