Echo Chambers in Investment Discussion Boards

[Hielko]

Interesting paper, certainly applies to this board as well: http://www.cs.ucsb.edu/~ravenben/publications/pdf/echo-icwsm17.pdf

 

Abstract:

We examine the quality of information and communication in

online investment discussion boards. We show that positivity

bias and skewed risk/reward assessments, exacerbated by the

insular nature of the community and its social structure, contribute

to underperforming investment advice and unnecessary

trading. Discussion post sentiment has negligible correlation

with future stock market returns, but does have a positive

correlation with trading volumes and volatility. Our trading

simulations show that across different timeframes, this

misinformation leads 50-70% of users to underperform the

market average. We then examine social structure in communities,

and show that the majority of market sentiment is

produced by a small number of community leaders, and that

many members actively resist negative sentiment, thus minimizing

viewpoint diversity. To improve generated information

content in online investment communities, we suggest

designing to increase diversity of opinion, minimize friction

around incorporating new information, and provide performance

feedback for self-correction.

[rolling]

Thank you. Never had thought about it  :-[ 

It should help avoiding some risks when thinking about investing in board ideas (either in this board others)

[merkhet]

Interesting paper, certainly applies to this board as well: http://www.cs.ucsb.edu/~ravenben/publications/pdf/echo-icwsm17.pdf

 

Abstract:

We examine the quality of information and communication in

online investment discussion boards. We show that positivity

bias and skewed risk/reward assessments, exacerbated by the

insular nature of the community and its social structure, contribute

to underperforming investment advice and unnecessary

trading. Discussion post sentiment has negligible correlation

with future stock market returns, but does have a positive

correlation with trading volumes and volatility. Our trading

simulations show that across different timeframes, this

misinformation leads 50-70% of users to underperform the

market average. We then examine social structure in communities,

and show that the majority of market sentiment is

produced by a small number of community leaders, and that

many members actively resist negative sentiment, thus minimizing

viewpoint diversity. To improve generated information

content in online investment communities, we suggest

designing to increase diversity of opinion, minimize friction

around incorporating new information, and provide performance

feedback for self-correction.

 

Basically describes the $VRX thread.

[StubbleJumper]

Haven't we always known that Yahoo Boards were crap?  At least there's decent thought by most posters on COBF, and there are precious few cheerleaders blindly pumping names.

[DooDiligence]

Haven't we always known that Yahoo Boards were crap?  At least there's decent thought by most posters on COBF, and there are precious few cheerleaders blindly pumping names.

 

++

[Cigarbutt]

I'm fairly new here but I submit that, overall, this forum is excellent.

There a many outliers that make it uncomfortable but that's the way to go, especially if done in a respectful manner.

But there is always a danger of self-reinforcing group think. Need humility and a strong inner score card at the same time. Not easy.

I like:"To improve generated information content in online investment communities, we suggest designing to increase diversity of opinion, minimize friction

around incorporating new information, and provide performance feedback for self-correction."

Please never hesitate to provide help for self-correction.

Who knows, maybe useful concept outside of investment circles?

[LC]

FYI: Calling the COBF forum excellent and somehow better than "lesser" forums i.e. yahoo, is the exact type of insular/echo chamber structure that the article is warning us about. Food for thought

[Cigarbutt]

I did not realize that feedback would come back so rapidly. Way to go LC. I owe you one.

[SharperDingaan]

Interesting paper, certainly applies to this board as well: http://www.cs.ucsb.edu/~ravenben/publications/pdf/echo-icwsm17.pdf

 

Abstract:

We examine the quality of information and communication in

online investment discussion boards. We show that positivity

bias and skewed risk/reward assessments, exacerbated by the

insular nature of the community and its social structure, contribute

to underperforming investment advice and unnecessary

trading. Discussion post sentiment has negligible correlation

with future stock market returns, but does have a positive

correlation with trading volumes and volatility. Our trading

simulations show that across different timeframes, this

misinformation leads 50-70% of users to underperform the

market average. We then examine social structure in communities,

and show that the majority of market sentiment is

produced by a small number of community leaders, and that

many members actively resist negative sentiment, thus minimizing

viewpoint diversity. To improve generated information

content in online investment communities, we suggest

designing to increase diversity of opinion, minimize friction

around incorporating new information, and provide performance

feedback for self-correction.

 

Basically describes the $VRX thread.

 

Describes the various FFH threads as well.

Lack of diversity of viewpoint, & pressure to conform to the group think.

 

SD

[Jurgis]

I finally read the paper. They have some good results and some very so-so results.

First, positives:

 

Resistance to Viewpoint Diversity

 

This is well researched and real issue IMO.

 

These are so so:

 

Poor User Generated Information: sentiment in the discussion boards does not correlate with future stock price

movements, and in fact stock trading strategies based on the sentiment in these boards underperform the overall

market.

 

This is claimed based on stock prices 5 days ( ! ) from the posting. "Strategies" based on sentiment are really crappy IMO ("buy when someone posts positive sentiment on stock, sell when someone posts negative sentiment"). It would be great if they got some long-er term stock return results. But then it's very difficult to research: what is "long-er term"? When would you buy/sell/etc.?

 

Failing to Incorporate New Information: sentiment in the discussion boards does not correlate with recent past stock price movements.

 

This is also not very useful. Clearly if you are value investor, you would have a positive sentiment on dropping stock. Or not. Depends on the reasons for drop. So correlation is not expected and lack of it does not show much if anything.

 

Mistimed Activity Adds Noise: posting activity correlates positively with future stock trading volume and volatility.

 

This might be true. Not sure it is useful insight. BTW, there might be 3rd cause that is real cause of both posting activity and future trading volume and volatility. In other words, correlation, but not causation. (At least they don't claim causation :) ).

 

[rkbabang]

Poor User Generated Information: sentiment in the discussion boards does not correlate with future stock price

movements, and in fact stock trading strategies based on the sentiment in these boards underperform the overall

market.

 

This is claimed based on stock prices 5 days ( ! ) from the posting. "Strategies" based on sentiment are really crappy IMO ("buy when someone posts positive sentiment on stock, sell when someone posts negative sentiment"). It would be great if they got some long-er term stock return results. But then it's very difficult to research: what is "long-er term"? When would you buy/sell/etc.?

 

 

Breaking News: Shocking new research shows that online discussion board posters can not accurately predict the 5 day movement of stock prices!!!!!

 

[DooDiligence]

Poor User Generated Information: sentiment in the discussion boards does not correlate with future stock price

movements, and in fact stock trading strategies based on the sentiment in these boards underperform the overall

market.

 

This is claimed based on stock prices 5 days ( ! ) from the posting. "Strategies" based on sentiment are really crappy IMO ("buy when someone posts positive sentiment on stock, sell when someone posts negative sentiment"). It would be great if they got some long-er term stock return results. But then it's very difficult to research: what is "long-er term"? When would you buy/sell/etc.?

 

 

Breaking News: Shocking new research shows that online discussion board posters can not accurately predict the 5 day movement of stock prices!!!!!

 

 

How do they do a decade out?

 

I say they as though we are not they...