Stop runs and liquidity pools
Lesson 18 · about 10 min
Module 1 noted that stop orders are not in the book until they trigger, at which point they become market orders. That single fact explains one of the most frustrating patterns in trading: price spikes through an obvious level, fills every stop sitting there, and reverses. This lesson explains why that happens as a matter of mechanics, where stops cluster, and how a large participant uses them.
Where stops sit
Traders place stops in predictable places, because the textbook tells them to:
- Just below a swing low (for longs).
- Just above a swing high (for shorts).
- Just beyond round numbers.
- Just outside a range that has held several times.
- Just beyond the previous day's high or low.
Because thousands of traders use the same logic, stops cluster at these prices. A cluster of buy stops above a swing high is a liquidity pool: a large amount of aggressive buying that will happen automatically if price gets there.
Price | Resting sells | Stop buys (invisible) | What is really there
--------+---------------+-------------------------+------------------------
5015.00 | 40 | |
5014.75 | 55 | |
5014.50 | 60 | ~ 1,400 buy stops | <- above swing high
5014.25 | 80 | ~ 600 buy stops |
5014.00 | 210 | | <- swing high
5013.75 | 180 | |
The DOM shows 60 and 80 contracts offered at 5014.50 and 5014.25. What it does not show is roughly 2,000 contracts of buy stops parked there by shorts. If price ticks to 5014.25, six hundred market buys fire, then fourteen hundred more, and there are only 195 contracts of offers to absorb them within three ticks. Price spikes.
Why large participants want the run
Now imagine you need to sell 2,000 contracts and you would like to do it at a good price without moving the market against yourself. Above the swing high sits 2,000 contracts of guaranteed aggressive buying. If you can get price to 5014.25, those buyers will come to you. You lift enough offers to get price there (which costs you some contracts of buying), the stops fire, and you sell into them at 5014.50 and above. You have sold your 2,000 at the top of a spike that you started.
That is a stop run. It is not a conspiracy and it is not illegal; it is a participant who needs liquidity going to where the liquidity is. Stops are the largest and most predictable source of aggressive liquidity in the market.
Key idea: A stop cluster is a pool of guaranteed aggression that any participant can see is probably there. Large sellers go to buy stops to sell; large buyers go to sell stops to buy. Price is drawn to liquidity, not to lines.
What a stop run looks like in order flow
The signature is distinctive once you know it.
| Phase | Price | Footprint | Delta | DOM |
|---|---|---|---|---|
| 1: approach | Grinds toward the high | Moderate buying, nothing unusual | Mildly positive | Offers thin above |
| 2: trigger | Ticks through the high | Burst of buy prints, stacked buy imbalances | Sharply positive | Offers consumed fast |
| 3: absorption | Stalls 2 to 4 ticks above | Very heavy volume, buying meets a wall | Still positive but price stops | Large offers appear or hidden size reloads |
| 4: reversal | Drops back through the high | Selling takes over, finished high | Turns negative | Bids pulled below; fast |
Phase 3 is the tell. Heavy buying that does not move price, just above a level everybody was watching, is the stops being sold into. The reversal in phase 4 is fast because the shorts who were stopped out are gone, the late longs who chased are trapped, and the seller who wanted the fill has it.
A stop run that does not reverse, where phase 3 never happens and price keeps accepting higher, is a genuine breakout. The difference is whether a wall appears. That is why breakouts cannot be judged from price alone; the next lesson takes this further.
Sizing the pool
You cannot see stops. You can estimate them:
| Clue | Suggests |
|---|---|
| Many bars of consolidation under a high | More shorts placed, more stops above |
| Round number just above the high | Retail stops cluster there; larger pool |
| High is visible on daily charts | Multi-day participants' stops; much larger pool |
| Recent failed break already ran the stops | Pool is partly drained; smaller |
| Level tested many times and held | Very large pool; obvious to everyone |
The last row is the important one. The more obvious a level, the bigger the pool above it, and the more attractive it is to run. "Everyone can see it" is an argument that a level will be tested, not that it will hold.
Protecting yourself
Three practical adjustments:
- Do not put your stop where everyone else's is. A stop a few ticks beyond the obvious cluster survives the run. It costs a little more per trade and saves the trade far more often than it costs.
- Treat the first break of an obvious level as a test, not a signal. Wait to see whether phase 3 appears. Absorption just beyond the level means run, not breakout.
- Look for the run as an entry. A completed stop run (phase 4 confirmed, price back inside the range) is one of the highest-quality reversal entries in order flow, because the participants who could push further have just been removed.
Try it: Find five instances on your product where price spiked through an obvious swing high or low and reversed within a few bars. For each, look at the footprint just beyond the level. Was there absorption (heavy volume, no progress) in the spike? Then find five instances where the break held. Was the absorption absent? Write down the counts.
Recap
- Stops are invisible market orders that cluster at obvious places: beyond swing points, round numbers, range edges and prior day extremes.
- A cluster of stops is a liquidity pool: guaranteed aggression that large participants use to fill size.
- A stop run has four phases; absorption just beyond the level is the tell that distinguishes a run from a breakout.
- The more obvious the level, the larger the pool and the more likely it gets tested.
- Place stops beyond the cluster, treat first breaks as tests, and use completed runs as reversal entries.
See it drawn
Original diagrams for the ideas on this page. Illustrative, not real market data.