Has anyone had any experience with this? It looks remarkably powerful. Does it work as stated? The demo seems to have glitches. http://www.interactivebrokers.com/en/software/tws/twsguide.htm here is an article by Thomas Peterffy on it: http://www.interactivebrokers.com/en/p.php?f=landing_algos&ib_entity=llc Many thanks for any insight.
I've used it extensively and yes, it is quite powerful. Anything in particular you are looking to do?
The A/D algo is good but it used to be better. You can simultaneously run multiple instances of it on the same or multiple tradables and you can create and name templates so you won't have to reconstruct everything every day. However, a little while back they made some changes to it and now the templates inexplicably no longer save all of the fields. For example, you used to be able to save the start time and end time fields in your templates. Now your algos' start times and end times will be lost when you save the template and they will have to be entered manually on each one each day. I have a feeling this was an oversight but it is very difficult to get the attention of anyone who can look into this at IB. (Hey def, can YOU get this fixed?) Anyway, to answer your question, notwithstanding any glitches that may be in the demo, the A/D algo does work as stated and it's likely to be worth your time to try it.
Many thanks for the insights. It seems that I cannot choose this in the top part of the algorithm choices: "take entire offer(up to number of shares remaining) if the offer size is..." if I specify anything in the position condition below it. The link below is different than the current A/D algo but the best I could find. http://www.interactivebrokers.com/e...fnotes/PDF-Accum-Distribute.php?ib_entity=llc From the Peterffy article: "Keep in mind that the Accumulate/Distribute algorithm is designed to be able to run unattended, as one trader may be running several hundred instances (i.e., versions) of it at the same time working on various instruments or even working on the same instrument." Any comments on this? I am wondering what sorts of moving average conditions others have imposed and found most useful? Has anyone used this algo to accumulate cheap options?
I'd like to use it to quickly accumulate larger positions for me on maybe 5 or 10 strong stocks on a stock market up day. It would be much easier to automate such a task than manually trying to fill all of those orders. I could have a sell algo running at the same time and it would serve as my stop on the positions as they accumulate. Is this practical and doable?
Reviving an old thread but it's sad that this great addition is still very limited 3 years later. For example adding a condition that it would buy only when the price is lower than yesterday's close or today's open is still missing despite being very useful. And it should be able to be API compatible but is only partially so. Simple indicator based conditions would also be useful, there's moving averages but no RSI? why? Seems very simple and practical to have an order that only takes trades if the RSI is lower than X instead of a certain percentage.
Potential downside of Accumulate/Distribute Algo: Using this Algo exposes your future trade intentions. I would feel very uncomfortable moving 300-600 contracts this way, even though that is when it would be really useful. In general, I like and respect Interactive Brokers. But they do have a hugely profitable market making operation. If I am a market maker, I would find it irresistable to use the advanced info in Accmulate/Distribute Algo of all the IB clients. I wonder if I am being too paranoid here. Any thoughts and comments? Personally, I would love to see this function in a piece of desktop software that only tells my broker the current lot that I am trading, but do not tell anybody anything about my next batch of trades in advance. Does such software exist?
From my own experience, the fills have been decent and I haven't felt gamed yet. There is always potential for conflict, even when talking to CS reps who can see my trades. For your purposes, it would make more sense to just use the API and send each order at a time, therefore only you know your future plans.
Pondering on how to implement a basic stop order with A/D, seems impossible at this point for a couple of reasons: 1. It offers only the opportunity for the order to execute in set price ranges while with a regular stop order, it should execute no matter what the price is after triggering. 2. The options are "Give up permanently..." or "Resume order if all conditions become true again", where's "Keep order active if conditions are true only once". 3. This could be achieved if an A/D order could be attached to a regular STP order which isn't possible right now, I'm not sure why this is the case. Asking def or anyone else for suggestions on how to achieve this fairly basic order.