Fractional Pricing and the Teenie-Spread Era
US stocks used to trade in fractions of a dollar as small as a sixteenth — a "teenie" — and that coarse pricing grid mechanically forced wider spreads than decimal pricing ever allowed.
Until 2001, US stocks did not trade in cents. They traded in fractions of a dollar — eighths, going down to sixteenths for the most actively traded names, where a sixteenth of a dollar, 6.25 cents, was nicknamed a "teenie." A market maker quoting a stock at $50 bid, $50 1/8 offer wasn't choosing that spread because it reflected the stock's true risk — the exchange's minimum price increment simply didn't allow anything finer. That coarse pricing grid was itself a source of guaranteed market-maker profit, and it is one of the clearest examples in market history of market structure rules directly determining who earns an edge.
Before decimalization, the smallest allowed price increment on US stocks was a sixteenth of a dollar (6.25 cents) — far coarser than a stock's true bid-offer risk required — and that gap between the minimum tick and the "fair" spread was pure, structurally guaranteed profit for market makers.
Why the fraction system inflated spreads
A market maker's true cost of providing liquidity — inventory risk, adverse selection, order-processing cost — for a liquid large-cap stock might justify a spread of only a few cents. But if the smallest allowed increment is a sixteenth of a dollar, the market maker cannot quote anything tighter than that even if competition would otherwise push the spread down further. Every stock effectively had a price floor on how cheap trading could get, unrelated to the stock's actual liquidity. Academic studies comparing spreads before and after the US moved first from eighths to sixteenths in 1997, and then to full decimal pricing in 2001, found that average effective spreads dropped sharply at each step — direct evidence that the fraction system, not underlying risk, was setting the floor.
Worked example
Under eighths pricing, a Nasdaq market maker quotes a heavily traded stock at $40 bid, $40 1/4 offer — a 25-cent spread, because the eighth-dollar increment forces at least a two-tick gap for a dealer unwilling to trade at the exact same price on both sides. A market maker who buys 1,000 shares at $40.00 and sells them at $40.25 pockets $250 on the round trip, largely independent of any real inventory risk taken, because the stock rarely moves 25 cents against them within the minutes the position is held. After decimalization, the same stock, with the same true liquidity, trades at a one- or two-cent spread — the same round trip now nets $10 to $20 instead of $250, a reduction of well over 90%, even though nothing about the underlying company's risk changed. That gap is, almost entirely, what decimalization took away from market makers and handed to investors as lower trading costs.
What this means in practice
Fractional pricing is worth understanding today because it is a clean case study in how a market's tick-size rule sets a floor on trading costs regardless of competition — the same tension shows up now in debates over tick sizes for illiquid stocks and in crypto markets choosing their own minimum increments. A market maker's edge was never really about skill in this era; it was about a regulatory grid that made a spread wider than it needed to be, and that edge evaporated the moment the grid got finer.
It's tempting to think decimalization simply "helped investors and hurt dealers," full stop. It's more nuanced: tighter spreads also reduced the profit available to compensate market makers for providing liquidity in less-active stocks, and some smaller, thinly traded names saw quoted depth thin out after decimalization because dealers had less margin to work with — the benefit was concentrated in the most liquid names.
Related concepts
Practice in interviews
Further reading
- Bessembinder, 'Trade Execution Costs and Market Quality After Decimalization'