Hypotheses
Quantiacs LSTM State Strategy: Faithful Specification, Currently Infeasible on This Platform
The supplied code is a tradeable, long-only NASDAQ-100 allocation strategy, not merely an LSTM utility. Credit: Quantiacs; licence reported as MIT in the brief. It trains separate neural networks for 12 low-volatility…
Hypothesis
The supplied code is a tradeable, long-only NASDAQ-100 allocation strategy, not merely an LSTM utility. Credit: Quantiacs; licence reported as MIT in the brief. It trains separate neural networks for 12 low-volatility eligible stocks and converts their raw outputs into daily portfolio weights smoothed against previous weights. A faithful executable port is currently infeasible: the documented equity feed does not establish availability of Quantiacs' historical NASDAQ-100 eligibility/is_liquid series, and the strategy sandbox forbids imports of torch, qnt, xarray and pandas. Do not substitute today's constituents, a hand-picked basket, a crypto universe or another model. This specification describes the supplied source; its performance is unverified.
Analysis
strategy-ml_lstm_state
Backtest and paper results are hypothetical. Trading involves risk of loss.