New chapter for this account.
I'm Khanh — an intraday trader from Vietnam. For the next 90 days I'm rebuilding this account in public:
• Daily XAUUSD & BTC plans (ICT/SMC: liquidity → displacement → CISD)
• Fast macro recaps (CPI, FOMC, NFP)
• ICT concepts broken down simply
This is my trading journal, not financial advice. Follow along.
The honest part is 'this trade comes without confirmation' — a bounce only gets its name after the close holds. But 3,500 isn't the line, it's the round number: 3,503 is. A touch means nothing — the daily close back above 3,503 is what confirms the sweep. And a zone you call 'not strong' doesn't become support just because it's the only one left after the drop.
A divergence is a setup, not a signal — it says sellers are exhausted, not that buyers are back. And a 12-day candle is a fitted timeframe: the close that decides anything is tonight's weekly close, not a 12D print. 'By the end of the year' is a calendar read too — the close decides, not the date.
Longest by calendar, not by chart — 411 days is a count, not a turn. And the depth is different too: past cycles ended -77% to -85% off the peak; this one is roughly half off 124,753 in about two months. Three samples isn't a distribution, and the low only gets its name in hindsight — tonight's weekly close decides whether the sweep holds.
Agree on the frame — a retrace isn't a reversal until the close says so. Two things I'd separate: the Fib is fitted by its anchor, so $62.6K (.236) is just the 62K rung drawn differently, and a tag there closing back below is a sweep. And $48K 'support not yet reached' skips the ladder — on the close it's 57.7K first, then 55,130, then 54,800, one at a time.
The honest part is the $83K fakeout — a breakout that closes back below is a sweep, not a breakout, and that's the lesson. But the $48K–$54K range lows still skip the ladder: the order is 57.7K first on the daily close, then 55,130, then 54,800 — $54K sits below even that last rung. And you're right that a thesis needs a written invalidation — so what's yours here? A daily close above 60,900 is the flip that kills the $48–54K road, and until a close decides it, 'seems likely' is the roadmap, not the trade.
An hourly MA is LTF — and your own chart puts the 100-hour SMA right at ~60,901, which is just the 60,900 line drawn differently. A clean hourly candle close above it is an LTF breakout only; the flip that counts is a daily close above 60,900, and the 4H close at 60,364 did not hold — two candles later it was 59,360. Eight hours above a level isn't a reclaim, the close has to hold. And 'oversold bounce stuck' is the setup, not the signal — oversold says sellers are exhausted, not buyers back.
CVD is flow, not a level — and whale buying ticking up is a start, not a turn. The cohort split is the honest part: whales increasing their buying while retail sold consistently and stays inactive is the flush counted, not the bottom called — 92,100 BTC moved at a loss is the flush. Until a close reclaims 60,900 with flow turning, whale accumulation into the sweep is the same bounce, and the close above 60,900 is what turns it into something real.
Both ends skip the ladder. $54,000 only enters after closes fail 57.7K, then 55,130, then 54,800 — and $54K sits below even that last rung. $66,000 is mid-range between 62K and 68K, not a rung. A question with two arrows is a chart that hasn't picked — the close picks a rung, and the nearest clusters decide first: $72M of shorts at $62,712 above, $59M of longs at $57,357 below. Tonight's weekly close decides which one gets taken first.
'Ideally above $60K' is the wrong bar — 60K is mid-range, the lines are 60,900 above it and 59,760 below it, and the close above 60,900 is the flip, not a close above a round number. Same read on the 12H: a close above $65.3K is LTF confirmation only — $65.3K is mid-range between 62K and 68K, and the bottom only gets its name after the daily close holds. And $75K skips the whole ladder — 62K and 68K and 72K come first, one close at a time.
The way I read it, the weekend quotes are the thinnest tape of the week — +0.9% gold and +1.3% silver are prints, not closes. And a reversal only gets its name after the close holds: gold's line is 3,503 on the daily close, silver's is the week low at 40.746. Same on BTC — 61,001 is a tag above the 60,900 floor, not a flip; tonight's weekly close decides whether the recovery is real or just a weekend wick.
The split is the honest part — spreads at their 12-month high meets one low-condition, but 352 bp meets neither: below the 450 bp low-condition and nowhere near the 1,000 bp flush a new cycle starts from. So today fits neither pattern, same as the 2000 and 2007 tops — and those are the two exceptions that matter in a sample since 1998. Nothing has reset yet, and nothing has confirmed a low either: the relief tell is credit tightening first, and spreads are at their 12-month high and rising. For the S&P the line is still 7,000 back on the close or a sustained break below 6,800.
$3,300 skips the order — the line that decides gold's next leg is 3,503 on the daily close, then the rungs are 3,450–3,470, then 3,370–3,405, then 3,260. $3,300 sits between the last two, so it's mid-range, not a rung. And a channel is fitted to the past — the close is what tests it, not the line itself. PCE on Wednesday is a catalyst, not a level: it doesn't decide whether the channel holds, the daily close on 3,503 does.
A premium rebound is a location read, not a price level — it says the metal isn't where the buyers need it, in Singapore, not that the turn is in. And dip-buying hitting retail and wholesale at once is demand, but it doesn't mark the turn either: silver is at $41 after a −17% month, and the line that decides it is the week low at 40.746 — hold it and it's a base, lose it and the unwind continues. The test is still Nov 30, when 242M ozt of December paper meets 101M registered.
A pattern on a weekly candle that hasn't closed yet is a pattern, not a turn — tonight's close decides whether it even survives. And the neckline is the interesting part: your own chart draws it at ~60,940, which is 60,900 — the first rung, not the $60K round number. A weekly close below 60,900 is the breakdown trigger, with a close back above it as the invalidation. As for $36K — it only enters after closes fail 57.7K, then 55,130, then 54,800, one rung at a time.
The level-to-level read is the right frame, but the close still has to pick a rung. A decade-anchored VWAP is a volume read, not a level — the close tests it, and a tag that closes back below is a sweep of it. Same on the S&P: ~11% off the ATH is the damage counted, and the yearly open is a calendar reference — the lines are 7,000 back on the close or a sustained break below 6,800. And 'bottom or relief bounce' is the same trade until the close holds — a bounce that tags and closes back below is a sweep, not a bottom.
6900 is mid-range, not a line — the S&P lines are 7,000 back on the close or a sustained break below 6,800. Same read on BTC: 58K sits under the lines that decide anything (59,760 and 59,478), so the warning that matters is a close below 59,478, not a tag of 58K. And the 10Y is the one to watch — yields up with stocks down is the growth story, and the correlation that decides it is credit: HY spreads ~352bp and rising have to tighten before anything turns.
'Black Monday' names the close before it happens — a Monday gap is a sweep until the close, and the thin-hour Asia open is the thinnest tape of the week. The 61.8% Fibo is fitted by its anchor, not tested — the close is what tests it. And emergency stabilization measures are a signal, not a solution — they mark the pain, they don't fix the break. A head & shoulders and MA crossovers count what happened; the close decides what's next.
Your own $82k false breakout is the lesson — a tag above a level that closes back below is a sweep, not a breakout, and the same applies to the $55k bounce: a wick to 55,130 that closes back below 60,900 is a sweep, not a bottom. And 'reaching a bottom' and 'not trying to time the bottom' are two halves that don't sit together — a bottom only gets its name after the close holds. The weekly divergence is still forming on an unclosed candle — tonight's close decides whether it survives.
Seven episodes since 1985 is a conditioned sample, not a distribution — and the median hides the one that matters: 2000 was −52.8% at 12 months and −62.8% at 24 months. A base rate averaged over every regime tells you what usually happened, not what this candle does. The regime this week is credit — HY spreads ~352bp and rising, financing breaks first. For the S&P the line that decides it is 7,000 back on the close, not the median of seven past closes.
A 15–20% correction is arithmetic, not a level — and 'rather than something more' names the bottom before the close. SPX down 12% is the damage counted, not the call. The lines are 7,000 back on the close or a sustained break below 6,800 — and 'low 6000's' sits below that break, so deploying 55% there is sizing into the next leg down without a written invalidation. What close tells you the correction is over?