From Physics, a Smarter Way To Trade Stocks?

SMITH BRAIN TRUST – There must be a science to stock trading, right? Maybe it’s physics.

In new research, Maryland Smith’s Albert “Pete” Kyle uses methods of physics to develop a dynamic model to measure financial market liquidity and make better investment decisions.

Working with Anna Obizhaeva of Russia’s New Economic School, Kyle applied physics’ dimensional analysis to the field of market microstructure. Their working paper, Market Microstructure Invariance: A Dynamic Equilibrium Model, is posted at SSRN.

In finance, market microstructure explores how shares are traded, how information is interpreted into price changes, and how optimal trading strategies are developed.

"You would think it's very easy to measure liquidity but in fact it's kind of hard to measure liquidity," says Kyle, speaking to the Australian Financial Review about his latest research. "You can look at bid/ask spreads, or if you're a gigantic asset manager you can look at thousands or hundreds of thousands of trades, and you can eventually accumulate so many transactions you can measure your market impact.”

So, where does the physics come in?

Cambridge physicist G. I. Taylor used dimensional analysis to calculate precisely the explosive yield of a U.S. atomic bomb test in New Mexico in 1945, by analysing a sequence of time-marked pictures of the mushroom cloud years later. And that’s where Kyle and Obizhaeva begin. The researchers start from the premise that the liquidity of the market is based upon how much information is coming into the market at any given moment.

If there’s very little information coming in, because there’s no new information available, the market should be more liquid, Kyle says. But if there’s a lot of information coming in, because of a surge of uncertainty, volatility will be higher, but liquidity lower.

And, according to Kyle and Obizhaeva’s model, prices of very liquid, or heavily traded, stocks are likely to be more accurate than the prices of stocks that are less liquid. If you can trade a stock easily, the reasoning goes, you’re more likely to do research.

"Our paper basically says that all you really need is volume and volatility. What we're trying to do is derive a formula that says your market impact or your bid/ask spreads or various measures of liquidity should be a particular formula that has these inputs – volume and volatility."

The formula that has come from the research is counterintuitive, Kyle says. The formula has powers in it of one-thirds and two-thirds, which he says mimics the dimensional analysis in physics.

"The theory that we developed basically says that in very big liquid stocks, people in some sense trade twice the volume, but they only have half as much impact and they only have half as much information that they're profiting on,” Kyle says.

"Everything in the model translates into powers of one third and two thirds."

Kyle – whose seminal 1985 paper, nicknamed Kyle '85 by many researchers, changed the thinking on the way markets operate – says his latest research has implications for asset management and for policymakers about how investment funds are managed.

In periods of high uncertainty and high volatility, which often make markets less liquid, Kyle says the research would recommend that investors cut back on quantities traded, execute trades more slowly, or simply make fewer trades overall.