Why is idiosyncratic stock market volatility so low?

We find that the historically low IR [idiosyncratic risk] can be explained by the changes in firm characteristics that take place since the 1990s.

…the number of listed firms has fallen dramatically and the composition of listed firms has changed considerably, with public firms becoming larger and older.  We show that there is a stable relation between firm-level idiosyncratic risk and firm characteristics…we find no evidence that IR increases with institutional ownership…

That is from Bartram, Brown, and Stulz, in the NBER working paper series.


"…the number of listed firms has fallen dramatically": is that important to what was called "dynamism" a few posts ago?

Its public policy of cutting costs and driving up profits.

The worst way to cut costs and jack up profits is actually paying US workers to make things to sell.

Eg, Elon Musk defies public policy. Exponentially increasing costs, zero profits. Note he keeps issuing and selling more Tesla shares to pay his higher costs of paying workers to build and work in bigger factories. SpaceX uses private offerings to pay more workers in bigger factories.

Jeff Bezos has exponentially increased costs to keep profit to near zero. Paying workers to build and run exponentially more server farms and warehouses.

But for most US corporations, cutting costs by closing US factories to jack up profits, and then buying back shares to keep prices from falling in line with declining assets. Eg Apple which fired Jobs for building US factories and now owns no factories to cut costs but has very high profits used to buy back shares. More public corporations follow the Apple model. GE, for example bought back shares when profits were high in excess of GE's liabilities today even as GE has far fewer factories, and thus lower costs.

The paper's behind a paywall, and the abstract doesn't specify what factor model they're using. Are these idio returns adjusted to the CAPM, Fama-French, FFM3, Barra or something else? That seems important enough that it should be specified in the abstract.

The late 1990s stock market was notorious for a large divergence between growth and value stocks. If they're just using CAPM, then this would be captured under idio volatility. The fact that the Nasdaq moved separate from the Dow in 1999, shouldn't really be considered a "firm-level" difference to today's market. I think any reasonable person would characterize that as a macro difference.

I second that opinion by Doug. I would add that maybe share buybacks have an effect on IR? Keeping volatility down due to "buy the dips" mentality, which also explains why the US stock market bounced back so quick after Monday.

They're using the five-factor Fama-French model. You can find an earlier version (five days earlier) of this paper on SSRN. https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3107798

The authors qualify the conclusion as applying "absent financial crises or irrational exuberance," which in most people's book would include the late 1990s.

As for buybacks, Ray, the word doesn't appear in the paper, but I would speculate that tends to be characteristic of older, larger companies, just return of capital by a different name.

As part of liquidating capital. Eg Sears/Kmart, GE,...

Thanks for the link, Tom! After reading the full paper, I am much more convinced by the claims.

would be a lot more helpful to provide a brief analysis of what is linked rather than just providing links to all sorts of random stuff on the internet a lot of which is behind pay walls. I would note that the investment universe today is way larger than it was in the 1990s given that it now includes lots of firms in Asia, South America and Africa, that were not investable 25 years ago. Though whether this is germane to the article is anyone's guess because the article is not readily available.

"The paper’s behind a paywall..."

On this topic, just read this Verge article

Meet the Pirate...

A quick read of the article leaves me feeling a bit incomplete. However, I admit to being an economic dilettante. (And as much as graduate programs in economics demand numerical fluency, I wish they also demanded greater language fluency. A good writing course should be mandatory. That is a general complaint with economic journal articles. Simply, I wish the authors would rewrite their findings for a more general audience i.e. like an article for the Wall Street Journal.) Still....

"However, while firm characteristics are helpful in understanding the dynamics of IR in normal times, our study does not address the puzzle that IR can become extremely large in crises or unusual times, a puzzle that parallels the puzzle of market volatility exhibiting such large movements as well (Schwert, 1989)"

"These results show that absent financial crises or irrational exuberance, idiosyncratic risk can be expected to be low because firm characteristics are much different from what they were when idiosyncratic risk was especially high in the late 1990s."

In my observations, increased concentration in the marketplace may contribute to more seismic shifts in the stock markets in reaction to a minor shift in indicators (i.e. interest rate changes). Rather then arguing that bubbles are bursting, I would wonder if giants can cause large ripples when they move in a small pool.

That may be a distinction without a meaningful difference for investors, but with significant differences for how policymakers should respond to events.

Diversification is an important key to building a stable portfolio. But in an age of giants walking the earth how easy is it to build a stable structure that can survive their movements?

John Cochrane on this topic


slightly off topic, but relevant:


Absent a lot of lousy judicial decisions, the market for corporate control would ordinarily produce volatility as oppportunities for beneficial gains from hostile takeovers were identified and pursued. But the courts are protecting entrenched management and firms are not subject to the policing that an active market would provide. It should be no surprise that "we find no evidence that IR increases with institutional ownership."


When have hostile takeovers been beneficial? Most mergers, takeovers end poorly, because they are done to cut costs.

idiosyncratic - peculiar or individual

I had to look it up. Start with occam's razor, if these are a random list of firms then all shocks are idiosyncratic, the null hypothesis.

So the question is asked wrong, the real question is why do official stock exchanges suffer aggregate shocks? Tell me the value added structure of the market because aggregate shocks happen to value added chains.

Comments for this post are closed