Part 1 of this series covered EMA, VWAP, and Volume Profile — three tools built directly from price and volume math. This part moves into more discretionary territory: tools that lean more on pattern recognition and interpretation, and where the gap between "well-established" and "popular but unproven" matters a lot more. Worth being upfront about which is which before going further.
RSI: momentum, with a real statistical pedigree
The relative strength index, developed by J. Welles Wilder in 1978, measures the speed and magnitude of recent price changes on a 0–100 scale — essentially a speedometer for momentum rather than price itself. The standard lookback is 14 periods, and the two levels worth knowing are above 70, conventionally read as overbought, and below 30, read as oversold.
The mistake worth flagging early: treating those levels as automatic reversal signals. In a genuinely strong trend, RSI can sit above 70 (or below 30) for a long stretch without price reversing at all — a stock in a powerful uptrend can stay "overbought" for weeks. Shorting purely because RSI crossed 70 in a strong uptrend is one of the more common ways new traders lose money to an indicator that was actually working correctly; the market just wasn't in the regime that signal is built for. RSI extremes are more reliable in range-bound or choppy conditions than in strongly trending ones, and pairing RSI with a trend filter like the EMA (from Part 1) — only treating overbought/oversold as a signal when the broader trend isn't strongly against you — meaningfully improves its reliability.
Divergence is the other core RSI concept: if price makes a lower low while RSI makes a higher low (bullish divergence), momentum is quietly strengthening even as price falls — a signal, not a guarantee, that the move down may be losing steam. The reverse (price higher high, RSI lower high) is bearish divergence. Divergence is a genuine, widely studied phenomenon in momentum indicators, though like everything else here it works better as one input among several than as a standalone trigger.
Killzones and AMD: popular, but read this part carefully
"Killzones" refer to the highs and lows set during specific trading sessions — Asia, London, New York — on the idea that each session's extremes mark real liquidity (clustered stop-losses and pending orders) that price is statistically likely to target before reversing. The broader framework built around this, often labeled AMD (accumulation, manipulation, distribution), describes a full-day narrative: a quiet accumulation phase, a manipulative move that sweeps one session's liquidity in the "wrong" direction, then a distribution phase that reverses hard the other way.
This framework — commonly associated with the trading educator known as ICT — is genuinely popular in retail trading communities, and the underlying logic (that clustered stop-losses represent real liquidity worth targeting) is directionally reasonable and echoes real institutional order-flow behavior. But unlike RSI or VWAP, it doesn't rest on a peer-reviewed formula or decades of institutional literature — it's a pattern-recognition framework, and its track record is much harder to verify than a moving average's. If you're drawn to it, the right next step isn't to trust it on faith — it's to backtest it the way the earlier piece on this site walks through: define session highs and lows precisely, code the AMD pattern as an explicit, nondiscretionary rule, and see whether it actually produces a statistical edge on real data before risking anything on it live.
Support and resistance: the oldest tool, still worth using carefully
Support and resistance mark price levels where past buying or selling was concentrated enough to plausibly repeat — support below current price, where buyers have previously stepped in, and resistance above, where sellers have. They're really a simpler, more visual cousin of the high-volume nodes covered in Part 1: both describe prices the market has previously treated as significant.
Two things are worth taking seriously here. First, a support or resistance "level" is really a zone, not a precise price — treating it as an exact number invites getting stopped out by ordinary noise around a level that's still holding in a broader sense. Second, a level breaking is meaningfully different from a level holding, and the difference matters for how you trade it: price rejecting a level and reversing suggests the level is still respected, while price breaking through on strong volume suggests new momentum entering the market and the old level potentially flipping roles — old resistance becoming new support, or vice versa. Round numbers and prior session highs/lows tend to act as psychological levels for the same underlying reason liquidity clusters around them in the first place: enough traders are watching the same obvious price that it becomes somewhat self-fulfilling.