My AB tick charts always seem to be different than charts
created by other traders. I have always attributed this to my real-time
quote source being Interactive Brokers. I know IB sends out quote data as
time based snapshots rather than every tick. However recently I’ve determined
that my AB tick charts are different than other charting software charts also
based on IB data. Since today was a short and slow pre-holiday trading
day, I decided to use the time to compare an AB tick chart with a MarketDelta
(i.e. Investor/RT) tick chart. I ran both charting packages using the
same IB TWS. AB was setup with a single 10t candle chart of the Mar 08 YM
futures contract. The chart also has a 10 period SMA and a colored volume
indicator. The chart was intentionally kept simple to minimize AB’s
CPU usage. By monitoring AB’s CPU usage, I could see that it would
fluctuate between 1% and 5%. So in other words, CPU usage should not be
an issue in AB’s charting since it’s no way near 50% (i.e. my
real-time chart refresh interval is set to 0).
I do not know Investor/RT very well, so I use the default
settings. The Investor/RT CPU usage vacillates between 0% and 1%.
The 10t chart is the same: Mar 08 YM 10t candle chart, 10 SMA, and colored volume.
I let both programs run through the trading day to collect data
from the same IB TWS. Next I decided on a small time slice from 11:00am
EST to 11:10am EST to evaluate the differences between the charts. The
first thing I could easily notice is that Investor/RT has more bars than AB.
AB created 14 10t bars and Investor/RT created 22 10t bars between 11:00am and
11:10am. Next I noted the timestamp, OHLC and volume of each tick bar for
both charts in a spreadsheet. The total volume between the two charts was
pretty much equivalent (it’s difficult to align the two charts exactly).
The thing I find suspect is AB’s timestamps. All
my AB timestamps are a multiple of 5 seconds where Investor/RT’s timestamps
look random. I know that if AB’s CPU usage is high it will throttle
down chart refresh intervals, but that should not be the case here with such
low CPU utilization.
This leads to my questions:
- Does anyone know why AB would
produce less tick bars than Investor/RT using the same data feed?
- Does anyone know why AB would
be aggregating ticks to a time slice that is a multiple of five? Are
there settings to control this?
- Are AB tick charts and tick
data collection dependant on the chart refresh interval? In other
words would a 10t chart with a chart refresh interval of 0 be different
than a 10t chart with a refresh interval of 3 seconds?
For reference, below are snapshots of the 10t charts, the
spreadsheet I created, and my AB settings.
Thanks,
David
A snapshot of each chart is below:
Tick bar data:
|
AmiBroker - Mar 08 YM - 12/24/2007 10t
|
|
|
|
Time
|
Open
|
High
|
Low
|
Close
|
Volume
|
|
11:00:35
|
13619
|
13620
|
13619
|
13620
|
25
|
|
11:00:55
|
13619
|
13619
|
13616
|
13618
|
32
|
|
11:02:00
|
13618
|
13620
|
13617
|
13260
|
42
|
|
11:02:55
|
13619
|
13620
|
13619
|
13619
|
15
|
|
11:03:35
|
13619
|
13619
|
13617
|
13618
|
21
|
|
11:04:20
|
13617
|
13620
|
13617
|
13618
|
29
|
|
11:06:00
|
13617
|
13619
|
13617
|
13618
|
14
|
|
11:06:40
|
13618
|
13621
|
13618
|
13620
|
48
|
|
11:07:35
|
13621
|
13622
|
13620
|
13622
|
67
|
|
11:08:05
|
13621
|
13624
|
13621
|
13624
|
25
|
|
11:08:50
|
13622
|
13624
|
13621
|
13622
|
46
|
|
11:09:25
|
13622
|
13626
|
13622
|
13625
|
41
|
|
11:09:50
|
13624
|
13626
|
13623
|
13624
|
45
|
|
11:10:35
|
13622
|
13625
|
13622
|
13624
|
54
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Investor/RT - Mar 08 YM - 12/24/2007 10t
|
|
|
|
11:00:02
|
13618
|
13620
|
13618
|
13620
|
13
|
|
11:00:42
|
13620
|
13620
|
13617
|
13617
|
24
|
|
11:00:49
|
13617
|
13618
|
13616
|
13617
|
14
|
|
11:01:20
|
13617
|
13617
|
13617
|
13617
|
29
|
|
11:02:02
|
13618
|
13620
|
13618
|
13619
|
24
|
|
11:02:57
|
13619
|
13620
|
13619
|
13619
|
12
|
|
11:03:26
|
13619
|
13619
|
13618
|
13618
|
15
|
|
11:03:55
|
13617
|
13618
|
13617
|
13618
|
23
|
|
11:04:30
|
13619
|
13620
|
13617
|
13619
|
18
|
|
11:06:23
|
13619
|
13621
|
13619
|
13620
|
12
|
|
11:06:33
|
13619
|
13621
|
13620
|
13621
|
30
|
|
11:06:59
|
13620
|
13621
|
13620
|
13621
|
51
|
|
11:07:33
|
13620
|
13622
|
13620
|
13622
|
30
|
|
11:07:56
|
13622
|
13623
|
13621
|
13622
|
20
|
|
11:08:12
|
13623
|
13624
|
13623
|
13623
|
19
|
|
11:08:50
|
13624
|
13624
|
13621
|
13622
|
27
|
|
11:09:19
|
13622
|
13626
|
13622
|
13626
|
19
|
|
11:09:32
|
13626
|
13626
|
13624
|
13625
|
28
|
|
11:09:51
|
13625
|
13625
|
13623
|
13623
|
16
|
|
11:10:20
|
13623
|
13625
|
13622
|
13625
|
42
|
|
11:10:33
|
13623
|
13625
|
13623
|
13625
|
22
|
|
11:10:50
|
13625
|
13625
|
13624
|
13626
|
23
|
Here are my AB settings for the test:
