I want to warn people that sites such as that linked here cannot possibly predict the direction of the real equities market in a reliable statistical sense -- apart from the outcomes one might get by flipping a fair coin. This is common sense -- if the operator of the site actually had an inside track on the market's workings, what possible reason would he have for revealing his findings to the public, instead of pri…
You're right, but it's even worse than that. This system is trivially vulnerable to sybil attacks. I can create 16 identities, each posting an opposing investment thesis on 8 securities. For the 8 identities that were correct in the first round, each posts a new thesis for 4 securities. For the 4 identities that were correct in the second round, each posts a new thesis for 2 securities. Now I have two identities with…
It's based on "Information Ratio Comparison with a long S&P Information Ratio". Information ratio is return/returnVol of your trade. This return/risk is compared to simulatenous S&PReturn/vol (a long position in S&P here ALWAYS). So you would have to outperform S&P portfolio buy on a vol adjusted basis. Which is a huge endeavor, because S&P has a much lower vol than indivdiual stocks due to diversification. I am trying to determine if you have "skill" picking stocks and outperforming S&P on a risk adjusted basis. In simple terms, even if you put buy and sell opposing views doesn't mean you outperformed S&P on a vol adjusted basis, so your accuracy rating would still go down. Which is veyr likely if your sell stock doesn't go down in price at all but just underperforms S&P returns. As stocks on average have a positive drift (i.e. 6-8% return over long periods) I hope this is clear. You would have to have inside information that a particular stock was going to move more than the market times the stock's beta with respect to the market. Please ask if this is not clear. So the coin flipping strategy doesn't apply directly. Please throw at me your next sophistication when you're ready (Hint: figure out probablity that a stock outperforms S&P on a risk-adjusted basis in a given period). We are always comparing it with S&P because we want to see if investors should hold the market portfolio or rely on a stock picker.
2) This issue is not UNIQUE to my site. There are millions of boards (yahoofinance, cnbc, barrons, stocktwits, bogleheads, ft, you name it) which provide no accuracy ratings for any of their commenters, bloggers and ppl can do what you are suggesting their as well. And in fact, this problem exists across the internet. Jim Cramer the biggest media analyst has an awful trackrecord but any new entrant doesn't know that. .So your concern applies to all the financial media forums in general! for the record unlike WSJ, cnbc, ft and other litter out there, my website at least attempts to track risk adjusted accuracy for everyone as per above metric not coin flippa.
3) Also most sophisticated INVESTORS ARE NOT DENSE. They will check your outperformance versus your reasons for outperformance in your writeups. People on the street generally know who got lucky and who called "it" and know her/his stuff.
4) VALIDATION TECHNIQUE DUH!!! Also I am currently testing beta traction I plan to add social media authentication (or cellphone code check) and minium trades before accuracy/crowd validation of articles. Eventually connect with brokerage to verify purchase.. which is not a problem. Also i plan to turn this into a crowdsourced efforts where analysts/writers get paid for their efforts.
The last part is where it gets interesting. I am trying to put a check across rampant misinformation on Wall st which hurts retail investors the most. That being said, I am open to suggestions to change things on my site and build it better.
Thanks