How do relevant electronic social networks affect individual investing? In their March 2012 paper entitled “Facebook Finance: How Social Interaction Propagates Active Investing”, Rawley Heimer and David Simon investigate the propagation of active investing strategies within a Facebook-like social network of retail foreign exchange traders. Registered users of this free network (who must have a qualified foreign exchange broker account) have access to: (1) an indicator of the aggregate positions of the entire network in specific currency pairs; and, (2) a real-time view of the trading activity of mutually accepted “friends.” The network receives information about user trades instantly from qualified brokers. Using a complete record of activities within this network involving more than 5,500 foreign exchange traders, two million time-stamped trades and 140,000 messages and friendships mostly between February 2009 and December 2010, they find that:
- The median user is about 36 years old with one to three years of trading experience, self-identifies as a technical trader and lives in the U.S. or Western Europe. The average (median) number of friends per user is 20.9 (8.0).
- Regarding user trading activity and performance:
- About half of trades last less than an hour and only 10% last more than a day. The euro-dollar pair accounts for 34.3% of trades. Average trade size is $34,580 with 8.6X leverage (excluding outliers).
- Average (median) return per trade is -$6.20 (+$0.22), with 63.4% of trades profitable. Over the entire sample period, only 21% of all traders are profitable, with the average trader losing $2,335.
- More active traders have a higher win rate and lose slightly less per trade than the average, but they experience higher weekly volatility and lose more cumulatively (over the sample period, only 17.8% are profitable with average loss $4,776).
- About 75% of all users quit trading during the sample period.
- Regarding relationships between short-term returns and social interactions (read the example below):
- The tendency of users to contact others increases with the sender’s short-term returns. There is no evidence that the messages cause the sender’s higher returns.
- Trading by message recipients increases with the returns of those sending the message. In other words, based on the results above, social interaction tends to stimulate recipients toward greater volatility and cumulative loss.
- Results are robust to controls for individual characteristics and market conditions.
- Despite the high dropout rate, average user trading frequency and return volatility increase over time.
The authors illustrate with…
“…the example of trader 478, a 34-year-old Indonesian who made 792 trades over the course of 10 weeks. One week in November 2009, trader 478 made a total profit of $72,303, becoming among the more successful traders of that month. Trader 478 sent messages to 128 other traders that week, more than twice the number contacted during the previous week, supposedly to celebrate his performance. Upon receiving these messages, the 128 recipients increased their aggregate trading activity the following week by almost 30 percent, from 1,707 trades to 2,148 trades. However, their collective gains dropped throughout the week by 176 percent, turning the group of recipients substantially unprofitable. Within a few weeks, most of the recipients had quit trading.”
In summary, evidence indicates that individuals tend to ape the short-term trading success of others and that social networks thereby amplify trading activity (and participant return volatility).
This trade-stimulating dynamic may apply also among investment/fund managers and between investment/fund managers and individuals.
Cautions regarding findings include:
- Individuals attracted to social networks may not be representative of all investors/traders.
- The social network dynamics may change if/when the supply of new users approaches exhaustion.
