Franklin Hugh Money

In 1994, two finance professors did the least glamorous thing anyone in finance can do. They counted. They counted the last digit of stock quotes, not the prices themselves. Just the last digit.

Half the Prices Were Missing

In those days, American stocks did not move in pennies. They moved in eighths of a dollar, and an eighth is twelve and a half cents. So a quote could end on any one of eight fractions, from a single eighth all the way up to a whole dollar.

The two professors, William Christie and Paul Schultz, looked at the biggest, most heavily traded stocks in the country, and they found something strange. Four of those eight fractions almost never showed up. Not rarely. Almost never.

If prices moved freely, you would expect all eight fractions to turn up about equally often. You do not flip a fair coin and get heads forty times in a row. Yet on the deepest, most competitive stocks in America, half of the possible prices had quietly gone missing.

There was an obvious explanation, and everyone in the industry offered it. They called it convention, or habit, the way it had always been done. Traders quoted in even eighths the way you say "quarter past" instead of "fifteen minutes past." No conspiracy, just custom. That explanation was not crazy. It was only incomplete, in the one way that turns out to matter.

The Mechanism

When you buy a stock, someone sells it to you at one price. When you sell, someone buys it from you at a slightly lower one. The gap between those two prices is called the spread. Hold onto that word. The spread is not a fee on your statement, and you never see it as a line item. It is baked quietly into the two prices, and it is how the dealer in the middle, the market maker, gets paid.

If a quote can land on any eighth, the smallest possible spread is one eighth, twelve and a half cents. But if the odd eighths never appear, the smallest possible spread doubles. It becomes a quarter. Twenty-five cents, on every round trip, on millions of them.

And that is the whole idea, the one worth carrying out of here. The smallest increment you are allowed to quote a price in decides who keeps the difference.

The Day It Stopped

Now, a convention is not a contract. Nobody had to meet, and nobody had to agree. That is what makes the next part the most remarkable thing in this story.

On the 26th and 27th of May, 1994, the newspapers reported what the two professors had found. And the pattern that had held for years, on the largest stocks in the country, broke in a single day.

On May 27th, the dealers in three of the most famous names on the market, Amgen, Cisco, and Microsoft, suddenly began using the odd eighths they had been skipping. Not slowly, and not quietly. That day. The spreads on those stocks fell by nearly half, almost overnight, and Apple followed the next trading day.

Nobody sent a memo. Nobody called a meeting. Nobody admitted a thing. The behavior simply stopped, all at once, the moment it was counted out loud in public.

Two years later, the government weighed in. The Securities and Exchange Commission issued a report, and it found that market makers had, in its own words, adhered to and enforced a pricing convention, one that discouraged them from competing on price. And then came the line that tells you exactly how this kind of thing works. The Justice Department ran its own investigation, and it was careful to say what it had not found:

"An express agreement reached among all of the market makers in a smoke-filled room."

U.S. Department of Justice, quoted in the SEC's August 1996 report

No smoke-filled room. No plot, no meeting. And yet the effect was the same as if there had been one. A convention had done the work of a cartel, and no one ever had to agree to join it.

The cost was not theoretical. Every one of those wider spreads came out of somebody's trade, not in a fee they could see, but in a price that was always a little worse than it needed to be.

The firms involved, including the largest names on Wall Street, settled the matter. The number was one point zero two seven billion dollars, across thirty-seven firms, covering roughly seven years of trades. At the time, it was the largest civil antitrust settlement in American history. The firms admitted no wrongdoing. They paid, they agreed to change how they did business, and some of them even agreed to record their traders' phone calls.

The Fix Was Arithmetic

But here is the part that should stay with you. The thing that actually fixed this was not the lawsuit, and it was not catching anyone. It was arithmetic.

To see the fix, you have to know where the eighth came from, and the answer is wonderful. Two hundred years ago, the money that circulated in America was the Spanish dollar, a silver coin worth eight reales. Eight. People called it a piece of eight. To make change, they cut the coin into physical wedges, into bits, which is why even today we say "two bits" and mean a quarter. When American stock markets opened, they simply inherited the fraction that was already in everyone's pocket. The eighth.

So the eighth was never a rule anyone chose. It was a two-hundred-year-old habit, carried over from a coin that no longer existed.

And it turned out you did not have to prove a single person guilty to shrink the spread. You only had to shrink the increment. In 1997, the markets cut the minimum from an eighth to a sixteenth, and in 2001, they went all the way down to a single penny.

The moment they did, the spreads collapsed on their own. On the most actively traded stocks, the quoted spread fell from about six and a half cents to under two. No convention could survive a tick that small.

In Plain Sight

Consider what actually broke this open. Not a raid, not a wiretap, not a confession. Two professors, counting the last digit of a number that everyone could see and nobody had bothered to add up. The data had been public the entire time. The pattern sat in plain sight, for years, in the price of every trade. It only took two people willing to count it.

So here is what to do with this. You will never quote a spread yourself, but you are always paying one, in every market you touch, and you should know which direction helps you. A wider increment helps the person on the other side of your trade. A narrower one helps you.

The next time you hear that a market has become "more efficient," translate it in your head. It almost always means the same quiet thing. The spread got smaller, and a smaller spread means more of the money stayed with the person doing the trading.

That person is you.

← Back to Research