> loading data_
> loading data_
/methodology
DanioX turns a discretionary trading framework into published rules. Nothing is hand-tuned per day and nothing is hidden: the tables below are generated from the same code that scores the live dashboard.
Each season is a weighted checklist summing to 100. A condition scores 0–1 (most are yes/no; a few are graded). If a data source is down, its conditions are removed from both the numerator and the denominator, the score is rescaled, and the gauge shows the share of the checklist that was evaluable. A season needs at least 40% of its weight available to be scored. The dominant season is the highest score; conviction is strong at 70+ with a 15-point lead, moderate at 50+, otherwise weak.
| DXY above 50-day EMA | 20 |
| DXY rising (20d) | 20 |
| US 10Y yield rising (20d) | 15 |
| VIX above 20 | 15 |
| Gold falling (20d) | 10 |
| Crypto breadth below 40% | 20 |
| Gold above 50-day EMA | 25 |
| Gold rising (20d) | 25 |
| Gold outperforming S&P 500 (20d) | 25 |
| Dollar weak (below EMA50 / falling) | 25 |
| BTC daily uptrend (close > EMA50, EMA20 > EMA50) | 30 |
| BTC dominance rising (7d) | 25 |
| ETH/BTC falling (7d) | 20 |
| BTC 4h close above 4h EMA50 | 25 |
| BTC.D falling while BTC holds above EMA50 | 20 |
| ETH/BTC rising (7d) | 10 |
| ETH/BTC above 50-day EMA | 10 |
| Breadth ≥ 40% (coins in daily uptrend) | 20 |
| Alts beating BTC over 30 days | 15 |
| Altcoin Season Index (90d) | 25 |
Graded conditions: VIX scores 1 above 20, 0.5 between 16 and 20 when rising. “Dollar weak” scores 0.5 for each of (below EMA50, 20-day change < 0). Alt outperformance (30d) and the alt season index (90d) scale linearly from 0 at 25% to full credit at 75%. BTC dominance moves smaller than ±0.05 points over 7 days count as flat.
For Gold, Stocks (average of S&P 500 and Nasdaq), BTC and an equal-weight alt basket we compute the 20-day return divided by (daily volatility × √20). The link with the highest volatility-adjusted momentum leads, provided it exceeds 0.25; otherwise the chain is “stalled”. Note that stock and gold series have ~5 bars per week, so “20 days” means 20 sessions (about four weeks) for them and 20 calendar days for crypto.
Universe: 42 liquid Binance USDT pairs (ETH, SOL, XRP, BNB, DOGE, ADA, TRX, AVAX, LINK, DOT, TON, SUI, NEAR, APT, LTC, BCH, UNI, AAVE, ATOM, FIL, ARB, OP, INJ, TIA, SEI, FET, RENDER, PENDLE, CRV, LDO, HBAR, ICP, ETC, XLM, WLD, JUP, ENA, ONDO, TAO, POL, ALGO, VET). Symbols Binance rejects are skipped. The index is the % of coins whose 90-day return beats BTC's; ≥ 75 = alt season. Breadth is the % of the universe in a daily uptrend.
The current value is CoinGecko's market_cap_percentage.btc. For the 7/30/90-day trend we model each large coin's past market cap as today's share × (past price / today's price): stablecoins are held flat, staked/wrapped tokens follow their underlying, and every other coin plus the long tail follows the equal-weight alt basket. Supply changes are ignored, so it is labelled approx.
The signal board publishes only rule setups that made money after fees in a long backtest. Trend Breakout (from Zarattini, Pagani and Barbon, "Catching Crypto Trends", 2025): buy the daily close above the highest close of the previous 20 days; the stop is the midpoint of that closing range, raised after every daily close and never lowered; sell at the first daily close below it. Smart Trend: the same breakout on today's top 100 coins, skipped while BTC closes below its 200-day SMA or when the coin's 50-day SMA fell more than 5.93% in 10 days. Regime Trend: the same breakout on each day's 20 most traded coins, only while BTC closes above its 200-day SMA and at least half of those coins close above their 50-day SMA; every open trade is also sold at the first daily close with BTC below its 200-day SMA. A bot scans after every daily close (00:00 UTC), an AI review can reduce or veto a candidate, at most 3 new signals a day and 6 open per setup are published, and the bot closes each signal at its trailing-stop exit (Regime Trend also at its regime exit). CTP, FGA, AA, AMD, AMD Structural and Spot Swing were retired on 19 October 2026 because their backtests were about break-even or negative after fees.
Every live setup is backtested with one method: the live rules replayed on Binance daily candles from January 2019 (or a coin's listing date) to today, closed candles only, a 0.1% fee per side plus 0.02% slippage on every fill, results in R (the planned loss of one trade). The headline numbers are the full period; each calendar year, 2019-2022 against 2023-2026 and the trades a follower could take with the live caps are reported separately. Trend Breakout and Smart Trend use today's coin list (survivorship bias); Regime Trend uses every Binance pair ever listed, dead coins included (from July 2019). Newer coins join on their listing date. The table puts each backtest next to the live success ratio of the signals actually published. Details, equity curves and every trade: signal backtests.
| Setup | Trades | Win | Avg R net | Before costs | PF | Max DD | 2019 | 2020 | 2021 | 2022 | 2023 | 2024 | 2025 | 2026 YTD | Live caps | Live success |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| RTREND 1D trendSep 2019 – Oct 2026 | 830 | 38% | +0.59R | +0.61R | 2.13 | 64R | -0.59R24 trades | +0.81R162 trades | +0.53R186 trades | — | +0.59R148 trades | +0.68R170 trades | +0.19R104 trades | +1.47R36 trades | +0.65R359 trades | none yet |
| TREND 1D trendJan 2019 – Oct 2026 | 2,391 | 33% | +0.92R | +0.95R | 2.43 | 235R | +0.61R92 trades | +2.02R190 trades | +2.09R253 trades | -0.46R228 trades | +1.37R336 trades | +1.31R397 trades | +0.15R391 trades | +0.57R504 trades | +0.73R749 trades | 0% (0/13) |
| STREND 1D trendOct 2019 – Oct 2026 | 2,135 | 39% | +1.18R | +1.20R | 2.99 | 202R | +0.13R17 trades | +1.64R211 trades | +1.53R331 trades | — | +0.95R428 trades | +1.80R512 trades | +0.20R432 trades | +1.19R204 trades | +0.88R454 trades | 0% (0/7) |
All numbers after costs, per trade in R (the planned loss of one trade); trade counts under each value. Trades / Win / Avg R / PF / Max DD = every backtested trade on Binance daily candles from 2019 (or a coin's listing). Costs: 0.1% fee per side and 0.02% slippage on every fill. Live caps = only the trades a follower of the live signals could have taken (at most 3 new a day and 6 open, most liquid first). A year without trades shows —. Coin lists are today's (survivorship bias: coins that died are missing, which flatters a long-only system).
Live success = published signals that closed with a profit at their trailing-stop exit, out of those closed. Live signals pass an AI review the backtest can't replay, and trend following wins only about one trade in three, so a handful of live signals says little either way.
Every source is cached for about 60 seconds on the server and fails independently — the UI marks any source that is unavailable. See the disclaimer.
DanioX is built by Daniotech, founded by Asghar Ali. For Pakistan's stock market (KSE-100), see our sister site danioPSX: live PSX market data, news and dividends.