A clearer view of what’s changing.
Find strength that lasts beyond a few green trades. Build a portfolio, then test the tradeoff.
Algorithm outlook
These are descriptive change labels, not buy recommendations. Use the Decision scorecard before choosing. Short-term momentum and monthly strength are separate. “Recovering” means a rebound inside a weaker month.
| Compare | Build |
|---|
The market, and what happened next.
Nifty, Sensex and India VIX — connected to a test of the selection rules.
Rules and data timing
Existing scorecard: the previous payoff and recent-profit rule. Regime-conditioned: weights historical outcomes by matching index trend and volatility, shrinks the estimate toward all-regime performance, and requires three exact-regime trades. Momentum + risk gate: ranks recent net profit among sufficiently sampled, slippage-positive algorithms, but sits out when India VIX is at or above its trailing 90th percentile.
Trend uses Nifty relative to its 20-day average (±0.5% is neutral). VIX low/middle/high buckets use trailing 60-observation percentiles. Index agreement compares Nifty and Sensex against their own 20-day averages. Five-day return and 20-day annualized realized volatility are context indicators, not additional fitted selection rules. Every historical entry uses only the previous completed market session. Selection is made before the subsequent five-session block. Market data missing at a cutoff prevents a regime-conditioned pick.
These are fixed research rules, not thresholds optimized against the displayed test outcomes. Surviving catalogue, revised published signals and small samples still bias retrospective tests. No synchronized intraday index path, broker fills or open mark-to-market is available. No future-return probability is claimed.
Source: Yahoo Finance daily chart data for ^NSEI, ^BSESN and ^INDIAVIX. This is indicative daily history, not exchange-certified tick data. Freshness is checked before an upload replaces the saved data.
Choose for the next trade. Check the evidence.
Profit per trade, payoff and resilience — at your allocation, with whole lots and costs.
Known additional plan estimates are added: Delta-Leverage ₹1,999, Lyapunov ₹1,349, Nash ₹1,499. Other paid plans are unverified; increase the fee allowance as needed. Each algorithm is tested as a separate account at the same allocation. Fees are not pooled.
What this measures — and what is still missing
Available: published signal entry/exit times, legs, lot sizes, signal utilisation, recorded margin, closed outcomes, costs, recent momentum, monthly deterioration, loss streaks, concentration and overlap. The green/red count is only one input.
Not yet connected to this score: intraday adverse excursion, actual broker fills, changed or withdrawn signals, and archived historical listings. Directional signal accuracy cannot be established from profit alone. Daily market features and fixed regime-aware selection tests are available in Market & tests. Intraday directional accuracy remains unverified.
Ranking: 100 × (0.6 × 20-session net + 0.4 × window net − 0.65 × window drawdown) / allocation, damped for fewer than 30 trades and a winner contributing more than 35% of gross wins. A fixed research heuristic, not a calibrated probability. Qualifying requires 12 closed trades, ≥80% modeled entry coverage, recent activity, positive net expectancy, positive month and 20-session net, and positive 5-point-slippage net.
Drawdown excludes open and intratrade losses. Published signals may be revised and the catalogue has survivor bias. More combinations can produce a convincing historical winner by chance: research on backtest overfitting. No rule can guarantee the next trade or remove drawdowns.
Every trading day, in one view.
Click a date heading to rank that day. Click a result to inspect its trades.
Green / red = realized published return, before costs. Dash = no final close, not zero exposure. Returns shown in percentage points (pp).
Side by side. On equal terms.
Compare up to six algorithms at the same allocation, with whole lots, utilisation, trading charges and slippage.
A mix to examine. A reason for every choice.
Find a mix that passes recent-profit and higher-slippage checks. Allocation can vary freely within your funds and reserve; ₹8.4 lakh is only the reference benchmark.
Separate subscription assumptions
Earlier quotes from 19 October, not refreshed prices. Other separate plans are unverified; increase the base allowance to cover them before relying on the net comparison.
Fees are editable assumptions, not verified current subscription quotes. Separately paid analysts may increase the total. Baseline is always ₹2.8 lakh each in Delta-Leverage, Ratio-Return and Delta-Shift. This does not read or change your saved portfolio.
Your portfolio, your choices.
Add algorithms from Discover. Enter allocations; results use whole lots and each signal’s utilisation.
Subscription total is editable; add any separately paid plans. History starts flat, with no inherited trades.
The tradeoff
Costs & monthly results
Modeled entries & quantities
| Algorithm | Entry IST | Allocation used | Lots | Quantity / leg | Recorded margin |
|---|
Meaningful changes. Less noise.
A change normally needs confirmation on two trading days. A sharp decline can be flagged sooner.
Daily snapshots are retained on the private server. The scheduled source fetch runs on your Mac and requires Codex to be available. Historical date exploration uses the revised source feed; saved daily outlooks are the record of what this dashboard actually observed.