Module 2F
Session times and the Silver Bullet windows
Leads into Phase 3, where the read you build here becomes something you act on.
The window discipline and the read sequence are this curriculum's own. Liquidity, sweeps and locations build on 1B, 2B and 2D.
This is where the previous modules come together
Up to now you have learned the pieces one at a time. Operator behaviour and liquidity pools in 1B. Conditions in 2A. Structure in 2B. Locations and states in 2D. Day type in 2E.
2F is the first place they get applied together, in context, inside a single window. It is a worked example of the read rather than a new concept, and if the earlier modules are not solid this one will not make sense. Go back rather than pushing through.
- Silver Bullet window
- The defined one-hour windows, primarily 10:00 to 11:00 ET, where institutional activity is most readable.
- Liquidity sweep
- A fast spike through a pool of resting stops that then reverses.
- BSL / SSL
- Buy-side liquidity above the highs, sell-side liquidity below the lows. The pools a sweep targets.
- Flow confirmation
- Order-flow evidence, delta turning, absorption, compression, that a swept level will actually reverse.
- Map feature vs trigger
- Locations mark where to look. The flow read there is whether to act.
- Confluence
- Several independent signals lining up at one spot to create a high-probability moment.
Time is a filter before price is a signal
Before you can execute anything you need to know when the market is most likely to deliver a clean, readable move.
The Silver Bullet is not a pattern and it is not an indicator. It is a time-based filter that anchors attention to defined one-hour windows. Most beginners trade all day because they do not know which hours are structured and which are noise, and the window solves exactly that problem and nothing else.
Why a window would matter at all
Markets are not uniformly active. Institutional participants work to structured schedules tied to session opens and overlap periods, and when orders of significant size get executed they leave readable evidence behind: sweeps, displacement, and footprints in the order flow.
That is the whole claim, and it is a modest one. Not that a hidden mechanism activates on a timetable, but that more size trades at some hours than others, and size is what leaves a trail you can read. If a window ever stops producing readable sequences, the honest response is to notice that rather than to keep believing in the hour.
What you are actually looking for
The window is where you learn to identify the area for expansion: the zone and moment where price is most likely to expand in the intended direction.
Intended direction means the higher-timeframe draw, where price is ultimately being pulled, set by your bias read. Not wherever the next spike happens to point.
That expansion can resolve three ways, and at this stage your only job is to recognise which.
The three windows
| Session | New York time |
|---|---|
| London | 03:00 to 04:00 ET |
| New York morning | 10:00 to 11:00 ET |
| New York afternoon | 14:00 to 15:00 ET |
All times New York. Deliberately not given in UTC, because the offset changes twice a year with daylight saving and a fixed UTC table is wrong for several months of it.
For ES and MES on a New York schedule, the 10:00 to 11:00 window is the primary one. Post-open volatility has settled, the initial order flow is established, and participants are actively managing positions. It produces the cleanest sequences.
The afternoon window is secondary. It aligns with the London close and often sees a reversal or continuation as European participants close out. The London window is valid and mostly irrelevant if you are running a New York day.
The window defines the universe of valid setups. Outside it, the model does not apply. You are not looking for this at 09:15 or at 11:30. You wait.
And the window has a second job. It is the confirm-or-deny mechanism for the bias you formed in the first thirty minutes. Either a clean sequence validates it, or the failure to set up invalidates it. Both are information.
Three outcomes, and recognising which
Expansion, or continuation
Price pushes in the direction it is already being drawn and keeps going. The existing bias plays out.
Reversal
Price first runs the opposing liquidity, traps whoever chased it, then expands the other way, back toward the higher-timeframe draw.
Balance
Price rotates with no real commitment. Nothing is being delivered. This is a stand-down and it is a completely valid outcome for a session.
This module is not teaching you to trade any of these. Execution is Phase 3. Here you are training your eye to read the window. The walk-through below uses a reversal as the worked example because it puts every piece on display in one sequence. The same pieces, read differently, are how you recognise the other two.
Before the window: context
Done before the window opens, from your daily bias and your opening type read. This tells you whether the environment is even structured to deliver today.
- Higher timeframe structural read: bullish, bearish or balanced
- Prior session profile reference: value area high, value area low, point of control
- Where price opened relative to yesterday's value and range
- Overnight activity and any bias it built
Before the window: map the locations
These are the map features. Places where the question gets asked when price arrives, and never entries by themselves.
- Buy-side liquidity above recent swing highs, equal highs, prior session highs
- Sell-side liquidity below recent swing lows, equal lows, prior session lows
- Prior session value area high, value area low and point of control, from 2A
- VWAP, the session's volume-weighted fair price, from 2A
- Low volume nodes from previous sessions, which are travel zones rather than reaction zones
- Order blocks and fair value gaps, from 2D
- Round numbers and other psychological levels
This step is done before 10:00. It is preparation, not a reactive decision inside the window. Opening your chart at 10:02 and trying to work out where liquidity is while the move is already happening means you are reacting rather than reading, and the difference shows up in what you take.
Inside the window: the sequence
A sweep
Watch whether price probes one of the pools you marked.
A sweep is a purposeful move into a liquidity pool that takes out the orders resting there before the market reverses. It is aggressive and fast. It looks like a breakout or a breakdown, and it reverses quickly.
This is the behaviour 1B named. After a sell-side sweep in a bullish context, someone with size has bought at a discount using those stop orders as their fill. After a buy-side sweep in a bearish context, someone has sold into the buyers chasing the breakout. The stops are the counterparty. That is the whole point of running them.
The sweep alone is not an entry signal. It is confirmation that the model is potentially activating.
The common error: entering the moment a low breaks, on the assumption that breakdown means sell. The sweep is often the last thing that happens before a reversal, so entering in its direction is entering into the hand of the participant who caused it rather than alongside them.
Read the flow at the level
After the sweep, do not jump to the candle chart looking for a structure shift as your trigger. That is confirmation, not the trigger. The trigger is what the order book and delta are doing while the sweep is happening and immediately after.
What you are reading:
- Cumulative delta direction. After a sell-side sweep you expect to reverse upward, delta should be turning up, or already turning while the sweep finishes. If delta is making new lows while price sweeps, the move is real and your reversal hypothesis is wrong
- Delta swings. Strong swings in the reversal direction confirm aggression on the right side. Weak or absent delta means nobody is committing. Wait
- Absorption. Large resting orders defending the level, aggressive orders hitting that wall without breaking it. The most direct evidence that size is positioned against the sweep
- Volume at the sweep level. High volume and large prints mean real participation. Low volume sweeps are more likely to keep going, which is 2D's tell arriving in practice
- Compression after the sweep. If price balances immediately after with bid and ask at equilibrium, positioning is being gathered before the expansion. If price keeps going in the sweep direction, the hypothesis is failing
- Tape speed. A sweep that finishes with the tape slowing and then re-accelerating the other way is confirmation in real time
- Auction state. Is the auction exhausted here, having tried and failed to find new participants, or still strong? Exhaustion supports reversal. Strength supports continuation. That is 2A's language doing work
No flow confirmation, no trade. This is the hard filter. A sweep without delta turning, without absorption, without compression is a sweep that will keep going. Stand aside.
A structure shift, and where it actually sits
After the sweep and the flow read, a structure shift may appear on the candles: a confirmed break in the opposite direction to the sweep. After a sell-side sweep it is a bullish change of character, and the mirror after a buy-side sweep.
The structure shift is the candle chart catching up to what the order flow has already told you. In this framework it is one valid confirmation among several, not a required gate. It sits in the confluence stack alongside delta direction, delta swings, volume, absorption, compression, tape speed and auction state.
- With three or four aligned factors at the level, you have a valid read whether or not the shift has printed yet
- If it forms, treat it as one more factor strengthening the read
- If it does not form but other factors align, the read can still be valid
- If no factors align, the sweep is just a sweep. No trade, shift or no shift
The deeper point. The candle chart is a visualisation of what order flow already did. Treating the structure shift as the trigger means waiting for the slowest confirmation in your stack, and by the time it prints the best entry has often gone.
The wrong framing: no shift, no trade, full stop. That collapses a multi-factor stack into a single-factor candle dependency, which is the same error as treating a fair value gap as a trigger rather than a location.
The right framing: three or four aligned factors is a valid read, and the shift, if and when it forms, is one of them.
When the window does not deliver
These are the signs the window is balancing or unclear. There is nothing to read for a trade and the right response is to stand aside.
- The window closes with no sweep at a mapped location. The model did not activate
- A sweep occurs but flow does not confirm. Delta keeps going in the sweep direction, no absorption, no compression. The sweep was real, not a trap
- Multiple sweeps in quick succession. Clean moves are usually clean. Price whipping both sides of liquidity inside one window is noise
- Higher timeframe bias conflicts with the sweep direction. Low conviction at best. Skip it
- Rotation. Delta near zero, small swings, price accepted in a range with no need to expand. The stand-down signal
- Compression fails to break. Price sweeps, compresses, then breaks back into the sweep direction. The trap was the failed reversal, not the original sweep. Move on
News events suspend the model
The model assumes normal price delivery. High-impact macro releases interrupt that assumption. During and immediately after a major release, price is driven by a binary fundamental catalyst rather than by the liquidity sequencing this model reads, so the sequence becomes unreliable and entries carry unquantifiable spike risk.
The rule is binary and non-negotiable. If a high-impact event is scheduled within thirty minutes before, or during, the window, the model is suspended for that window. No trade. The session is logged as calendar-suspended.
Events that trigger suspension: FOMC decisions and press conferences, CPI and core CPI, PPI, non-farm payrolls, GDP, and any red-flagged event on the calendar between 09:30 and 11:00 ET. Apply the same rule to the afternoon window, where FOMC decisions often land at exactly 14:00.
Check the calendar every morning as step one, before you look at a chart.
Observing a suspended session without trading it is encouraged. How price behaves during and after a major release is context you cannot get any other way, and logging it costs nothing because you were never going to trade it.
Self-check
Grade yourself honestly
1. Why does a time window produce more readable moves than an arbitrary hour?
A correct answer is modest: more size trades at some hours than others, and size leaves a readable trail. If your answer invoked a mechanism that switches on at a set time, you have described something you cannot check. The honest version also implies its own falsifier, which is that if a window stops producing readable sequences you should notice rather than keep believing in the hour.
2. A sweep happens at your mapped level and cumulative delta keeps making new lows. What is that telling you?
A correct answer says the move is real and the reversal hypothesis is wrong. Stand aside. This is the single most useful line in the module, because it is the one that tells you when your own idea has been falsified while it still feels like it is working.
3. No structure shift has printed, but delta has turned, absorption is visible and price is compressing. Is that a valid read?
A correct answer is yes, and explains why: the candle chart is a visualisation of what flow already did, so waiting for it means waiting for the slowest confirmation you have. If your answer was no, you are holding the single-factor version of the model, which is the same error as treating a gap as a trigger.
4. The window closes with no sweep at any location you marked. What happened?
A correct answer is nothing happened, and that is a complete outcome. The model did not activate. If your answer involved finding something else to do, note that. Most windows do not deliver, and a session with no trade in it is the normal case rather than a wasted morning.
If question 2 was the hard one
Go back to the flow section before moving on. Everything else in this module is preparation. That one question is the entire live skill: knowing what evidence would tell you that you are wrong, and being willing to see it while the trade still looks available. Phase 3 is unteachable without it, and every give-back and every moved stop in trading traces back to not having an answer to it.
One rep before you continue
Watch five windows and take nothing
In replay, step through the 10:00 to 11:00 window on five separate days.
For each one, mark your liquidity locations before letting the window play, then watch it. Write three lines. Did a sweep occur, and at which location. What followed it. Which of the three outcomes was this: expansion, reversal or balance.
Where no sequence formed, write what was missing. The sweep, the flow, or the structure.
You are not looking for trades and you should not take one. You are training the recognition that Phase 3 executes on, and the specific thing being trained is telling a window that delivered from a window that merely moved.
What comes next
Reading the window is where this module stops. Acting on the read is Phase 3, once you can identify the sequence reliably enough that you are not deciding what it was after the fact.
Two modules complete Phase 2 first: the daily bias protocol that builds the thesis this window tests, and the pre-trade write-up that makes the whole thing something you committed to before the open rather than something you assembled during it.
Log the module
Where did you get stuck?
Not a test, and nothing is marked. This is the module where every earlier piece has to come together at once, so it is the most likely place to discover that one of them did not land. About a minute.