I drew this without specific dates because I don't intend this as any kind of timing indicator. Rather, I wanted to illustrate the tremendous growth of herding behavior in the stock market since 2000.
In the chart, I'm looking at a moving window of 60 days and counting the number of days within that window that either have 2/3 or more of stocks traded as advances or 2/3 or more as declines (NYSE issues only). So we're looking at relatively one-sided days in which advances lead declines (or vice versa) by a ratio of roughly 2:1 or better.
In 2000 and 2001, such one-sided days were the exception; because stocks traded in quarter point increments, many issues remained unchanged. The ratio of unchanged stocks to advancers and decliners has steadily fallen over the years. Now, out of over 3000 issues traded, it's unusual to have 100 unchanged stocks; in 2000, over 500 unchanged issues were the norm.
Interestingly, the ratio of unchanged issues to total issues traded has fallen significantly since July, 2007, so it's not just decimalization that has led to the shift. Program trading and the inclusion of more stocks in baskets that are traded--not to mention the inclusion of more stocks in ETFs (including leveraged ETFs)--may well account for this phenomenon. Small cap issues are no longer a market backwater.
The average number of issues traded daily since 2000 has actually fallen. Nevertheless, there is far more money--and far more money managers--chasing the same returns. It does, indeed, appear that they are chasing returns in part by chasing each other. Incredibly, we're getting close to the point where nearly half of all trading days are relatively one-sided. For that to happen, sectors and indexes need to be more tightly correlated with one another.
It used to be called a market of stocks, not a stock market. Now it seems to be a market of stock groups. When the biggest bets are being placed on entire indexes and sectors via futures contracts and ETFs, shares are likely to move in unison. That dynamic does not appear to be changing any time soon.
.
Posts
6 comments:
very interesting research. Do you think that this perhaps increases the influence and importance of technical factors in relation to fundamental factors over time?
Im thinking that this type of data can also be useful in providing a frame of reference for longer term trends vs shorter term trends in TICK correlations.
As always, great post
so.. and on this market we can expect more from sectors confirmation when market breakout, dont saying about stock breakout - setor confirmation
Hi Watson,
Yes, good points, I think that short term traders have to think about how they can front run the herd on these intraday moves.
Brett
Hi DmytrenkoAV,
Yes, I think you're pointing to one reason I am so skeptical of market moves that *don't* have good sector participation. In this environment, the trending moves really should take the broad list of stocks with them.
Brett
Hi Brett.
I put this post on Angry Bear...can you come over to explain what it means and any implications for us?
Rdan
http://angrybear.blogspot.com
As always, there is the question of causality. That is, are the higher correlations in stock price co-movements caused by "herding behavior"? Or alternatively, are the higher correlations merely because at this point in the economic cycle, values are effected more by macroeconomic influences than microeconomic influences? Just because the co-movements are greater doesn't mean that they are forecastible...
Post a Comment