Prediction markets have a fairness problem

When Elbert Gary, chairman of the newly formed US Steel, proposed releasing the company’s financial results to directors only when they were released to the public, several board members were reportedly “outraged.”

In the plutocratic days of 1902, US Steel’s non-executive directors considered the money they could make trading on non-public information a perk of the job — “some of them had added handsome slices to their fortunes by this kind of maneuvering,” Gary’s biographer writes.

They were far from the only ones.

In 1904, the famed financial writer Edwin Lefèvre observed that insider trading by company directors was “taken as a matter of course, without indignation, without even passing comment.”

Lefèvre, however, did pass comment: “It is unsportsmanlike. It is fighting a smaller man for money.”

It was not, however, illegal.

That same year, after shareholders discovered that tens of thousands of dollars were missing from the Lillooet Gold Dredging Company, a “pretty stenographer” came under suspicion because she’d been seen spending money well in excess of her $60-a-month salary.

In her defense, the stenographer explained that she hadn’t stolen the money — but she did know it was gone.

“She declares that she knew of the shortage a long time ago, and that she is innocent of any wrongdoing,” The Minneapolis Journal reported. “She admits that she has made a great deal of money, but that she had made it legitimately by scalping the stock of the company on account of the inside tips which she has been able to secure by reading the mail that passed thr[ough] her hands.”

“No one seemed to question the legitimacy of her actions,” Michael A. Perino writes in The Lost History of Insider Trading — and she was correct in thinking she hadn’t broken any laws.

Prior to 1900, as Perino explains, courts had “almost uniformly” held that directors and officers owed no duty to shareholders when they dealt in a company’s shares, even if they were in possession of material nonpublic information.

As late as 1933, the Massachusetts Supreme Judicial Court ruled the same.

In 1934, however, the Securities Exchange Act made it illegal “to use or employ, in connection with the purchase or sale of any security...any manipulative or deceptive device or contrivance.”

That broad wording would seem to cover trading on privileged, nonpublic information. But it wasn’t until 1961 that the SEC prosecuted its first public-markets case of insider trading.

(Interestingly, it was not a board-sitting plutocrat who was made an example of. Instead, it was a retail broker who, in the SEC’s own words, “probably assumed, without thinking about it, that the dividend action was already a matter of public information.”)

The long gap between the Securities and Exchange Act in 1934 and the precedent-setting case in 1961 created the impression that insider trading was widely considered acceptable behavior before the SEC began prosecuting it.

In fact, it was widely condemned right from the start.

In 1915, The New York Times opined that, while insider trading was legal, “Christian gentlemen with a very fine sense of honor” would refrain from using non-public information to enrich themselves.

In 1906, when directors of the Union Pacific Railroad were accused of delaying the announcement of a dividend increase to give themselves time to buy the company’s shares, it was a front-page scandal. The directors faced no legal jeopardy, but the uproar against them suggests people agreed with Elbert Gary's assessment that their privileged dealing “was akin to robbery of their own stockholders.”

Perino further documents that critics argued against insider trading in modern-sounding terms as early as the 1880s, warning that it discouraged participation in the stock market and created misaligned incentives for business executives.

It took decades of debate before the SEC finally did something about it, but a consensus emerged that insider trading undermines the stock market.

Now that we have prediction markets, we get to debate it all over again.

The unlevel playing field

This weekend, the crypto-analytics firm Bubblemaps identified six Polymarket accounts that made $1.2 million betting that the US would strike Iran. 

In each case, the bets were placed so close to the event that it’s presumed they had insider knowledge of it.

In theory, that’s exactly the kind of information that prediction markets are meant to surface. It’s useful to know how likely it is the US will attack Iran, and, after it happens, how likely it is the attacks will topple the Iranian regime (only 48% likely, apparently).

“The promise of prediction markets is to harness the wisdom of the crowd to create accurate, unbiased forecasts for the most important events to society,” Polymarket says about markets related to events in the Middle East.

Polymarket also says they’d prefer that the crowd not include insiders. But that’s impossible to regulate on a peer-to-peer platform that accepts crypto and does not require KYC.

Polymarket’s semi-decentralization has also allowed anonymous insiders to trade ahead of military actions like the extradition of Maduro from Venezuela (placing their bet just five hours before the event), and last year’s attack on Iran.

These kinds of markets are popular with bettors: The Iran bets that resolved this weekend did at least $679 million of volume on Polymarket alone. 

But they’re unpopular with everyone else. Maybe existentially so: Senator Chris Murphy says he’s drafting legislation that would ban “corrupt and destabilizing prediction markets, where insiders who know the outcome…can rig the game to favor certain bets.”

I can’t tell you what the odds of that legislation passing are (there’s no Polymarket on it, sadly).

But beyond the political backlash, prediction markets face rising reputational risk — because who wants to bet against people with inside information?

This, Perino explains, is the primary reason why insider trading is banned in equity markets: “The government invariably explains that the rules against insider trading are about leveling the playing field and encouraging investor confidence in the fairness and integrity of the marketplace.”

The rules appear to be effective. 

“Comparative studies have shown a correlation between robust insider trading enforcement and firm valuations or widespread investor participation in equity capital markets,” he adds.

Prediction markets don’t have a firm-valuation problem, because that’s not where firms are valued (yet). 

But they do have a fairness problem.

If semi-decentralized platforms like Polymarket want widespread participation in prediction markets, they’ll have to take a lesson from US Steel: Stop the insiders from trading.

If not, history suggests the government will do it for them.

— Byron Gilliam