Hi Jim,
I'm seeing an anomaly in daily and monthly OBV as reported on their Java charts.
I personally don't see a problem, but I understand what you are getting at.
First of all OBV is an ancient indicator designed for end of day charts and great for indices.
Second, the daily value is dependent on the timescale used as the starting point, but the daily difference as in the calc you did and I did below, should be the same. And they are. Its just the volume for yesterday.
Third, because it was designed as end of day indicator, its usefullness as a intraday indicator is suspect. I only fire up the java daily real time chart if I day trade which is rarely. I normally position trade, but when the markets are super volatile, then even swing stocks 'compresses time' for price movement, then I am forced to become the 'day' trader. In this case, I use obv intra day only to confirm direction of prices and whether advances or declines are going to stick or reverse. Its not an indicator I would base buy/sell decision on on a daily chart.
Tracking today (17th Jan) OBV on the real-time Java chart, OBV at the last trade was -10,325 ie., 10,325 more sold, than bought.
I assum this number is end of day and guess you were expecting +82415. What time scale were you on when scooping the data? The answer lies in the OBV rules.
At the beginning of the day, obv starts at 0. Stockwatch does follow the OBV rules correctly of adding or subtracting volume correctly. It'll take paragraphs to regurgitating the formula. The anomally appears in the following case:
assum a 5 min bar
start at obv=0
first 5 min vol=2,000 shares, tick up or down doesn't matter
end of 5 min obv=0
second 5 min interval vol=3,000 tick up
end of 10 min obv = 3000
lost the first 2k volume
third 5 min interval vol=5000 tick even
end of 15 min obv = 3000
lost another 5000 shares because it was neither positive nor negative from the last period.
fourth 5 min interval vol = 1000 tick down
end of 20 min obv = 2000, volume traded totalled=11,000
Final anser is: Yes, when tick even, current volume for the period is tossed.
1 - OBV for the month is UP strongly - indicating accumulation (ie., more shares bought than sold) - OBV as of 2 Jan is +268,386 (!) and as of 17 Jan is +464,084.
NICE isn't it. Check this out, do the same calc but start from Nov 18, the day before the recent low and trend change because of the FDA plant inspection schedule rumour, till yesterday!
2 - OBV difference between 16 and 17 Jan on the monthly Java chart shows [(464,084) - (381,669)] = +82,415
Jan 17 obv=500,275
Jan 16 obv=417,860
difference = +82415
Yesterday volume = 82415
same value as you posted
Furthermore, the monthly OBV strongly suggests accumulation. This is (duh!) the opposite of what the Accumulation/Distribution lines are saying.
Something is wrong, here. Does anyone have any ideas?
Observing a divergence between the indicators. Nothing wrong.
True, OBV is suggesting accumulation but more accurately it is showing the price advance is tracking volume. Therefore, very healthy chart.
The A/D line created by Chailkin is similar to OBV but with a subtle difference, it is formulated not to accumulate it values from the last period's close like OBV does. So its down side is it misses the gap days, but OBV accounts for gap days. Personally I like A/D both for intra day as well as historical. With out going into exact formula, A/D measures the strength/weakness of only the period (whether day on the historical, or bars on the daily) with a value ranging from +1 to -1.
It is tracking the activities of the 3 banitos well.
Earlier when the 3 Banditos were indiscrimately selling, A/D tracked them to the T. Now that they have moderated and the draw maybe relative smaller (than the 4mm draw), they can 'callibrate' their hedging activities. Again A/D is tracking this through the relatively stronger closes on the day.
Its interesting to note that AD recently track sentiment/psychology well. Notice that the A/D was declining as people expected a large draw which in fact put heavy daily pressure when the last draw began early Dec. The it reversed rising from its mid Dec bottom as the last draw was coming to an end, expected 1st week of Jan for $3mm. Then Monday Jan 7 AD began its decline because the expected draw did not materialize, ie doubt, indecision and second guessing for various reason posted on the BBs. The guesses changed to a $3.5mm to $4.0mm to $4.5mm draw. Since the $4mm draw announcement, AD is rising again.
This next draw in progress will be an annoyance. A daily speed bump along the way up.
Since Sept 11 draws began. FA wise, we are progressing, FDA around the bend, much more $$ cash in the bank. Larger trades, real time bids/offers are flashing (only the pros can do this) so a few big boys have DMX on radar. Hoping for more, especially the funds. Where's Graham?
Next is 2Q financials either end of today or early next week. Monday?
The clock is ticking, we are past mid way in 3Q. "We want more news, we want more news,..." Wolf K.
Its been fun to talk TA with ya Jim. Broad but precision type topic, hope I'm clear. No time to proof read and edit.
Have a great day!
with great comfort and great confidence...FDA on the way! |
