Jan

19

Reading Sartre's La Nausee, I came across a wonderful "counting" sentence:

"Three o'clock is always too late or too early for anything you want to do." (obviously untested and his opinion)

Is there such a dead time or point in the trading day to which it's more advantageous not to do anything? In the six and a half hours of the NYSE, does there exist a slice of the pie that is such an apex?

Looking at things, it's nine thirty to lunch, then lunch to three o'clock (bond market closing), and three o'clock till closing. Any hypothesis stick out at ya'll? How do you divide the day? What statistical tools should be taken into consideration?

Furthermore, can you break up the week, month, or quarter to find a dead spot that's too late or too early for entry?

Food for thought.

Russ Herrold comments:

Reading Sartre's La Nausee, I came across a wonderful "counting" sentence:

"Three o'clock is always too late or too early for anything you want to do." (obviously untested and his opinion)

Is there such a dead time or point in the trading day to which it's more advantageous not to do anything? In the six and a half hours of the NYSE, does there exist a slice of the pie that is such an apex?

I think of it differently — there are better times for some strategies, as a trader, than others within the pulse of a day.

Any hypothesis stick out at ya'll? How do you divide the day? What statistical tools should be taken into consideration?

Let's go to the track, and lean on a rail for a few days. The horse to watch: An ECBOT future, and only on non-major scheduled news days. The methodology: Counting track conditions with a tally sheet for a month or two. The stopwatch: five minute intervals (I have my computer on a weak tone sound every minute and increasingly stronger ones at the five's and quarter hours to remind me to look up and rate the market when I do this). Observations are in New York time.

Taking the notation of conditions, and doing reduction of the patterns after several days, I find this:

8:30-9:25 'pre-open' light volume, often scalpable, but also with no material liquidity if one ends up on the wrong side of an exogenous event (most of the news issues at 8:30, and so a gapping move play is often available)

9:25-9:30 'at the post' untradeable

9:30-9:50ish 'from the gate' choppy and not overly tradeable until all the underlyings in NY are both open and have finished running the market and are easily reachable overnight resting LMT and STP orders

9:50-11:50 'around the first stretch' trending and mean reversion plays will often emerge (about 1/3 of the time), or if not, range bound movements will grind up a player seeking to trade — a 'three strikes, you're out' approach works well here

11:50-noon, noon-13:00, 13:00-13:30 'on the back stretch,' see 'Feeding Time for Giants' discussed on this list last June (9-Jun-2006: A Cyclopean Workshop, also read more here)

Eminently playable — The chair's comments about mid-day loss of liquidity, and my observations at the time have covered this time frame already (watch the volume)

13:30-15:00 'coming into the far turn,' similar to 9:50-11:50

15:00-15:45 'down to the wire' as people realize the day is drawing to an end, and wind up intra-day or assume overnight positions before the 'rush hour' about to occur

15:45-16:00 'at the finish,' very hard to trade well as liquidity is bleeding off by the second (watch the volume)

16:00-16:15 'back at the paddock,' earnings of an index component will often move the futures, and laggards with resting LMT or STP can be picked off (watch the order book for such plays)

Similar patterns with different time bands exist on the FTSE100 (Z.FUT.LIFFE), the DAX (DAX.FUT.DTB), and the Aussie SPI (SPI.FUT.SNFE), for which I maintain tallies.

Furthermore, can you break up the week, month, or quarter to find a dead spot that's too late or too early for entry?
 
Well known dead spots are standing aside from using such time based expectation plays in front of major scheduled news; other trading approaches exist in their stead — setting up LMT's on either side of historical gap ranges to enter a post-Fed meeting announcement and to trail up an exit waiting for the peak of irrational exuberance worked well in the last year, but that play has been dead now for a while.


Comments

Name

Email

Website

Speak your mind

1 Comment so far

  1. Ryan Maelhorn on January 22, 2007 1:30 pm

    I am starting to think that the standard Monday through Friday work week can be used to aid one in trading.

    I don’t have the ready access to data that others have here, so I can merely comment on my own experience. It seems to me that there must be a lot of traders/investers that view the work week not too much unlike the standard 9 to 5 corporate cubicle dweller does. Every work week is a project. The goal is issued on Monday, and the Brass expects a completion by Friday.

    I have witnessed this in the markets I deal in by noticing that Mondays tend to be narrow ranged and turnaround days. A market that was crashing all last week may start to head back up, but not wildly, on Monday, and vice versa. Tueday through Thursday often provide the wideset swings and the best times to take profits. Fridays are usually also narrow range days, I personally chalk this up to the fact that most traders dread holding stock over the weekend, as during those 2 days a lot of news can occur. So, most people who have played the stock will be out by Friday, and most people interested in playing the stock on Friday will hold off until the next week.

Archives

Resources & Links

Search