This is an explainer of a trend metric, including how this site uses it. It predicts nothing. Nothing here is investment advice. Decisions are yours.
A moving average is the mean of the last N closes, redrawn each bar: a smoothed trail of where price has been. In a 24/7 market the humble detail is the N: fifty daily bars here span seven calendar weeks, not the ten trading weeks an equity chart means, so identical labels describe different lookbacks across markets.
The instrument, from zero
The simple moving average adds the last N closes and divides; the exponential variant weights recent bars more heavily so it turns faster and whipsaws more. Both are smoothing filters: they trade lag for noise reduction, always, with no setting that escapes the trade-off. A price above its MA50 says one thing precisely: the current price exceeds the average of the last fifty closes. Everything else read into the line, support, magnetism, institutional levels, is interpretation layered on arithmetic, some of it self-fulfilling because enough eyes share it, none of it a property of the mathematics.
The close that had to be invented
Equity MAs stand on a natural fact, the session close. Crypto has no session, so its daily "close" is a convention, and in a 365-day market this site's convention is 00:00 UTC. Two consequences follow. First, the calendar divergence: crypto MAs count every day, weekends included, so MA50 spans about seven weeks of real time against the equity MA50's ten, and cross-market analogies ("the 200-day means the same thing everywhere") quietly compare different windows. Second, source disagreement: a venue whose daily bars roll at a different hour computes visibly different MAs on volatile assets, so two readers can argue about whether price is "above the 50" while both are right on their own charts, a cousin of the no-consolidated-tape problem in the universe construction rules. Published methodology, which bar, which venue, which clock, is what makes an MA reading reproducible rather than an opinion.
What moving averages are actually for
The defensible use is state description: above or below, stacked or tangled, rising or falling. That is how the monitor's methodology publishes trend state, price relative to MA50 and MA200 and their order, and how the regime dial builds its BTC condition, close above MA50 with MA20 above MA50. Note what those formulas do not claim: that the crossing of two averages predicts anything. They define a regime boundary, chosen in advance, so that "in an uptrend" stops being a vibe and becomes a computable, arguable statement that commit-reveal attestation can freeze at publication. The golden-cross folklore, MA50 crossing MA200 as an omen, earns its skepticism: sometimes it front-runs a trend, sometimes it top-ticks a chop, and its record is regime stew. Definition, not divination, is the job.
The whipsaw tax and the lag tax
Every MA use pays two taxes. Lag: an average of the last fifty closes registers a turn well after the turn, so MA-based exits give back a slice of every top, the fee trend following BTC pays for catching the middle of moves. Whipsaw: in ranging tape, price crosses an MA repeatedly without meaning it, and each cross exacts costs, spreads, and false starts, which is exactly why breakouts fail in chop. Crypto's volatility widens both taxes: the 4-8% daily ranges ATR measures make casual crosses of even slow averages routine, which is why single-line signals are noise here and why composite conditions, stack order plus a second confirming measure, are the minimum honest structure. An MA is a ruler laid on the past. Rulers are useful. Just do not ask one for the future. Decisions are yours.