This article has been written with the benefit of insights provided by a seasoned trader, on Twitter as @CypherSpook
There is a wealth of literature in behavioral finance that suggests the relevance of psychological biases that affect decision making in investment. There are far too many to make mention here, but perennial favorites include the winner’s curse, which describes the outcome when the winner overbids for an asset in an auction; loss-aversion, which is the key to understanding why we systematically overweight losses compared to equivalent gains and, consequently, it feeds into insight over the well-established equity risk premium, and, of course, herd behavior, or the tendency of investors to congregate and base their own behaviors on the decisions made by others.
Relatedly, there are firm reasons to suspect that there are real biological bases for decision-making. Broadly, the role of the prefrontal cortex in the human brain in making sound decisions has long been established; specific to investment, it has even been suggested that a site within that region — the ventromedial prefrontal cortex — plays a role in modulating our vulnerability to the ‘money illusion’, or the propensity for holding exaggerated valuations for an asset.
What is often missed in such academic investigations are some of the finer mechanisms at play that rely on the fact that these biases can be systematically exploited. This is hardly a surprise to anyone who harbors a sympathetic nerve for Nassim Taleb’s contention that most pontificators lack skin in the game and are, therefore, hamstrung by partial views of the world.
So, in this article some of these mechanisms are considered and reconciled, at least in practice, with the thrust of the arguments proffered by behavioral social science and their biological bases.
Darwin’s Dangerous Idea
The premise worth examining is simply that of natural selection. As Dawkins put in his seminal, The Selfish Gene, humans are mere survival machines for the genes they carry, which are, in a sense, immortal. The point he was trying to make was that of correctly identifying the unit at which natural selection operates. It occurs not at the level of the society or group, nor at the level of the individual — though those levels all provide very useful mathematical heuristics for the underlying dynamic — but at the level of the gene.
This was a brilliant insight for two reasons.
One, it teaches us how easy it is to be fooled into applying evolutionary logic at levels of analysis that are simply easier to see, but fundamentally incorrect. Two, the rationale for genes was on the basis of information. Information is hard — impossible, perhaps — to destroy. It can morph, of course, but resists destruction. This is why Dawkins coined the word ‘meme’ to suggest the enduring relevance of information to societies. (A sad, but inevitable commentary on the idea of information is, indeed, the fact that the word meme itself has been hijacked!)
The reason to bring this up here is to make vivid a simple idea: the simplest way to understand the complexities of investor behavior in markets is to adopt a Darwinian perspective. Behavioral and psychological biases, as well as their biological bases are subsumed within a more fundamental unit of information. In a financial market, this is quite simply price.
A herder causes herd behavior
When we see herd behavior being suggested as the phenomenon that is at the seat of market movements, it is surprising that we look past the obvious question: What is the agent that causes the herding?
For the most part this is overlooked in theory. Some exogenous event or external agent is vaguely alluded to, and its significance usually deemed to be of second-order importance. Who cares about the pebble that started off at the brow of the hill, after all, when what matters is the massive snowball at the bottom!
However, even simple examples from nature suggest that this may not be the most insightful approach if we wish to develop a real understanding for rampaging herds.
Let’s consider two.
A pride of lions hunt a wildebeest on the savannah in a very specific and rehearsed manner that is hardwired into what makes a lion a lion. Some members systematically ‘herd’ the group of wildebeest by causing a panic among them; the latter then collectively and frantically head off in the general direction of the stronger members of the pride, who have been laying low and well concealed in the tall brown grasses.
The following video illustrates the strategy that a pod of humpback whales takes in its approach to hunting a school of fish. There are essential similarities, and some additional nuances beyond the example of the lions that are worth observing. The whales emit sounds from the depths of the sea that frighten the fish, who dutifully panic and head in the opposite direction. The whales then encircle them with a cordon of bubbles, and, finally, and quite expertly, they coordinate their kill.
Are the lions and the whales relying on psychological biases that afflict their prey? Almost certainly they are. They are relying on the skittishness of their prey.
Are they also aware that their prey have biological disadvantages? Yet again, it stands to reason, that they are aware of these as well. If they weren’t, they’d be hunting elephants and sharks instead.
Market Makers as the Herders
We can apply much of the same logic to financial markets.
The Bitcoin (henceforth, BTC) market is especially illustrative as an example, because, in contrast with legacy markets, it is in its infancy. Moreover, the interest of the institutional and professional trader in BTC has been relatively recent. We shall collectively call this latter group of institutional traders the Market Makers, or simply MMs for ease of exposition.
The first of the observations to consider is that the application of Darwinian natural selection is precisely what drives the market, and the unit of analysis is, quite simply, that of price.
The second is the one that is harder to swallow: MMs are better at exploiting the informational content within the price data than is the average individual or retail investor. And by many orders of magnitude! This provides them with a survival advantage, just as Darwin might have predicted, had he been a trader himself.
MMs are the humpback whales, and sadly, retail investors are the hapless school of fish that obligingly coalesces into a bait ball.
Let us consider, with the help of a professional trader, CypherSpook, a few of the mechanisms that they use in practice.
Mechanism 1: Discipline and Coordination
The coordinated attacks that are the hallmark of our lions and whales is common among the MMs, as well. It bears noting that, as CypherSpook puts it:
Institutional traders are taught from their first day that they buy at good levels or they don’t buy at all.
Such discipline is usually in short supply among retail investors. Compounding this disadvantage is the fact that MMs are also either explicitly or tacitly coordinated. Lions don’t have to willingly collude. The prospect of a kill is sufficient to provide the impetus for coordination.
A rather vivid example of this are the ‘kill zones’ in the forex market, whereby an office in Asia operates not merely when the office in North America is resting, but, crucially, when most retail investors in North America aren’t at their most vigilant.
You have to think of each major financial timezone as somebody’s “turn” at working the market. Remember, MMs buy and sell good “levels”. We had a rally in the London timezone… but now New York is awake! They say to themselves “not so fast guys, you aren’t leaving the station without us.” So what do they do? Sell into the rally and push price down to accumulate at lower levels, while keeping the bullish trend intact.
You’ll see this game happen across all time zones.
Sometimes it skips a zone and the rally continues if there is good news and bulls are strong. But if this happens too much we get overextended. Then it’s the MMs opportunity to find the bedrock of the trend.
Mechanism 2: The Stop-Loss Cascades
An integral component of the MM trading approach is that of the pride of lions or pod of whales that learns how intensely a group of its prey needs to be triggered before it behaves like a herd.
MMs use ‘liquidity pools’ as their parameters in defining a hunting zone. These pools are zones where MMs buy BTC to go long precisely when retail investors begin stepping out of the market, and where they sell BTC — that is to say, go short — when the average retail investor is in the mood to buy BTC.
Liquidity pools should not be confused with dark pools, which are yet another mechanism by which MMs can access trades ‘off the regular order books’, using their access to much larger stores of cash.
CypherSpook explains it thus:
Unfortunately most retail traders will wait for some confirmation before they get in a trade. This could be a trend line break, moving average cross, candle pattern. When retail traders enter the market at the level of this confirmation — let’s say they all buy when the price of BTC is $10,000 — they will all set their stop loss price at a predictable level, perhaps in the vicinity of $9,000.
Now here’s where it gets ugly. MMs see this activity and manipulate it.
As an example, let’s say that A is an MM, and the BTC price is currently at $8,000. A operates by buying heavily to artificially drive the price up to $10,000 (which, recall was the entry level for the retail traders), and provide the impetus for a breakout to $15,000.
Now, what A has just done is set a huge trap between $8,000 and $15,000. An ideal hunting ground! A started that rally and trapped all the people that bought in that range.
Since A has more buying and selling power than that group of trapped retail traders, it can start aggressively selling to shake them out of the trades and, quite literally, take their money. All A needed was a few people to lead the way to the cordon it has erected, before they all start crumbling. A can do this because it knows where the retail investors have put their stop losses and it simply has more firepower at its disposal than them.
Mechanism 3: Disinformation
A trigger for herd behavior according to standard economics theory is an information cascade, where correlated behavior is triggered on the basis of a cascade of information that is, in itself, correlated. Such an information pattern obviously thrives wherever there is the possibility for linking across individuals, either serially in a queue, or simultaneously in a real or virtual social network.
But information cascades are not a theory of the propagation of ‘good information’. It is a theory of how available information can be ignored in favor of information that is merely perceived to drive herd behavior.
MMs know this intuitively. In a nascent market like that of Bitcoin, where, by its very nature a central authority does not exist to vet the information in the public domain, disinformation is easy to create and relatively easy to leverage it to inspire herd behavior, as well. Nowhere is this more evident than on Twitter, where ‘good’ information on the technology competes alongside disinformation created by some traders who claim to be MMs, and who seek to generate a cascade that favors their trading strategy.
Mechanism 4: Mechanical Lions
If these and other mechanisms that MMs use in their strategies appear somewhat too mechanical and scripted, then it should not come as any surprise that they can, indeed, be mimicked.
As a matter of fact, Bitcoin is especially afflicted by such mimicry, since it is a relatively new market and is regulated very unevenly across the globe. This makes it a fertile ground for applications of artificial intelligence in the form of high-frequency trading bots.
These bots can be run from anywhere in the world, and they can be taught to learn from the investment behaviors of both retail investors and institutional investors.
There are several examples of how these bots might set about exploiting herd behavior in the market. A common tactic is that of ‘spoofing the market’, where the order book is rigged with several trades to inspire buy or sell action from retail investors, but the order is then retracted as the price begins to move in the desired direction.
Survival of the Smartest
When futures were introduced to Bitcoin late last year, understandably the average individual investor took this to mark a sea change in the game. After all, institutional interest had finally arrived!
The key lesson to extract from the foregoing is not that the doors were inadvertently opened to the predators, who shall now inevitably pick off retail investors — one by one — and sully the vision of broad ownership of Bitcoin across the globe.
Rather, the lesson is simply that being part of a herd of investors often requires acquiring a new mental fitness that separates you from it.
We can rely once again on our motivation from nature. Animals have been endowed with a fight-or-flight response, which is based on the physiology of how they react to a stressful situation such as an attack by a predator. It involves a neural response that brings about the onset of a nasty set of side effects on the body, and which all inspire rapidly fleeing or resorting to fighting.
This instinctual, unthinking binary choice fits in perfectly with the herder’s design: A set of animals facing the same stimulus will all, more or less, be similarly programmed in their responses. If all retail investors panic sell, driving the price down, MMs find themselves in the happy position of being able to buy the asset at discounted price.
However, there is a simple source for seeking the mental fitness necessary to fight a knee-jerk inclination to join the stampede. Quite simply, it is to note that a heightened degree of awareness of the source of stress provides a longer timeframe for planning a response that is unyielding to the herder.
This article hopefully makes us a little more aware of the games MMs play. However, educating ourselves on the true value of our investments is the larger reserve from which long-term awareness obtains.