Lumi Trading research workflow: backtesting, simulation and risk checks

Understand the current product stage

Treat Lumi Trading as a research product in internal testing rather than an available live-trading service. Website workflows describe research direction; access and downloads follow the current official entries. Confirm technical development services separately from product availability. A demonstration page does not prove that a market or broker is connected.

Inspect historical-test assumptions

Check data sources, missing records, adjustments, time zones and possible look-ahead leakage. Document fees, slippage and execution assumptions, and distinguish in-sample tuning from out-of-sample evaluation. Results may change materially with these assumptions, so explain a performance chart alongside its data, parameters and test conditions.

Use simulation to examine the process

Simulation can help evaluate signals, order states, logs and exception handling, but cannot fully reproduce market liquidity, latency or execution. Preserve strategy versions and experiment records, and examine recovery, duplicate signals and boundary cases. These are evaluation methods, not a claim that every research tool is available in the current version.

Separate software from investment decisions

Technical platforms and development services help build and test processes; they do not determine investment suitability or guarantee capital or returns. Any future connection requires separate checks of data rights, account authorization, risk limits and manual stopping procedures. Do not make real-money decisions from an article example or a simulated metric.

Frequently asked questions