Trend & Structure Analysis

SMA-20/50/200: SMA_N = (1/N) * sum of the last N closes. The label comes from the ORDERING of price against the three averages. Price above all three with SMA20 > SMA50 > SMA200 is conventionally read as an uptrend, the reverse as a downtrend. Everything else falls into a residual bucket. Read that bucket honestly: a mixed stack does not establish that price is moving sideways - it also occurs during reversals, pullbacks and volatile trends - so 'sideways' names the absence of a clean stack, not a demonstrated condition.

On the window lengths: 20 trading days is about a calendar month, but 50 is about ten weeks (not a quarter) and 200 about nine and a half months (not a year). A trading quarter and year are nearer 63 and 252 days. In crypto, which trades continuously, all three are simply calendar days. They are conventions, not optimised parameters.

20- and 50-day ranges: the maximum daily high and minimum daily low over the window.

RSI-14, J. Welles Wilder Jr. (1978). Gains and losses are magnitudes: g_t = max(C_t - C_{t-1}, 0), l_t = max(C_{t-1} - C_t, 0). The seed is an arithmetic mean of the first 14 of each, so 15 closes are needed; thereafter each is smoothed recursively and separately, G_t = (13*G_{t-1} + g_t)/14 and likewise for losses. Then RS = G_t/L_t and RSI = 100 - 100/(1 + RS). Wilder's recursion is not a rolling mean; substituting one gives a different indicator.

RSI is BOUNDED BETWEEN 0 AND 100. A figure outside that range is not an RSI, and the prices and averages beside it are not momentum readings. The bounds are real readings - an all-gain window is 100, an all-loss window 0 - but a completely flat window makes both averages zero, the ratio 0/0, and the indicator undefined. There is no universal convention there, so this analysis publishes no reading rather than a midpoint, and the warmup before 15 closes is absent for the same reason. Absent means undefined, never neutral.

The labels - overbought, elevated, neutral, subdued, oversold - are bands at 70, 55, 45 and 30: inherited conventions, not derived thresholds, and crossing one is a change of label rather than an event.

Read the oscillator and the moving-average stack as two readings of ONE series. Both are transforms of the same closes, so their agreement is not independent confirmation and their disagreement is not a contradiction: price can sit above all three averages while momentum has been running against it. That is a fact about the asset, not a fault in either measure.

Swing pivots: a local extreme identified by comparing a bar with its neighbours. A full definition requires the number of bars checked either side, whether highs and lows or closes are compared, how ties are broken, and when the pivot becomes available - and those specifics are part of the method rather than a detail. A CENTRED pivot needs later bars before it can be declared, so it is confirmed with a delay. That is not itself an error: it becomes look-ahead bias only if an analysis treats the pivot as known at its own timestamp. Nearest support and resistance are normally the closest confirmed swing low below and swing high above the current price.

Support and resistance are heuristic reference levels where price previously turned, not barriers with physical force. Their predictive value is contested in the academic literature.