StockTwits and the Second Derivative of Market Data

Two big types of feedback we get at StockTwits :

1) There is a lot of noise on StockTwits

2) There needs to be a way to track performance and/or rate traders.

While we take all feedback very seriously both are valid and we are constantly building tools to filter and use the data and discover new traders, both miss the point of what StockTwits truly is – the second derivative of market data.

StockTwits is market data with a semantic layer on top.  It is a centralized stream of information and ideas consisting of decentralized analysis.  Using this context let’s address the points above:

Regular Market Data is Noisy

StockTwits is data and information, so of course there is noise.  Market data is mostly noise so why wouldn’t the second derivative of market data consist of noise as well.  Standard market data is an older and more established form of noise that has been around for a while and already has the tools to filter and analyze it so people don’t see it that way. Most of the daily ticks, trades during the day are immensely noisy, yet people pay top dollar for this data and extract alpha from it.  Our job is to build tools and enable others to build tools to be able to transform  and use this new second derivative of market data.

Noise is Relevant

Real time data is very contextual and real time which adds to the noise factor for someone who isn’t there.  Just like “you had to be there to get the joke” you have to be on StockTwits to understand the information coming through.  The more time you spend the more signal there is.

StockTwits is Work

Anyone who has spent any time investing or trading know that its work and a full time job.  Whatever your strategy is there is a lot of time conducting and reading research, testing models and looking for ideas.  I fear that many people think that StockTwits is the “easy”.  It’s not.  Nothing in the market it easy.  StockTwits is the second derivative of market data.  It’s market data refined to a usable form.  Just like there is a lot of work needed to take refined oil and make your car run successfully, there is work that needs to be done to take this refined data and ideas and put them to work in your portfolio.

As a student of the stream, I know that certain traders are better in certain markets.  I know that @wwwstockrake awesome in markets that are volatile and turning (like this one).   I also know that swing traders like @zortrades and @traderstewie are better in trending markets.  I know to discount information from them depending on the type of market you are in.  This is something that I have picked up after spending time watching StockTwits.

Just as a portfolio puts in work to calculate the information ratio on the stocks in his portfolio, someone using StockTwits must do the same.  Using StockTwits information may be a bit easier and different, but it is still work.

There is Nothing More Transparent than a Someone’s Stream

This week a very respected and uber-transparent trader in our community, @johnwelshphd, proposed a ratings system for traders.  He’s not the first to have this suggestion and we’ve had many people request performance for traders, and while perhaps in the future we may implement something along the lines, there is nothing more valuable and transparent than a user’s stream.

There is nothing more transparent than someone’s StockTwits stream.  If someone only tweets exits it takes only a minute of reading their stream to see this.  If someone never tweets losing trades it only takes a few minutes to see that.  Any rating system can be gamed.  Sites like Investimonials, while a great idea and somewhat useful are getting gamed hard. Heck, the ratings agencies that are the backbone of the debt markets were gamed hard (and contributed to the fall).  Gamed ratings systems obscure reality and sometimes create false realities.  Often times less is more, and there is nothing more transparent than a user’s stream.

Plus, ratings systems shifts the focus away from the ideas and information and towards the trader themselves.  New and even bad traders can sometimes have great ideas, and vice versa.  The great thing about golf is that even the worst golfer can often times hit a shot like a pro. Focusing on the individuals is what CNBC and the mainstream media do.  Each day they trot on a hedge fund or mutual fund manager who is a better salesman than investor who goes on and talks their book.  In the investment world, celebrity brings liquidity and the guru becomes a self fulfilling prophecy.  Focusing on the idea and not the individual is much more democratic and disruptive.

Tom Brakke writes in Cave and the Flow:

These days, it is not hard to be pulled into the flow of information.  Digital devices bring it to wherever we may be, at all hours of the day or night.  The flow is a raging river that sweeps us along.

The question is whether or when or how often to seek higher ground and retreat to a cave of our own.

As someone whose mission is to create original ideas, I know that I will do my best work when I spend a lot of time in the cave.  But I must also heed the lessons of the river, so I need to gaze down upon it to gauge its features, to approach it close enough to hear the roars of the raging waters, to reach down into it for a handful of water to taste, and sometimes to take a dip in an eddy of information.  And yes, occasionally I need to swim out into the main current, hoping that it doesn’t sweep me into raging rapids of confusion or over a waterfall of wasted time.

StockTwits is a place where various market participants come out of their caves and add ideas to the flow.  In addition, for those willing to work, it can provide great fodder for those who are going back to their cave to perform analysis. It’s our job to foster the building of tools to make this process better.


  1. The stream is transparent, but can still be deceptive. If I’d made a few bad recommendations in the past, I might fill up my stream with index-hugging bets. If you could automatically tally the results, this would be easier.

    One ST feature I mentioned to Soren the other day, when he asked for suggestions, was the ability to append “+” or “-” to a symbol to indicate sentiment. That way, you could literally parse their feed to see if they were GLD+ last month, and SPY+ this time last year. Then get some kind of expected holding period or expected # of bets in a given portfolio, and you can get an automatic estimate of their results.

    This presents some issues (i.e. stop losses, position sizing, and the fact that some people have long-term positions plus positions they trade in and out of), but it’s a start.

  2. zerobeta

    Yes I agree that it can be deceptive, but it is completely objective and pure. I agree sentiment analysis would be very helpful in that case. It’s just a hard thing to implement.

    Bullish and bearish are sometimes too simple and could be a distraction. Some ideas are more complex or just a tweet that is thinking out loud. Sometimes these are most valuable.

    Definitely something to think about though.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s

%d bloggers like this: