Robert Parham
banner
kn.owled.ge
Robert Parham
@kn.owled.ge
Statistician, financial economist, computer scientist, astronomer, burner, human. Facts are more important than feelings.

Web: https://kn.owled.ge
PS FAQ: “Wasn’t insider trading illegal long before 2001?” Yes. Trading on inside info has been illegal for decades. What changed is that selectively giving material nonpublic info to outsiders (who might then trade) became a problem for *the firm* under fair-disclosure rules.
August 24, 2025 at 10:19 PM
TL;DR: Fair disclosure made news days matter again. Prices now jump around when the news is public, not when it’s whispered.
August 24, 2025 at 10:19 PM
3. Show that fair-disclosure rules re-centered information release on public announcements, nudging markets, by design, to *reduce* efficiency from “strong-form” toward "semi-strong-form" efficiency.
August 24, 2025 at 10:19 PM
2. Unify familiar empirical proxies (R², idiosyncratic vol, mean-absolute deviations, turnover) under one coherent, model-based measurement of informed trading, yielding a cleaner, better-scaled metric across time and markets.
August 24, 2025 at 10:19 PM
Along the way, we:

1. Generalize Kyle (1985) to a noisy private signal so the binary distinction of “informed vs. uninformed” becomes a continuum of informedness.
August 24, 2025 at 10:19 PM
One implication: a lot of finance/accounting work that leans on “news days” (especially earnings announcements) pre-2001 deserves re-evaluation, as “news days” weren’t truly news.
August 24, 2025 at 10:19 PM
Before that, “news” often wasn’t news: material info leaked to favored listeners, got traded on privately, compounded into prices, and by announcement time… there was little left to move.
August 24, 2025 at 10:19 PM
What changed? Regulation. The U.S. adopted Regulation Fair Disclosure (Reg FD) around 2001, and Europe followed with the Market Abuse Directive (MAD) around 2005. These rules prohibit selective briefings of outsiders.
August 24, 2025 at 10:19 PM
Translation: after those dates, news days started to matter a lot more for firm-specific prices, and markets started behaving the way Roll expected.
August 24, 2025 at 10:19 PM
We track the gap between news-day and non-news-day firm-specific volatility across four markets (U.S., Europe, Japan, Australia), 1980–2018. The gap jumps ~5× in the U.S. around 2001 and in Europe around 2005, but is flat in Japan and Australia.
August 24, 2025 at 10:19 PM
Enter our paper.
August 24, 2025 at 10:19 PM
Back in 1988, Richard Roll pointed to a puzzle: stock returns on news days didn’t look much different from non-news days. If public announcements move prices and non-news is mostly noise, why weren’t announcement days much more volatile? The “R² puzzle” was born and lingered for decades.
August 24, 2025 at 10:19 PM
Given that AI can do it so well, I’m guessing human researchers could have been doing it just as well for some time, as in Novy-Marx 2014 🙃
December 17, 2024 at 3:32 AM
Still trying to figure it out: can you explain the broad-based practical application of Gold and why Fort Knox exists?
November 28, 2024 at 4:37 AM
שלום,
network effects
November 24, 2024 at 9:22 PM
If you resort to claiming lack of knowledge on my side without knowing pretty much anything about me, I'd say there's a bit of smugness problem going on...
November 23, 2024 at 4:38 PM
There's a difference in your mind ("how things should be"); not in reality ("how things are").
November 23, 2024 at 4:21 PM
Alternatively, those other oligarchs were putting it in the service of the previous ruling party, as the media-politics nexus worked for millennia.
November 23, 2024 at 4:20 PM
Who are they buying them from, Filipe?
November 23, 2024 at 4:12 PM
את יכולה לייצר פיד משלך, שמחקה את הפיד הראשי אבל מכבה ריטוויטים של יוזר ספציפי.

SkyFeed is your friend.
November 19, 2024 at 1:20 PM