Derivatives Data Essentials: Part 2. Covering Liquidations, Volume Delta, Long-Short Ratio & Taker Buy Sell.
Welcome back. This is the second part of Derivatives Data Essentials. Please check out part one if you haven't already, it covered Open Interest, CVD, Funding and Orderbook Depth. In this part we move on to some less widely used things. In this one I'm going to talk about the following:
Liquidations happen when a trader’s margin is no longer sufficient to hold their leveraged position. When this threshold is hit, the exchange forcefully closes the position using a market order. This creates forced flow, not discretionary flow, which is why liquidation-driven moves often happen fast and overshoot levels.
Essentially they give insight into three things:
A big drop in OI during a move usually reflects long liquidations (if price is falling), and the same holds true for short liquidations (if price is going up). This is why OI shouldn't be viewed in isolation.
Liquidations often stack in clusters because traders tend to use similar leverages, put their stop-losses in the same place (swing highs and lows for example), etc..., and when price trades into one of these clustered zones it can create:
Crypto is especially prone to these events due to high leverage, high volatility and thin books.
Since liquidations are executed as market orders, when the book is thin there isn't enough resting liquidity to absorb those orders at nearby prices. Thus, the forced order has to climb further up or down the book to get filled, creating an outsized movement.
This is why liquidation cascades almost always happen during periods of:
In short, forced market orders + thin liquidity = exaggerated moves. This is the core mechanic behind liquidation cascades.
The raw time series can jump around a lot. Some assets naturally show larger liquidation values than others (more volume, more participants, higher OI, etc), plus high-vol regimes can skew everything further.
So what do you do to normalize it?
Well one way is the Z-Score:
zscore = (value - mean) / standard deviation
This transforms the metric into a relative measurement of extremeness, rather than an absolute value.
Well that brings us to why would you do it?
It's simple:
This is especially useful for:
In short it's giving you insight into whether the liquidation is strong relative to the last N periods. Even if the raw number looks small.
For liquidation Z-Score @abetrade has a great script for this. No BS plotting, simple & effective. I'd recommend anyone reading this guide should check it out. You can test out the concept and see where it fits into your system/style of trading.
Volume Delta measures the net difference between market buys and market sells within a single candle. Unlike CVD, which accumulates this over time, volume delta is bar-to-bar change. Basically giving insight into who controlled each candle and spotting short-term shifts in aggression that often occur during larger moves.
Put simply:
Some people can confuse the two, so here's a clean distinction:
CVD is the macro view of flow.
Volume Delta is the micro view.
Both are useful for different things.
Just like CVD, delta only becomes meaningful when paired with price. Here are the main interactions
Divergences with Volume Delta occur when price and delta move in opposite directions:
Delta becomes even more useful when viewed in clusters rather than single candles. Multiple bars showing:
Often indicate initiative behaviour.
Examples:
The Long-Short Ratio shows the proportion of accounts that are holding long positions versus those holding short positions on an exchange. It’s often viewed as a sentiment gauge, but it has major limitations and should never be treated as a directional signal.
In simple terms:
That’s it. It measures users, not position size, not conviction, and not leverage.
Because it counts the number of accounts, not the notional size of their trades, the Long-Short Ratio mostly reflects retail sentiment. Larger players, systematic traders, and hedgers barely register in this metric.
A high ratio often means:
A low ratio means the opposite: crowd leaning short.
But again, this is the number of accounts, not how much risk is on the table.
There are a few reasons traders treat this as background noise rather than a core signal:
For these reasons, the ratio tends to be used as a soft contrarian tool, not a predictive indicator.
Even with the limitations, the Long-Short Ratio has some practical uses:
Example:
If Long-Short Ratio is extremely high but OI is flat, it probably means a lot of small retail longs with insignificant size. Not a strong signal.
If Long-Short Ratio is extremely high and OI is rising fast and funding is positive, then you know retail is piling in with leverage. That’s when traps form.
There are only a few situations where Long-Short Ratio adds real insight:
Used alone, it’s weak. Used with context, it helps describe crowd bias.
Taker Buy/Sell Volume measures how much of the executed volume came from buyers crossing the spread (market buys) versus sellers crossing the spread (market sells). It’s one of the cleanest ways to see who is being the aggressor at any moment.
Unlike Long-Short Ratio, this is actual executed flow, not sentiment. And unlike CVD, it doesn’t accumulate over time—it simply shows the breakdown of aggression within a selected window.
In simple terms:
Whichever side is higher is controlling the immediate flow.
Because takers move the market, this data tells you:
This is much more direct than long-short ratios or funding because it is based on actual traded volume.
Here is how price and taker flow interact:
Think of it as a cleaner, shorter-term version of delta without the need for full footprint charts.
Most platforms present this as a ratio:
Ratio = Taker Buy Volume / Taker Sell Volume
This ratio is very fast to react, making it useful during breakouts, news-driven moves, or early-session volatility.
Taker flow is most effective when paired with structure or key reference points:
Taker data is very commonly used by short-term traders, but it scales well into higher timeframes if averaged or smoothed.
To prevent confusion:
Taker flow is the cleanest, because it is not affected by candle openings or closings.
That wraps up the second part of the derivatives data essentials. These tools are useful for understanding how positioning builds, unwinds, and reacts under pressure, but they're still just the context. Liquidations, volume delta, taker flow, and long-short ratio help you see the mechanics behind a move, not predict the move itself. Just like the metrics in part one, none of them should be traded in isolation. The best way to get real value out of these is by making a system, backtesting it so you can really understand the mechanics of your trading strategy and know whether it's actually +EV.