Suspicious data
Suspicious data
Heck of a day in Cocoa today. The EOD data checker reported "Suspicious daily low" but an intraday chart suggests it might be real. Wowzer.
- Attachments
-
- intraday
- 5_minute_chart.png (16.69 KiB) Viewed 19809 times
-
- EOD
- sdl.png (7.22 KiB) Viewed 19809 times
-
- Roundtable Knight
- Posts: 123
- Joined: Tue Nov 03, 2009 8:49 am
- Location: London
- Contact:
Cococa was recently in the news with Anthony "Choc Finger" Ward buying a bulk load of cocoa beans...
http://www.telegraph.co.uk/foodanddrink ... trade.html
Could it be his footprints? (you'd think it would not try to offload it all or a big part inside a 5-min bar though)...
http://www.telegraph.co.uk/foodanddrink ... trade.html
Could it be his footprints? (you'd think it would not try to offload it all or a big part inside a 5-min bar though)...
-
- Roundtable Knight
- Posts: 137
- Joined: Thu Sep 23, 2004 11:19 am
- Location: California
-
- Roundtable Knight
- Posts: 229
- Joined: Thu Jul 08, 2010 2:36 pm
- Location: Boulder, CO
- Contact:
Cordura21
That's an interesting chart. Which service provided it - or do you roll your own?
I am curious as to the data used to construct the chart.
I assume the histograms represent the cumulative trades that executed at a given price divided into those that went at the bid and those at the ask. But what does the smaller "mid" category mean? Do you know the mechanics of a trade being counted as "mid" - does it simply mean that the bid = ask and thus the trade cleared?
Finally, how does the gray line related to the VWAP, how is it constructed? The chart makes it look as though there is some relationship between them - is it supposed to be a confidence interval around the estimated average price?
Thaks in advance!
That's an interesting chart. Which service provided it - or do you roll your own?
I am curious as to the data used to construct the chart.
I assume the histograms represent the cumulative trades that executed at a given price divided into those that went at the bid and those at the ask. But what does the smaller "mid" category mean? Do you know the mechanics of a trade being counted as "mid" - does it simply mean that the bid = ask and thus the trade cleared?
Finally, how does the gray line related to the VWAP, how is it constructed? The chart makes it look as though there is some relationship between them - is it supposed to be a confidence interval around the estimated average price?
Thaks in advance!
The chart is from Bloomberg. I am pretty sure that you can roll your own with something like eSignal, Tradestation, and I am 90% sure I saw it on the free Quotetracker, at least for stocks, and in Stockcharts.com.
The bars count the amount of contracts (y axis) traded at the days' prices (x axis). You can also see number of trades instead of number of contracts.
The mid price category means, according to Bloomberg, those trades executed in between the best bid and offer at the time. I told them that I thought you could only trade at the bid and the offer, they said "there are many was to trade the markets these days, sir. Is there anything else I can assist you with?" So I'll further investigate unless somebody in the forum knows better (mainly everybody).
The gray line shows the theoretical normal distribution of prices, given the vwap and the stdev. I wonder if the stdev is also volume weighted or not.
I think it looks like those "delta charts" turned sideways. I find it pretty neat. Do you think there's a way to get money out of these things?
Cheers, Cord
The bars count the amount of contracts (y axis) traded at the days' prices (x axis). You can also see number of trades instead of number of contracts.
The mid price category means, according to Bloomberg, those trades executed in between the best bid and offer at the time. I told them that I thought you could only trade at the bid and the offer, they said "there are many was to trade the markets these days, sir. Is there anything else I can assist you with?" So I'll further investigate unless somebody in the forum knows better (mainly everybody).
The gray line shows the theoretical normal distribution of prices, given the vwap and the stdev. I wonder if the stdev is also volume weighted or not.
I think it looks like those "delta charts" turned sideways. I find it pretty neat. Do you think there's a way to get money out of these things?
Cheers, Cord
-
- Roundtable Knight
- Posts: 229
- Joined: Thu Jul 08, 2010 2:36 pm
- Location: Boulder, CO
- Contact:
Thanks Cordura,
Typical delphic response from Bloomberg.
As for whether there is money to be made from Price-volume charts, I don't know except to say that IF the volume data contains information THEN one ought to be able to exploit it!
I remember a chap named Steve Woods who invented (or at least promoted "Float Turnover Analysis"). The concept was that you measure time for a given stock in terms of the share float and daily volume. Basically you create a price channel whose duration was such that the sum of the daily volume equalled the float. The longer the price action stayed within the price channel (i.e. the lower bound started to rise while the upper bound started to fall) the more significant a break out of the channel. He would use price-volume charts for the action within the channel to determine congestion areas and support / resistance. A quick Google turned up loads of lonks so I guess the system is still out there.
Typical delphic response from Bloomberg.
As for whether there is money to be made from Price-volume charts, I don't know except to say that IF the volume data contains information THEN one ought to be able to exploit it!
I remember a chap named Steve Woods who invented (or at least promoted "Float Turnover Analysis"). The concept was that you measure time for a given stock in terms of the share float and daily volume. Basically you create a price channel whose duration was such that the sum of the daily volume equalled the float. The longer the price action stayed within the price channel (i.e. the lower bound started to rise while the upper bound started to fall) the more significant a break out of the channel. He would use price-volume charts for the action within the channel to determine congestion areas and support / resistance. A quick Google turned up loads of lonks so I guess the system is still out there.
I don't know if you read this old stuff about Eckhardt:
http://www.elitetrader.com/vb/showthrea ... genumber=1
He said they only use price, they don't use summary stats like moving averages, and that things like breakout levels take into account just two pie es of data, ignoring the "structure" that lies in between them.
Maybe he uses the kind of ideas these graphics represent but in the trend following timeframe. I guess he just concentrates on price clustering instead of using volume?
A casual observation about price clustering and level "stickiness" seems reasonable. What do you think about it?
http://www.elitetrader.com/vb/showthrea ... genumber=1
He said they only use price, they don't use summary stats like moving averages, and that things like breakout levels take into account just two pie es of data, ignoring the "structure" that lies in between them.
Maybe he uses the kind of ideas these graphics represent but in the trend following timeframe. I guess he just concentrates on price clustering instead of using volume?
A casual observation about price clustering and level "stickiness" seems reasonable. What do you think about it?
the bloomberg person might have been refering to block trades.cordura21 wrote: I told them that I thought you could only trade at the bid and the offer, they said "there are many was to trade the markets these days, sir. Is there anything else I can assist you with?" So I'll further investigate unless somebody in the forum knows better (mainly everybody).
Cheers, Cord
for ICE
https://www.theice.com/publicdocs/futur ... de_FAQ.pdf
I don't think SGX Nifty closed at Zero
CSI is saying the Nifty stock index futures closed at zero on Wednesday. I think that's probably wrong.
- Attachments
-
- Nifty_SGX.png (30.78 KiB) Viewed 18794 times