Satellite Radar (Lead-Lag, 1-3h lag)
What it shows
The satellite radar measures how a sector's satellites (smaller coins) follow its driver (the sector's largest coin) with a 1, 2 or 3 hour lag. For each pair it shows the optimal lag, the strength of the lagged-return correlation, and beta. A map of "who follows whom and after how many hours". Resolution is hourly, so it's an indicative picture, not a recommendation.
How to read it
Start with the driver: when the sector's main coin has just made a clear move, the radar shows which satellites historically repeated it after 1, 2 or 3 hours. Trust most the pairs with high correlation strength and large n, as those have a repeatable pattern. Beta hints at scale: a satellite with β > 1 reacts more strongly (more risk and volatility). Remember the hourly resolution; it's a tendency, not a stopwatch. Practical takeaway: watch the driver's move as the first cue and treat high-correlation satellites as the ones whose reaction is most predictable.
Parameters
- Sector Driver
- The sector's largest, most liquid coin is the one whose moves usually lead the rest.
- Neutral reading: The radar's reference point: watch its move as the first cue for the satellites.
- Lag
- After how many hours (1/2/3h) the satellite most often repeats the driver's move.
- Neutral reading: Shorter lag = faster reaction; it's the time window in which the pattern usually plays out.
- Correlation Strength
- How repeatable the lagged-following pattern is (% bar): a higher value means it is more reliable.
- Positive reading: High: the pattern recurred often, the lead-lag signal is more trustworthy.
- Negative reading: Low: the relationship is weak and random, not worth relying on.
- Neutral reading: Medium: the pattern is sometimes present but not stable.
- Beta (β)
- The satellite's average move per 1% of the driver's move: the strength of the reaction.
- Neutral reading: β > 1 = the satellite amplifies the driver's move (more volatility); β < 1 = reacts more mildly.
- Sample (n)
- The number of observations the pair was computed on: the basis of result reliability.
- Positive reading: Large n: statistics based on many cases, less noise.
- Negative reading: Small n: the result can easily be chance, low reliability.
Example
L1 sector: driver SOL. Satellite SUI has a +2h lag, 81% correlation strength and β 1.4, n=180: when SOL rose, SUI historically followed with a two-hour lag, and more strongly. Ignore another satellite at 22% strength and n=15. The pattern is too weak and based on too small a sample.