Start with the data timestamp
Crypto prices come from Binance public market data; US and Korean stock charts come from Yahoo Finance. The tracked universe is limited to 24 crypto assets, 30 US stocks, and 30 Korean stocks. Data providers may delay or revise prices. Stock pages can display the previous trading day's close during a weekend or market holiday. Check the last successful scan time and the source bar time before treating a card as current.
Conditions are evaluated from completed daily bars, not from a still-moving intraday bar. The displayed current price may come from a different time than the close used to detect a condition. Missing, reversed, future-dated, stale, or invalid price bars can be rejected or flagged, so a missing condition does not prove that the market was unchanged.
Detection price is not an executed trade price
The detection price is the reference close of the bar where a rule became true. It is not a price at which Signal Market, or any visitor, necessarily bought or sold. A later current price is a second observation. The percentage difference between the two is not an investor's realized return: execution time, position size, spread, fees, taxes, and currency conversion would also matter.
Illustrative example: a condition is detected at a reference price of 100 and the later display shows 104. This means the displayed price changed by 4% between those observations. It does not establish a 4% trading profit.
A trend strategy and a timing strategy can record separate conditions for the same asset. Crypto long and short labels describe opposite simulated directions. The stock pages do not imply that a downward move is an actual short sale. An exit label tells you why a model state ended; it is not, by itself, a gain or loss.
LIVE history and backtests answer different questions
LIVE condition history contains conditions detected and stored from public prices after the current rules were put into service. It is not brokerage execution data. A backtest replays rules against earlier prices. Different start dates, asset universes, and market regimes mean the two series should not be joined into one continuous real-money equity curve. Open simulated positions should not be included in completed-trade win rates.
In the site's simulations, detection on a completed bar and a hypothetical entry at the next tradable price can differ. The model includes assumptions for entry and exit costs: 0.15% each way for crypto, 0.06% for US stocks, and 0.20% for Korean stocks. It accounts for the gap to the next open, but does not add a separate slippage allowance, tax, currency-conversion cost, or funding cost. These are model assumptions, not a quote from a broker or exchange.
Use several performance measures together
| Measure | What it shows | Important limit |
|---|---|---|
| Win rate | Share of completed simulated trades with a positive result after modeled costs | A high rate can coexist with large occasional losses |
| Profit Factor | Gross positive returns divided by the absolute value of gross negative returns | Unstable with a small sample; undefined without a losing observation |
| Maximum drawdown (MDD) | Largest peak-to-trough decline of the completed-trade simulated equity curve | Does not capture every intraday or open-position loss |
| Sample size | Number of completed observations in the chosen market and period | Results can shift markedly between bull, bear, and sideways markets |
For example, seven +1% results and three −5% results have a 70% win rate but a negative simple sum. Compare the completed sample count, testing period, market regime, cost assumptions, and drawdown before interpreting any headline percentage. A benchmark such as BTC, SPY, or KOSPI gives context but is not a risk-matched alternative.
What to do when a number looks wrong
Record the asset, page URL, data timestamp, and the value you expected, then use the contact page. Data gaps, symbol changes, new listings, trading halts, and provider revisions can change or suppress a condition. The dashboard does not independently verify every exchange announcement or model your personal portfolio.
This material is general market information. It does not account for your financial situation, risk tolerance, or investment horizon, and it cannot predict future returns. For the source market views, use the links below.