Thursday, June 07, 2007

Stock Market Performance by Hour of Day

Here we're looking at the Dow Jones Industrial Average on a daily basis (light blue line) from 2004 to the present. I've broken the Dow down into three components and set these to equivalent prices as of 2004 to highlight differences in performance (click chart for greater detail).

The first component is the Dow's performance during its first hour of trading (dark blue line). Note the steady ascent, as we've seen solid performance during the Dow's first hour. Indeed, nearly three-quarters of the entire bull market since 2004 has occurred during the Dow's first hour of trading. Had you done nothing more than buy at the previous day's close and sell at the end of the first hour of trading the following day, you would have picked up almost 3800 Dow points.

The second component is the Dow's performance during its middle hours (pink line): the hours between the end of the first hour and the start of the last hour. Observe that this time period has not participated in the bull market whatsoever. In fact, from 2004 to the present, the middle hours have *lost* a total of about 42 Dow points. What that tells us is that a good part of the daytrading market has not seen a directional bias, despite the significant ascent of the Dow over that period.

The final component is the Dow's performance during the last hour of trading. Rennie Yang of Market Tells notes that this has been weak of late, which he interprets as caution on the part of the smart money. We can see that the last hour has underperformed over the last two years. About a quarter of the Dow's overall gain since 2004 is attributable to the last hour of trading; since 2005, however, the Dow's last hour has actually lost money.

My interpretation of these findings is that the U.S. market has been a follower, not a leader, of other world markets. When overseas bourses rise in value--and we've seen stronger gains in EAFE than in the States--that strength is reflected in the next day's open in the U.S. and thus the first hour of trading in the Dow cash index. If that is the case, then an important implication is that participation in the bull market has hinged upon the trader's ability to hold U.S. positions overnight and capture that overseas strength. My breakdown of S&P 500 returns by overnight vs. day session supports this conclusion.

The takeaway? Just because a market is trending overall doesn't mean it's trending at the time frame you happen to be trading. Extrapolating from the day to day market to shorter time frames can be hazardous to your wealth.


The Last Hour of Trading as a Sentiment Measure

Who Controls the Markets?


Johan said...

Wouldn't it be interesting to see if this is a contemporary issue or to check this on a longer time-frame to see if it has been like this for decades?

Patrick Watson said...

It would be interesting to see this same analysis with the Diamonds (DIA) rather than the DJIA. I'm not sure how feasible it really is to buy the Dow at the exact opening price consistently.

ryang said...

This isn't quite the same chart as Brett showed this morning, but it does give you a good long-term look at the market's performance during the first hour vs. last hour. See the chart here.
The 'Last Hour' indicator, as we call it, is calculated by comparing the market's performance during the first & last hour of the trading day via the formula (today's close - today's 3pm price) - (today's 10:30am price - yesterday's close). A rising line means the last hour is outperforming the first hour. Note that it generally acts as a good lead indicator for the market, albeit lead times can vary. As Brett noted, the last two years has seen major underperformance during the last hour, which has historically had longer-term bearish implications for stocks. History suggests that the current divergence is likely to shift dramatically at some point in the not-too-distant future as the last hour starts to consistently outperform the first hour. This will send the Last Hour line shooting higher, and if history is any guide that will coincide with prices skidding lower. It's only when the divergence of the last couple of years is completely unwound that the Last Hour and price are likely to start moving in tandem again (as they did throughout most of the 90's.)

Brett Steenbarger, Ph.D. said...

Hi Johan,

Yes, I do have long-term data and that would be interesting. When I've looked at the relationship in the past, I've noticed shifts during different market periods. Rennie Yang's comment below speaks to this--


Brett Steenbarger, Ph.D. said...

Hi Patrick,

Yes, buying the Dow at the open would require DIA to get an accurate performance read. Otherwise, per my analysis, you'd have to buy the Dow stocks the prior day, MOC to benefit from the first hour of trade the next day.


Brett Steenbarger, Ph.D. said...

Great historical perspective, Rennie; thanks much, and thanks for the link--


Vlad said...

Interesting evidence, completely unproven explanation.

In general, if A and B exhibit a certain amount of co-movement in prices, it is not trivial to decide whether A leads B or B leads A.

Your interpretation that the US market has been a follower of other world markets does not seem to follow from your evidence alone.

Brett Steenbarger, Ph.D. said...

Hi Vlad,

It is interesting indeed. The liquidity provided by low rates in Japan has fueled not only the purchase of higher yielding currencies/debt, but also the purchase of higher beta equity assets (emerging markets). That has made the US the lagging market relative to EAFE. Nearly all the rally in the US market is attributable to movement from the US market close to the US market open: periods in which the superior-performing markets in Asia and Europe have been open.

To be sure, there are important earnings stories underlying the U.S. strength. International markets are not the whole picture. But if you look at dollar inflows into funds over the last year or so and if you talk with professional portfolio managers at global/macro hedge funds, you'll find that international bourses are leading the way vs. U.S. It's very difficult for the average U.S. retail trader to see such dynamics; we all tend to focus on what is most readily apparent--and that's usually the market we're trading!