My exploration of complex adaptive systems started with Per Bak’s infamous sandpile experiment, which captures all the essential characteristics of a complex adaptive system.
With the sandpile example, we saw that you couldn’t only consider the individual grains of sand in understanding its state. Eventually, when you add enough grains of sand to the pile and those grains interact with each other, the pile itself takes on a new set of characteristics that are difficult to predict.
The consequences of how the grains interact with each other go beyond what you would expect to happen if you analyzed each individual grain within the pile.
To better understand how patterns and behaviors emerge, we can examine some examples of complex adaptive systems in nature.
Getting to Know Emergent Behavior Through Nature
An ant colony is a common example used to showcase emergent behavior in a complex adaptive system.
There’s no hierarchy of command in the colony. There is no elected group of bureaucrats to pass down orders. Different ants have specific tasks. Younger ants work inside the nest to care for the queen and her brood. Older ants work outside the nest to gather food and defend against enemies.
Just for fun, let’s imagine an interview with Anthony the Ant, a forager from Park Ant Colony.
—
Me: Anthony, thanks for being here.
Anthony the Ant: No problem, Peter. I’m excited to be here. I never get this type of attention.
Me: Anthony, would you mind sharing with me what your typical day looks like?
Anthony the Ant: Absolutely! I’m a forager that is responsible for finding food for everyone back at Park Ant Colony. Every morning I leave the nest and walk around randomly until I find food. Once I find food, I take it back to the nest and leave a trail of pheromones in case any of my colleagues are in the area looking for food.
Me: That’s great, then what?
Anthony the Ant: Well…that’s pretty much it. Once I drop the food off, I head back out of the nest and wander around until I find a food source. Sometimes it takes me longer than other times, but I just keep wandering around until I find food. Once I do, I return to the nest with it and leave behind a trail of pheromones.
—
What did we just learn about an individual ant and the rules governing his day?
We learned about what the individual ant does. But we don’t know anything about the colony as a whole or how (and why) it works as a large, cohesive system made up of countless individuals that can all interact with other individuals.
By studying this individual agent in the system, we miss out on the behaviors that emerge from interactions with other agents.
In this case, other forager ants walk around randomly until they find food. If these ants encounter the pheromone trail left by our friend Anthony the Ant, then they follow it toward a food source and leave their own pheromones behind as they carry food back to the nest. This creates a positive feedback loop that enhances the success of the trail and leads to the emergence of a highway of ants.
On the other hand, Anthony’s trail will disappear if other ants don’t follow it, bring back food, and add their own pheromones. This may occur if Anthony stumbled upon a food source that isn’t as close to the pile as a food source found by another forager ant.
Because of the interactions between the individuals and the simple rules governing the lives of the forager ants, food sources nearest the nest will be entirely consumed first.
The queen in the colony isn’t telling ants to find the nearest food source, nor is there a Head of Foraging within the colony that directs ants to the nearest pile of food. Rather than being planned or controlled, this behavior emerges from the local interactions of the agents in the system.
There Is No Leader in a Complex Adaptive System
Prior to learning about complex adaptive systems, I assumed that a bird at the front of a flock was the leader. This isn’t the case, though. By following a few simple rules, the birds naturally organize into a pattern with no specific leader or hierarchy.
As birds take flight, they simply try to avoid crowding. If they are too close to their neighbors, they turn their direction to increase the separation. If they are too far from the flock, they change direction until they are turned in line with their neighbors and move a bit closer to the group.
The birds interact with each other using these simple rules and emerge as an organized group.
In both the ant colony and flock of birds, there is no planning involved nor is there a centralized group calling the shots to make the system go. Instead, the system emerges unbeknownst to the individual agents that are simply following their own individual rules.
Emergence in Financial Markets
Much like the anthill, financial markets are comprised of millions of participants. Each participant is, theoretically, bringing diverse tastes and trading rules into the system.
These rules adapt over time based on feedback. What emerges from interactions of investors is what Adam Smith called the “invisible hand.”
Adam Smith’s invisible hand suggests that when people are allowed to trade freely, then self-interested traders in the market would compete with each other for profit opportunities, which in turn would drive supply and demand towards an equilibrium price.
In other words, the price of a given security is an accurate reflection of the present value of its future cash flows. The collective wisdom of financial markets can lead to highly efficient markets in which current prices reflect all available information and any inefficiencies in market prices cannot be systematically exploited.
Unfortunately, the rational investor economic textbooks rely on doesn’t exist. Market efficiency is strongest when diversity among individual agents is strongest, but there are times diversity among market participants breaks down for behavioral reasons or economic reasons. We see booms and busts when this happens.
Complex adaptive systems aren’t necessarily complicated. The emerging patterns may have a rich variety, but like a kaleidoscope the rules governing the function of the system are quite simple. Small changes in the initial conditions of the system, however, can have significant effects after they have passed through emergence.
Forecasting and Predictions Rarely Work in Financial Markets
The market is greater than the sum of its parts. You can take all the fundamental research in the world and have your finger on the pulse of the market, but the properties that emerge will always be difficult to predict.
This is a big part of what makes forecasting in financial markets so difficult. People place too much importance on explaining individual pieces of the market and not enough on how people perceive those pieces will interact with each other.
Even if you are aware of this dynamic, very few people have the capability to master the ever-changing mix of calculus and psychology.
Related articles:
Thinking About Markets Like Piles of Sand
