3 ms·
Not OP, but Matt Levine "wrote about me" [0]. I made a Matt Levine insider trading game [1]. It was inspired by a Matt Levine article [2] where he mentioned how
by bko 6y ago
Not OP, but Matt Levine "wrote about me" [0]. I made a Matt Levine insider trading game [1]. It was inspired by a Matt Levine article [2] where he mentioned how some professional traders had access to early earnings but even with that data, they were still highly inconsistent in their trade performance. The point was stock prices are notoriously difficult to predict even with early earnings.
So I made this app where you get early earnings and a graph of prior stock prices and you try to guess one of two paths (they were inverse of each other). Even with all of those myself (and Matt Levine) were correct about 50% of the time.
I emailed him and he actually told me he was worse than 50% , which seems like a good thing because then you could just do opposite of your initial guess and have a profitable strategy. He was very nice in his correspondence, although he didn't give me a heads up saying he was going to blast it out. It's a static site [3] so it wasn't a problem, nor do I really care too much. But it gave me the ultimate flex when another super-fan at work spotted it and spammed it out.
From the article.
> Elsewhere here is “Matt Levine’s Insider Trading Game,” though I actually have nothing to do with it.
[0] https://www.bloomberg.com/opinion/articles/2019-12-16/we-kept-almost-making-money https://www.bloomberg.com/opinion/articles/2019-12-16/we-kep...
[1] https://insider-trading-game.netlify.app/ https://insider-trading-game.netlify.app/
[2] https://www.bloomberg.com/opinion/articles/2019-11-26/knowing-the-future-isn-t-that-helpful https://www.bloomberg.com/opinion/articles/2019-11-26/knowin...
[3] https://github.com/breeko/stox https://github.com/breeko/stox
- harmegido 6y agoThis is a very cool app, but one thing I'm noticing is that you aren't providing any numbers about expectations (expected earnings/share, etc). Stock prices are forward projections, so it's less about how much money made/lost and more about how much money made/lost beyond current expectations. And this is certainly information a real insider trader would have. Perhaps that would improve players results?
- bko 6y agoThats funny because that was the exact comment Matt Levine made in our email correspondence. I mentioned I'd be happy to include it if he got me a Bloomberg api subscription, but no luck! My guess would be that expectations wouldn't help much. The analysts at banks that provide estimates are relatively conservative (don't want to stick out from the crowd too much). For instance, it's much better to be way off from actual as compared to other analyst estimates. Because you can always say everyone else believed the same thing as an excuse. If you're way off from others and you happen to be wrong, then thats more embarrassing. The analysts also rely on internal estimates pretty heavily, and companies aren't dumb, so they usually downplay expectations. So much so that the majority of companies beat their own estimates (over 70% of the time). So I think these numbers are gamed and without being a sophisticated practitioner, they wouldn't help IMO [0] https://insight.factset.com/record-performance-vs.-eps-estimates-by-sp-500-companies-in-q2 https://insight.factset.com/record-performance-vs.-eps-estim...
- yayitswei 6y agoCool game! I was able to get over 65% by looking at just the year, that should be hidden in the guessing phase.