Bitcoin, Ethereum, stablecoins, exchange tokens, and small altcoins were built to do very different things. Yet when the market moves fast, those differences can seem to vanish. Traders stop picking assets one by one. They buy or sell “crypto” as a single block.
The phenomenon has a name: herding, the state in which investor decisions increasingly follow the direction of the market, and asset prices move more uniformly than they otherwise should.
When different assets start moving together
In a normal market, every asset should respond to information in its own way. Strong projects can rally harder, weak ones lag, and stablecoins barely move at all. The larger the market move, the more visible those differences usually become.
Herding inverts that pattern. Once trader attention locks onto the direction of the market, the characteristics of each individual coin start to matter less. A rising price is read as a reason to join the buying. A falling price triggers selling because everyone else appears to be heading for the exit.
For a while, traders stop asking “which coin is worth buying?” and start asking “is crypto going up or down?”
This does not mean every asset delivers the same return. It means their returns cluster more tightly together than you would expect under ordinary market conditions.
How the research detects herding
The study Herding in Different States and Terms: Evidence from the Cryptocurrency Market examined up to 200 crypto assets from July 2013 through January 2020. The researchers compared market behaviour across daily, weekly, and monthly data, and across different market states.
Herding is not observed through trader accounts or individual transaction histories. The researchers used a measure called Cross-Sectional Absolute Deviation, or CSAD, to see how far each coin’s return sits from the market average return.
- High CSAD means returns are dispersed and coins are moving more independently.
- Low CSAD means returns are tightly clustered and coins are moving more uniformly.
- CSAD failing to rise the way it should when the market makes a large move is the signal consistent with herding.
The method matters, but it has limits. Uniform movement can appear because traders are following the crowd — or because every trader is rationally responding to the same piece of information.
The herd effect moves fast
The study’s headline finding shows up in the difference between observation windows.
- Daily data: herding detected.
- Weekly data: herding not generally detected.
- Monthly data: herding not generally detected.
That pattern suggests the herd effect is primarily a short-horizon phenomenon. When the market moves quickly, traders react to the overall direction. Given a little time, asset-specific information starts getting priced in again and the moves become more differentiated.
Put plainly: on a day full of stress or euphoria, traders can treat hundreds of assets as one and the same trade. Over longer horizons, the distinctions between Bitcoin, altcoins, stablecoins, and other tokens begin to reassert themselves.
Herding is not born of panic alone
In the daily data, herding shows up when the market is falling but is not generally present when the market is rising. That makes intuitive sense: when prices drop, a trader’s objective can shift from finding opportunity to reducing risk. Selling happens broadly, and the differences between projects matter less.
But the research also found that herding can emerge during some periods when US equity market volatility, as measured by the VIX, sits at low levels. In other words, herd behaviour does not always originate in fear.
A calm market can make traders more comfortable taking risk. Rising prices attract attention, attention triggers buying, and buying reinforces the rally. Under those conditions, FOMO can bind asset movements together just as tightly as panic does.
What needs separating is moving in the same direction and herding in the statistical sense. In a crash, many assets can fall together but by wildly different magnitudes. If one coin drops 10% and another drops 60%, the direction is shared but the dispersion of returns is still enormous.
When diversification only looks like diversification
Holding five or ten crypto assets does not automatically mean your risk is spread. If every asset responds to the same market factor and moves more and more uniformly, that portfolio really carries one large exposure: the direction of the crypto market.
The research also found patterns consistent with short-term reversal. Highly uniform positive moves tended to be followed by weaker subsequent returns, while highly uniform selling could be followed by recovery. One explanation is that crowd buying or selling pressure pushes prices too far, and part of the move corrects once that pressure ends.
This is not a ready-made trading strategy, though. The research does not prove those reversals remain profitable after transaction costs, spreads, slippage, liquidity differences, and out-of-sample testing.
What traders should be watching
When every chart starts looking the same, the question is not only which asset is rising or falling. The more important question is whether decisions are still grounded in the characteristics of the asset, or only in the direction of the crowd.
- Several coins can represent one and the same risk.
- A market-wide rally does not prove every project has strong fundamentals.
- Correlation can rise exactly when diversification is needed most.
- Uniform movement is information about risk, not an automatic buy or sell signal.
Herding does not make asset analysis useless. Quite the opposite. When the market stops distinguishing one coin from another, traders need to be more alert to the possibility that a portfolio which looks varied is really the same trade wearing different names.
Primary source: Kyriazis, N. A. (2022), “Herding in Different States and Terms: Evidence from the Cryptocurrency Market,” Journal of Asset Management.
