Historical simulation
Current version 2026.09 · Macro + DCA + Adaptive Shield
Backtest on BTC/USDT historical data from March 15, 2018 to July 27, 2026. Starting capital: 2,000 USDT with a fixed initial 60/10/30 split, followed by independent compounding of each bot's settled net results. Lending interest is not compounded. Figures use a conservative multi-exchange baseline. Daily decisions use completed candles only and execute in the following period. Stress costs double current taker fees and add 0.06% slippage. History was used for development; live tracking is the prospective validation.
Simulated performance (2018–2026)
Backtest simulation only, through July 27, 2026. Current taker fees doubled, 0.06% slippage. Past performance does not guarantee future results.
Methodology & transparency
Data and scope
BTC/USDT data from March 15, 2018 to July 27, 2026. Source: exchanges through CCXT, with a shared BTC proxy where native history is unavailable. Shield uses completed signals with delayed execution; fees and slippage are stressed. The replay checks hedge liquidations against intrahour highs and includes liquidation fees plus available funding history. Optional lending and exchange outages are not reproduced.
Stressed costs
The conservative baseline doubles each exchange's current taker fees and adds 0.06% slippage. These are research assumptions, not fees charged by TIYlab.
Chronological validation
Signals use only information available at the close and simulated orders occur in the following period. Walk-forward controls and cost stresses are published, but history remains retrospective development evidence, not a pristine future sample.
Conservative reading
Over the same period, Buy & Hold returns +690.29% with a -76.63% maximum drawdown. In this historical simulation, TIYlab delivers a higher return and materially reduces drawdown. This result is not a promise of future performance.
DEX execution compatibility study
The portfolio backtest above is not presented as a native historical Hyperliquid or Aster backtest. TIYlab separately validates execution compatibility for these venues: spot and perp fees, Hyperliquid UBTC spread/slippage, funding, liquidation rules, precision, minimum size, latency and account mechanics. Periods before a DEX existed are never described as real data from that DEX.
Any replay using a historical proxy is labelled an execution-cost simulation. DEX venues are available in Live mode with account-specific security validation; this availability does not turn historical proxies into native DEX data or alter the CEX live history published since April 2026.
Same Macro, DCA and Shield engines as the portfolio backtest. Historical periods predating the DEX venues use a shared BTC proxy. The stress scenario doubles venue-specific base fees and applies 0.06% slippage per execution. Non-native historical funding uses a conservative proxy.
Hyperliquid is open for Live use after funded-testnet validation. Aster uses controlled per-account Live activation: Spot and Perp reads, Agent permissions and a real Perp lifecycle passed on Testnet. The official internal Spot-funding transfer returned Aster maintenance code 024003, so the funded Spot lifecycle is not presented as validated. Every Aster account must pass a Mainnet Spot, Perp, market-data, authenticated-balance, IP-allowlist and permission preflight before activation. Any failure blocks real orders.
Research history
TIYlab retains prior results whenever the architecture changes. Replacing Funding with Adaptive Shield materially changed simulated behavior. V2's higher result is not independent validation: the same broad historical period contributed to both versions. The versioned live record provides prospective validation.
📡 Follow signals in real-time
⚠️ These results are from a historical simulation (backtest). Past performance does not guarantee future results under any circumstances. Invested capital may be partially or totally lost. TIYlab is a software tool, not financial advice.