Register for free to join our community of investors and share your ideas. You will also get access to streaming quotes, interactive charts, trades, portfolio, live options flow and more tools.
Register for free to join our community of investors and share your ideas. You will also get access to streaming quotes, interactive charts, trades, portfolio, live options flow and more tools.
I tested Oddball on the Nasdaq I only had hourly data going back to Sep 4, 2002.
It loses 198 Nas points in that period. I used six hourly intervals starting at 0930 Eastern. The authentic Oddball recommends using six hourly periods starting at 10:00 Eastern.
Ignoring the first signal of each day doesn't help the results.
I could easily have amde a mistake and haven't yet done any thorough checks of the coding I used.
I doubt the short period in my study is representative. It is my beief that the PPT have propped this market and thus Oddball may have been going short with the reasonable expectation of an unmanipulated market going down.
Am going to try it for the S&P as per the original Oddball.
My eSignal intraday data only goes back a couple of months so I can't do a longer test. If anyone has a source of hourly intraday data going back further than the beginning of September please let me know.
MM, u posted on another thread today about the oddball method and Zeev boxes, specific vol etc. I think this is praps a better thread to continue this train of thought due to less clutter.
The oddball system looks interesting and I haven't tested whether it works but the author claims it does!
It doesn't use volume or price as inputs but simply looks at the last hour of trading and compares the advancing issues on the NYSE in that hour to the same hour of the previous trading day.
If the advancing issues is up by 3% or more it is a buy. If it is down by 1% or more, it is a sell. It is always either long or short depending on the last signal.
If something so simple works then there would be a few ways I might tweak it to try to increase performance.
1) introduce the concept of stop losses
2) introduce a neutral state where it was 100% cash
3) allow the 3% and -1% triggers to move adaptively
I have a couple of ideas how I might implememnt tweaks 1 & 2, and some vague thoughts on tweak 3
If $NDX breaks this week's high than both $NDX and $COMPQ will have done so. That will be my signal to get out.
It is possible that QQQ can break without $NDX doing so, by the exuberance of amateur traders. That would, IMO, be a false breakout. Whether it makes any difference is unlikely, but I see no advantage to basing the stock on the QQQ apart from it not requiring monitoring.
Back in short QQQ $28.03
Stop if Nasdaq 100 recent high taken out by 2 pts.
Another QQQ short here $27.90 with stop at $28.03
IBM, F etc that are having to take money from operations and put them in the employee pension funds. These companies have announced they are having to do this.
Shorting QQQ $27.58 here with stop at $27.68
$14B in overnight RPs drain today.
Fed just announced a new $5B RP this morning for total drain of $9B so far.
Wouldn't be surprised to see another one announced in the next hour or so as that's a pretty big drain.
"I believe some proprietary swing algorithms that are quite popular triggered downswings today"
Can you tell us any more about those algorithms, or how their signals can be accessed and verified? If you can't tell us, then can you just tell me<g>?
If your data told you that an index would only close down more than 5% ten times out of a period of 1000 trading days, then would you buy it intraday if it were down 7% ?
I ask becaus of your earlier data about closes having predictable characteristics of previous closes, and then the post I am responding to subsequently mentioning bugs.
It could be the case that the bug involves ignoring the possibility that the ten times it closed down more than 5%, it was previously down 7% intraday. That might be an obvious thing to check, but do you have a process for trying to spot such bugs?
I believe it is more valuable to assume absolutes exist. Maybe they don't exist, but in that case there would be not much point in worrying about it, or at least I don't think there would be.
I think you are mistaking the simplicity of his system for luck.
Even though the system was very simple, it had some solid foundations that he didn't violate:
1) Buy only strong stocks
2) Buy just above support
3) Use trailing stops set just under support
BTW, have you missed the posts made on this thread over the weekend?
Stopped out again from QQQ for $0.13 loss. The EOD rally is not meant to be.
Real battle going on here to hold Nas LoD.
Probably the investment banks selling to the feds<g>
Bought QQQ at $27.16 again with $27.03 stop
My QQQ stop just hit for a $0.10 loss.
If some volume buying on the Nasdaq large caps comes in I might try another long until the close.
Agree on the reflex rally. Going long QQQ here at $27.16 for 90 mins or so. Stop at $27.06
Nasdaq 100, MSFT making new LoD.
Air beneath both for a couple of percent if volume buying doesn't come in v soon.
Lots of put buying last couple hours and Fed pumping today so there is a chance it will be saved.
Fed pumps $8.5B today. $4B drains today from Friday's pump.
Net injection today $4.5B
Today's $8.5B expires tomorrow for a total of $14B draining tomorrow.
Greenspan pumped $8B yesterday. Half of that drains back out today, the other half on Friday.
The $5.5B pumped in on Friday last week drains out tomorrow.
A further $13B already pumped this month is due to drain on various dates in December.
Greenspan is going to be busy pumping if he wants to maintain the status quo.
Put into perspective, the total amount of money that trades in QQQ and MSFT combined on a typical day is about $4B
Fed pumped $8B today.
VLNC could go either way.
If their technology is as good as they say then the company will survive.
If it isn't then they won't.
I don't see how as investors we can answer this question so unless you have specific technical knowledge I think it is a gamble. Obviously, the downside is low compared to how much you would make if VLNC ever went back to its former highs. But those highs were set in a rampant bull market so likely that the opportunity for such prices will no be seen.
Are you holding any? What do you know about the company?
Thanks for the responses so far. Seems that you are a very disciplined trader. Exit strategies can be more important than entry strategies IMO. What exit strategies do you have for scalps?
Do you have a mathematical system for setting stops at a level designed to minimise whipsaw?
By 90, i mean 90%. I was thinking of BPNDX, which consists of 100 charts so 90% is 90.
The free trial sounds interesting, but I'm not about to spend hours or days trying something out when I'm sure someone else out there has already tested their theories and documented the results. If you don't know of statistical results of PnF buy signals then any comments on your experiences with them would be appreciated, if you are OK with sharing that kind of info.
BTW, do you get an exit strategy to finish off the trade initiated by the buy signal? Or do you just wait for a sell signal?
Chun-Li
OJ, have you or someone else analysed how many of a sample of bullish PnF charts lead to a tradeable gain?
I have a vague understanding of what a PnF chart is but haven't used them. It seems that a buy signal (and I know the breakout signal you are discussing is more than a standard PnF buy) is likely to fail due to the reliability of high BP's to lead to lower prices.
If a BP of 90 tends to mark a top for instance, then that means that 90 PnF charts are "wrong" doesn't it? Or is it that the buy signals are old and have already yielded profits? This is just top of the head thinking, hence my question as to any statistical evidence of the usefulness of PnF signals.
One more thing I should mention.
The box measuring system I just described with example is only valid, in Darvas's view, for stocks breaking out above yearly highs.
If you want to measure boxes in a declining stock, then I guess the appropriate similar system would be to apply it to stocks making new yearly lows, and adjust the algorithm to find the bottom of the box first by looking for a low not beaten for 3 days, then find the top by going back three days from the setting of the bottom, and looking for a high not beaten for 3 days.
Zeev's system seems to be more flexible and useful in that
1) boxes do not require the setting of new yearly highs or lows
2) the top/bottom of one box becomes the bottom/top of the next box up/down
3) the limits of the box are influenced by where the box limits were last time the stock was trading around the area of the current box, if that was recent. In other words, if a stock breaks up into a new box and then comes back down past the breakout point, it is then considered to be back in thte old box, rather than having to wait to define where the current box is.
4) Zeev uses intraday charts to form boxes, rather than just daily
5) Zeev's boxes don't have to be horizontal, they can be part of a sloping channel
I will let Zeev know about this discussion so he can comment on or correct my impressions above, if he feels so inclined.
Top is when stock has broken out from previous box and has subsequently not made a new high for 3 days.
Then bottom of box is any point after the top when a new low hasn't been made for 3 days. So the bottom is set after the top, but can be set before the top is known.
Consider the following daily price ranges below for an imaginary stock that prior to Day1 has an all time high of 19:
The top is at Day3. You only find this out on Day6, when Day3's top has not been breached for 3 days.
The bottom is on Day4. You only find this out on Day7, when Day4's bottom has not been breached for 3 days. It is the first daily low since the top was set on Day3 (even though this fact was not known until Day6) that has not been breached for 3 days. So the top is 27 and the bottom is 21.
The gerryco.com and sethi.org algorithms would have started looking out for the bottom on Day5 and Day6 respectively. gerryco.com would have the bottom as 22 set on Day5 and dicovered on Day8. sethi.org would have the bottom as 23 set on Day6 and discovered on Day9.
Day1: 20 - 25
Day2: 22 - 26
Day3: 22 - 27
Day4: 21 - 26
Day5: 22 - 25
Day6: 23 - 26
Day7: 24 - 25
Day8: 24 - 26
Day9: 25 - 26
Day10: 24 - 26
easiest way is to look for the top first, then once u find a top, immediately go back 3 days and start the bottom search from that point.
i.e.
If we start at Day1 and by Day7 we decide the top was Day4, then we go back to Day1 and start counting for 3 days of not finding a new bottom.
Maybe not a good explanation but no more time to go into it right now!
the algorithm does not correctly set the bottom of the box. See my post on the darvas thread for details.
http://www.investorshub.com/boards/read_msg.asp?message_id=594693
The algorithms at the sethi.org and gerryco.com sites DO NOT correctly identify Darvas boxes, particularly with respect to defining the bottom of the box.
Specifically, the problem is that although the bottom of the box cannot exist before the top exists, it can exist before the top is known. But the methodologies above don't start looking for the bottom, until the top is known.
Example scenario - A top is set on "Day 1" and a bottom is set on "Day 2":
On "Day 2", sethi.org & gerryco.com have not yet found the top of the box, so haven't started to look for the bottom yet. gerryco.com starts looking for the bottom on "Day 3" when it has only a tentative top, and sethi.org starts looking on "Day 4" when it has a confirmed top. Both miss the "Day 2" bottom and incorrectly define the bottom as the bottom of a later day's bar.
Not any more. It went off.
http://www.afxpress.com/afxpress2/afx/bn243598.xml.html
Haven't seen any mention of the bombing near the US embassy in Bogota on any US news service. It has been an hour now.
This IMO is PPT action designed to get some quick buying in before news of the bomb in Colombia broke.
No fave oil stocks but I trade the crude futures occasionally.
What are your thoughts on price of oil. My guess is up into an Iraq conflict, maybe even hitting $40. But I'm not going to trade it on the way up. Plan is to wait for escalation, then short as long term it is liekly to get back into the $18 - $24 range.
Incidentally, looks like US is stocking up. Rigs normally seeing one tanker a week are now seeing three. This is second hand info. I am not in the oil biz.
5 day RP for $5.5B just got posted. Net drain today now only $0.75B.
What are the particular characteristics of accumulation you identified that tell u this? Thanks.
>>Looks like accumulation<<
Fed, have another look at the frb site...
Fed, what's the latest in the day you've seen RPs announced on the app.ny.frb.org site? I guess they can do it anytime in the day at will although they generally seem to post em before market opens.
No fed desk activity posted yet today so we are lookin at a $6.25B drain. What time of day does that money get taken back?
Someone's trying a scam with the premarket.
Bad ticks 40 - 50 cents higher than market being shown for MSFT, INTC, CSCO, QCOM & DELL. These just happen to be the five biggest weightings in the Nasdaq.
Here's the link you are after:
http://app.ny.frb.org/dmm/mkt.cfm