World Views | Senators’ stock trades would make a poker cheat blush

One dispiriting story of the Covid-19 pandemic is that four U.S. senators sold hundreds of thousands of dollars of stock in the days after getting a classified briefing on the expanding coronavirus in January. The trades paid off as the virus spread and stocks crashed. This development is more interesting in a positive way than suggested by the news coverage, in that it shows efforts to clamp down on potential insider trading by members of Congress have largely worked. Even so, more needs to be done.
I’m a poker player. Cheating in poker is harder than most people think. It’s relatively easy to deal yourself or a partner good cards. You don’t need much practice for that, but it has minimal value. You can accomplish the same thing by merely throwing away your bad hands and only playing the good ones. To make money against good players, you need to deal other people very good cards, and yourself better cards, in statistically unusual ways. Since good players are relying on statistics, by subverting the odds, you can make money. But this becomes obvious pretty quickly. No one hand proves anything, but the build up of unusual events needed to make significant profits stands out clearly.
This is what members of Congress have been doing at least since 1978, when the Ethics in Government Act required disclosure of stock transactions. Researchers who have studied the transactions that were subsequently disclosed found clear statistical evidence of misuse of confidential information.
Any single trade could be luck or even skillful use of public information, and any one member of Congress could be lucky or skillful. But the overall record was inconsistent with honesty. Profits were considerably better than Warren Buffett managed. It defies credulity that the average member of Congress is that much better at stock picking than Warren Buffett.
I was skeptical when Congress passed the Stock Act of 2012, which bars insider trading by lawmakers, because I don’t trust reforms passed by the people who need reforming. The act did not say much beyond the old ethics rules, and didn’t have sharp teeth. On top of that, even most of the dull teeth were quietly pulled a year later without a recorded vote.
A new study published on April 22 by the National Bureau of Economic Research titled “Relief Rally: Senators As Feckless As the Rest of Us at Stock Picking” shows the Stock Act worked. Congressional stock trading dropped 77%, and abnormal profits disappeared. Incidentally, this is one more nail in the coffin for the claim that pre-2012 trading was innocent. After all, why would stronger reporting requirements discourage legitimate trading and destroy trading skill?
But bad behavior doesn’t disappear when outlawed, it just changes. Getting back to poker, a different suspicious circumstance is when a new player sits at the table and gets a royal flush on the first hand. This is different from the regular player who might be cheating. Here there is specific evidence from one hand, not accumulation over many hands. The Relief Rally paper documents that the stock trading after the coronavirus briefing was entirely unlike any other in the eight years since the Stock Act became law.
Normally, Congressional buying and selling balance out, and almost all trades are under $15,000. After the briefing, there were the normal number of small buys and sells, and these trades did not do abnormally well. But there was a flood of 60 large sells, without matching buys, that did very well—not just because the market as a whole went down, but because the specific stocks sold did worse than the market. The paper further documents that the happy results cannot be plausibly explained by known market factors unrelated to the virus.
This result meets standard tests of statistical significance in that it’s unlikely to be random chance. But that’s for all 60 trades combined. Any one senator could have been lucky or skillful. But when four people sit at the poker table and all get exceptional cards on the only hand they are dealt in eight years, it’s wise to find another game. Aaron Brown, Bloomberg