A Simple Model of Market Volatility
Credit constraints are enough, right?
Suppose that we are at equilibrium. Somebody receives a signal of the state of the world. They would like to run out and profit on it in the markets, but they are constrained by their lack of capital. Everyone knows that people face credit constraints, so a sudden shift in the price of a security is evidence that the true price shifted by more than is apparent. Thus, it would be on average profitable to “trade on momentum”. Further, if someone received only a small signal of the state of the world, then the price will shift beyond the true price.
Let me put it concretely. Suppose that there is a prediction market on an event – say, “will the Ayatollah Khamenei not be the leader of Iran at the end of 2025?” on Polymarket. Someone within the Israeli armed forces gets wind that they intend to assassinate the Ayatollah, and so he rushes out to buy as many yes shares as he can afford. However, he cannot buy enough to shift the market to the true probability, so other traders observe only a small change in the odds. Since everyone knows that there are credit constraints, this is itself evidence that there is a greater change in the underlying probability, so the price will change by even more as traders act. Of course, had this update been due to a trader making only a small adjustment to the price, and stopping there, people would still be right to, on average, continue moving the price along. Thus, we get volatility, even if on average the price will tend toward the correct price.
I cannot speak for the originality of this thesis. I have not gone and checked the prior literature. I do think it is interesting that credit constraints, and credit constraints alone, are sufficient to generate excessive volatility. That there is excess volatility is largely agreed upon, I think – my understanding is that Shiller’s work on stock market volatility is widely accepted.
I would like your views on this. What should I read? Should I pursue this as a paper?
The uninformed momentum traders have to undershoot the true price on average for the strategy to be profitable. I don’t know how excess volatility is defined, but shouldn’t the market’s underreaction when the credit constraints bind even out the overreaction when the momentum traders overshoot?
Copying a comment someone posted in the Manifold discord:
> “momentum trading causes volatility” is somewhat undermined by there being no predictable momentum in mature/efficient markets, yet there is volatility (more volatility than new information justifies as shiller shows) right? under this thesis i would expect there to be more momentum effects in bitcoin than in SPY yet there are ~none in either more broadly volatility is consistent with EMH but momentum isn’t