Retail traders still learn by watching the ‘smart money’

Michael Maiello | Jun 08, 2021

Sections Finance

Since the 1980s, equity markets have evolved from dealer-based models to more complex interactions among myriad traders pursuing separate and often opaque agendas. Digitization and algorithmic trading have created a complex and nuanced market structure, one that researchers have struggled to understand and model.

As these changes occurred, and particularly when physical trading floors went virtual, many in the market were concerned about losing transparency and creating an environment where some investors could maintain an unfair edge. Baruch College’s Ayan Bhattacharya, a visiting professor at Chicago Booth, and Cornell’s Gideon Saar have created a theoretical, dynamic model that seeks to explain an important aspect of financial markets, and they use it to demonstrate that less-informed traders, such as retail traders, can still learn from more-informed ones, including professional money managers. 

In the older, dealer-based models, each exchange-listed stock had a market maker providing liquidity for investors seeking to trade it. Anybody buying and selling could rely on the market maker to complete trades in return for part of the difference between the bid and ask prices on the stock. The market maker was officially neutral, trading regardless of what they did or didn’t know about the stock and its movements. 

Now, however, stocks no longer have a single market maker. Instead, the market is divided into “makers” and “takers.” Investors also more often put in limit orders, which specify the prices at which they would buy or sell a stock, and then frequently cancel or modify the orders.

“The old model was static and easier to understand,” Bhattacharya says. “What happens now is dynamic as people come in, trade, and go away, with everyone competing.”

Limit order books, in which traders specify prices for buying or selling, provide the canvas for a model in which “a subset of traders knows more than the rest of the market,” as Bhattacharya describes it. Because this is theoretical rather than empirical research, the model leaves unspecified what the smart money knows about a stock. It could be about the next quarter’s earnings or about a supply/demand imbalance forced by the liquidation of a large and leveraged hedge fund, for example.

Informed traders tend to “make” liquidity in illiquid markets and “take” liquidity from liquid markets, the research finds. This means that in an illiquid market, the informed traders know what price they want and will walk away if they don’t get it. When the market is more liquid, informed traders will take the trades—buying from and selling to what the researchers call uninformed traders. In short, the smart money knows its prices and won’t accept anything less, or will pounce if its prices are abundantly available.

Uninformed traders, meanwhile, want to figure out what’s motivating the smart money and turn that information to their advantage. That has always been the case, and the research finds it continues to be so. 

In Bhattacharya and Saar’s model, the evolving bid and ask prices from order books, including the volume of trades executed and those canceled, all contain information that informed and uninformed traders use in competition with each other. Order books are more complex now, in a world without a single market maker, but they still contain useful information. 

“Why do we have messages in the market that don’t lead to immediate order execution?” Bhattacharya asks. “We suggest in our model that the entire limit order book readjusts prices to reflect information in the market.”

The model helps academics understand the structure of an increasingly complex and competitive market and could eventually form the basis for empirical studies of these dynamics, including questions about what information is transmitted, how quickly it flows from one trader to another, and even when, or if, an uninformed trader becomes an informed trader. By introducing a new methodological toolbox—the core of which is a recursive framework for the analysis of order-book prices—their work enables researchers to explore fresh territory in this area. 

Reliable models for how markets spread information would be valuable to investors— especially those trying to understand what moves stock prices—to regulators, and to public companies that are communicating with investors. The research paints a picture of the markets that’s far more nuanced and detailed than the simpler models of yore.