When you ask someone in the financial industry how his job creates value, he’ll often answer that it enhances liquidity. Why are high-frequency trading and other forms of algorithmic trading good for markets? Liquidity, we’re told. What’s the potential harm from the Volcker Rule, which prohibits banks from engaging in proprietary trading? It could decrease liquidity. There are endless academic finance papers discussing policies to improve liquidity.
But one question that’s rarely asked is whether liquidity is good. To many, this seems obvious. Liquidity is a notoriously hard concept to define, but broadly speaking, it means the ability to find someone to take the other side of your trade — without it, markets break down. A liquidity crunch in the banking industry is commonly blamed for touching off the financial crisis and the Great Recession. Obviously, more liquidity is good, right?
Well, maybe not. Liquidity isn’t the only desirable characteristic of a financial market. Another concern is efficiency. In finance theory, efficiency means that asset prices incorporate all available information — in other words, that prices represent the best possible estimate of fundamental value.
If markets are efficient, it’s good for the real economy, because they will allocate capital to the people best able to use it. In an efficient market, if a stock goes up in price, it’s because the company now has a better opportunity for making money in the future. The company now can raise more capital for each share of stock it sells, meaning that the money is going where the economy needs it to go.
For a long time, economists have assumed that a well-functioning financial market could have both liquidity and efficiency. The basic models presented in textbooks assume that markets are complete — a way of saying that all assets are liquid — and efficient too. In some classic models of the trading process, efficiency and liquidity
But what if this assumption is wrong? What if there’s a fundamental tradeoff between liquidity and efficiency? Some models of markets suggest that this might be the case.
The simplest model with a liquidity-efficiency tradeoff is Nobel-winning economist George Akerlof’s model of the used-car market. In this famous construct, used-car dealers know how good each car is, but buyers don’t. When buyers aren’t willing to take a gamble, they won’t pay high prices for a random car. Therefore, the dealers will only sell the low-quality, inexpensive lemons, and all the good cars end up sitting on the lot. There’s no liquidity for the good cars, but the market is efficient, because everyone buyers pay the right price for the lemons. On the other hand, if customers are willing to gamble, they end up overpaying for the lemons and underpaying for the good cars — lots of liquidity, but no efficiency.
Things get worse if some traders are irrational. One way to get perfect liquidity is to have a bunch of bad investors who are willing to pay the wrong prices for things. But these same bad investors — noise traders, as economists call them, because their actions can seem to a rational observer like random noise — will end up paying the wrong price for stocks, bonds, options, etc. And if smart, well-informed traders don’t have enough firepower to stop them, the hordes of noise traders can push prices out of whack, making the market inefficient. This theory, which was formalized by Brad DeLong, Andrei Shleifer, Larry Summers and Robert Waldmann in the 1990s, provides another way that high liquidity can lead to low efficiency.
Some economists have made even more sophisticated theories of tradeoffs between liquidity and efficiency. For example, Sergei Glebkin of INSEAD in France models how the tradeoff can arise from the presence of large traders who move market prices with their actions (think Goldman Sachs or Vanguard). When big traders try to act on their information — say, to buy a stock they know is underpriced — their trades are so big that they move the price a lot. But then small investors pull back, because they know the big traders must have information they don’t. It’s Akerlof’s lemons problem all over again.
All of these models rely on very specific settings, with lots of assumptions. But there’s a general principle here. The easiest way to get liquidity is for lots of traders not to know — or not to care — if they’re buying at exactly the right prices. In other words, the easiest way to get liquidity is at the cost of efficiency.
If this tradeoff really does exist, it means economists — and financial regulators — need to think harder about whether liquidity or efficiency is more important for the smooth functioning of the economy. Will companies be better off if there are always people buying and selling their stock and trading their bonds?
Or will they be better served by a market where prices closely reflect the companies’ actual worth? Are some of the traders who claim they help the market actually hurting it? Next time someone tells you that their job is to provide the market with liquidity, perhaps you should wonder whether it just a matter of adding noise.
Noah Smith is a Bloomberg View columnist. He was an assistant professor of finance at Stony Brook University, and he blogs at