There has been much talk lately in our personal circles about positive externalities in technology, also known as network effects. In particular, we have been thinking about this phenomenon because it is one of the fundamental characteristics of successful open source technology.
The effect is simple to intuit. Positive network effects occur in products which get more valuable as they attract more users.
Consider a modern example: a social network. The more people who use the network, the more value an individual user gains from being part of the network.
Moving closer to home, we find that trading is replete with its own network effects.
The classic example would be the trading pits. Nowadays they are largely an anachronism, but in the past the pit was the most valuable place to be in the trading world.
Why was the pit valuable? Traders in the pit had access to the best information because everyone used it. Thus, the more people who traded on the floor, the more valuable the pit became for each individual trader.
Then electronic trading emerged.
In the beginning, electronic trading was confined to a mixture of upstairs traders and enterprising locals who began to arb the electronic market to the more liquid open outcry market.
As more people began making the transition upstairs, electronic trading overtook the pit for the same reasons which made the pit valuable in the first place.
Trading strategies can also exhibit network effects. Suppose that you observe that two stocks, ABC and XYZ, are highly correlated with one another. You then design a statistical arbitrage strategy for pairs trading the two stocks.
Now consider two different scenarios:
- In the first case, you are the only person who identifies that ABC and XYZ are a tradable pair
- In the second case, you are part of a community of traders who has identified the relationship between ABC and XYZ
Now suppose that ABC and XYZ begin to drift apart: ABC is up and XYZ is down.
In the first scenario, you are the only trader who is fading the move in the spread between the two stocks. In the second scenario, any temporary shocks away from the long run equilibrium relationship are faded by a large group of traders whose actions are coordinated by the use of a shared trading model.
The first scenario is inherently unstable: you will bail out of the trade when your risk limits are triggered and that is that. No one else is going to step in front of that movement.
The second scenario can create a stable state, up to a point. When many people are trading in unison, the market is not reliant on the fortunes of one trader to keep the correlation in line with expectations.
However, network effects can run in the opposite direction, in which case the process is known as congestion.
For example, if too many traders are using the same strategy the returns can break down purely due to market microstructure. Additionally, if too many people know about an anomaly, it will disappear due to arbitrage.
Interested in resources for Pairs Trading?
Want to learn how to mine social data sources like Google Trends, StockTwits, Twitter, and Estimize? Make sure to download our book Intro to Social Data for Traders