Matt Levine on academic research into insider trading correctly observes that published papers always seem to conclude insider trading happens before X, for any X. https://www.bloomberg.com/view/articles/2018-02-15/when-everything-is-too-safe-add-risk … A tantalizing possibility: insider trading happens before accidental p-hacking.
-
Show this thread
-
Starfighter’s last level asked players to identify an inside trader on a fictitious stock exchange driven by a world simulation. We simulated insider trading by giving the insider oracular knowledge of future events in the simulation relevant to their stock.
3 replies 1 retweet 11 likesShow this thread -
And I knew, by construction, that there was one and only one insider per simulation, because I put him there. Do you know how many people tried to convince me they found all seven?
1 reply 2 retweets 19 likesShow this thread -
As long as I’m on the topic let me spill the beans of how our simulation worked, because it’s the coolest thing I ever built. I wanted stocks to trade in a way that “felt real” so I started with historical reality.
1 reply 0 retweets 15 likesShow this thread -
Every time a player played a level they’d be matched up with one year of history from 2000 to 2015 with one S&P 500 stock. I did a random walk based on that year, to prevent anyone from figuring out that this was happening.
2 replies 0 retweets 9 likesShow this thread -
This basically created a surprisingly lifelike companies; they’d show business cycles, sometimes blow up ($BAC in 2008 was a great time to play through), sometimes have a flat year, etc.
1 reply 0 retweets 7 likesShow this thread -
The random walk’s results were the true, oracular, platonic idealized Fair Market Value for the company. For each trader in the simulation, we gave them a distorted view on their *perception* of that value.
1 reply 0 retweets 4 likesShow this thread -
Trivial example: for a buy-and-hold bot, we’d let them estimate FMV by rolling randomly between N% of the true FMV, for some N configurable to demonstrate skill. When they saw the market diverge from their estimate, they’d put in buy orders.
1 reply 0 retweets 4 likesShow this thread -
A very simple ecology of very simulated bots was sufficient to keep up the charade of an actively, lively, not-terribly-disordered market... at least until acted upon by the player, at which point things could go off the rails in delightfully silly ways.
1 reply 0 retweets 4 likesShow this thread -
The insider was trivial to program given that the computer knew the true FMV for the entire simulated year in advance: he’s a value investor who buys when the stock is undervalued and sells when he has locked in a profit... buuuuuut he *literally* knows the future FMV, exactly.
1 reply 0 retweets 5 likesShow this thread
Interestingly I think we might be the only game which actually had a legit order book*, possibly ever. Most games just simulate the closing price, sometimes the bid/ask price, but it was important for our pedagogic purposes that there be an actual book. * aside from crypto ;)
-
-
You can’t show some of the real challenges of trading without there being an actual order book. For example: slippage. Putting in a sufficiently large order moves the market against you. You can’t teach people about that if you give them any quantity at “the price.”
2 replies 0 retweets 5 likesShow this threadThanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
Replying to @patio11
I've thought starfighter more than a few times when watching crypto order books.
0 replies 0 retweets 0 likesThanks. Twitter will use this to make your timeline better. UndoUndo
-
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.