bullish divergence in Accumulation/Distribution

arepo

sr. member

Activity: 448

Merit: 250

this statement is false

Quote from: bitcoinBull on April 15, 2012, 11:48:35 AM

This is exactly what happened in the correction from $7.2. That's why OBV is near the low at $2, but price is not.

In order for price to recede from $5 without the OBV plummeting to a new all-time low, bitcoin would have to trade downwards on very low volume.

so the obv agrees with your present bullish attitude, and that is why you like it? why should $5 "be the new $2"? my argument demonstrates that the simplicity of the formula glosses over information in the price which the ADL incorporates. this means that the obv is more susceptible to the danger of propagation of errors than the ADL.

further, even if $5 IS the new $2 (whatever that means), just because the obv has done an okay job SO FAR says nothing about it, i.e. the problem of induction. this is why i was addressing the formulae in general. do you have any responses to that effect?

bitcoinBull

legendary

Activity: 826

Merit: 1001

rippleFanatic

Quote from: arepo on April 15, 2012, 01:52:46 AM

obv sucks. it just adds/subtracts volume with no relation to the price spread for a given period.

I like OBV, it seems as good as any other indicator and has tracked bitcoin price pretty well in the daily chart above. Look at the all-time highs ($32 and $7) and the all-time lows ($2 and $4) - they are almost perfectly aligned at 2000 and 400, respectively. OBV isn't an oscillator, so its either been a fairly accurate cumulative measure of bitcoin money-flow, or that alignment was pure coincidence.

Quote from: arepo on April 15, 2012, 01:52:46 AM

what would happen if, for a number of periods, the median price is only slightly lower than the median of the last period, but during which a large volume of BTC is traded? during periods of high volatility, this can easily happen. this large volume is assigned a negative value due to the slight drop in price, so the obv plummets even though the information in the price spread might contrarily indicate strong buying pressure.

This is exactly what happened in the correction from $7.2. That's why OBV is near the low at $2, but price is not.

In order for price to recede from $5 without the OBV plummeting to a new all-time low, bitcoin would have to trade downwards on very low volume.

arepo

sr. member

Activity: 448

Merit: 250

this statement is false

Quote from: bitcoinBull on April 14, 2012, 01:26:41 PM

Let's throw OBV in the mix.

obv sucks. it just adds/subtracts volume with no relation to the price spread for a given period. noting the difference between this and the ADL, it's easy to see how the problem of propagation of errors can lead to a distorted shape in the obv. let me elucidate: what would happen if, for a number of periods, the median price is only slightly lower than the median of the last period, but during which a large volume of BTC is traded? during periods of high volatility, this can easily happen. this large volume is assigned a negative value due to the slight drop in price, so the obv plummets even though the information in the price spread might contrarily indicate strong buying pressure. this alone accounts for the differences between the shape of the obv and the adl in the above post.

bitcoinBull

legendary

Activity: 826

Merit: 1001

rippleFanatic

Quote from: S3052 on April 13, 2012, 01:11:48 PM

That is all fine, but to trigger the rally, there needs to be strong, and sustained buying volume

Ain't that a truism!

Technical charting works great on price. But when price is sideways, how can we quantitatively analyze volume?

Let's throw OBV in the mix.

S3052

legendary

Activity: 2100

Merit: 1000

That is all fine, but to trigger the rally, there needs to be strong, and sustained buying volume

bitcoinBull

legendary

Activity: 826

Merit: 1001

rippleFanatic

last one for now.

open to interpretation, but I see signs of an upward trend..

bitcoinBull

legendary

Activity: 826

Merit: 1001

rippleFanatic

Quote from: arepo on April 13, 2012, 01:17:05 AM

Quote from: ?? on ??

Acc/Dist = ((Close – Low) – (High – Close)) / (High – Low) * Period's volume

dont forget this

yeah i notice that in practice, the ADL responds to the shape of the candles in a very predictable way. its slope is generally proportional to the distance from the median to the bounds in that when the upper bound is far away from the median (top wick is tall) it increases rapidly, and does the opposite when a 'bearish hammer' (lower wick is long) forms.

even though it is an abstraction from the price, however, it does seem to do a good job detecting trends in the noise. it's actually pretty intuitive if you understand the formula that myself quoted.

Thanks! I see what you're talking about now.. so ADL has been rising for the past weeks because we've been seeing a lot of longer-tailed hammer-style candles (like today's candle) - from sells dropped onto bids that refill.

Sells have slowed to a drip whilst the bids keep filling up. I have a hunch that sellers are tapped (could be wrong though).

arepo

sr. member

Activity: 448

Merit: 250

this statement is false

Quote from: ?? on ??

Acc/Dist = ((Close – Low) – (High – Close)) / (High – Low) * Period's volume

dont forget this

yeah i notice that in practice, the ADL responds to the shape of the candles in a very predictable way. its slope is generally proportional to the distance from the median to the bounds in that when the upper bound is far away from the median (trading happened closer to the lower bound, top wick is tall) it decreases rapidly, and does the opposite when a "hanging man" (trading happened closer to the upper bound, bottom wick is long) forms.

even though it is an abstraction from the price, however, it does seem to do a good job detecting trends in the noise. it's actually pretty intuitive if you understand the formula that myself quoted.

EDIT because i accidentally reversed the relationship and for further explanation.

adamstgBit

legendary

Activity: 1904

Merit: 1037