Hi, the application I am writing is only really concerned with 15min, 30min, 60min bars. I do not really have a need for tick data. In fact, I only need the data to refresh every 15minutes.
I will be working with around 1000 symbols.

From looking at the API it seems to me that I have two options:
1) Start a level 1 stream for each symbols and collate the ticks into 15 minute bars
2) Fetch historical data every 15minutes

The issue with option (2) is that the API does not allow me to request only the latest n 15minute bars for a symbol; this means that by the end of the day I am potentially reloading 32 (8 hours * 4) 15minute bars, most of which I have already downloaded before.

Which of the above two strategies should I follow? Or any other strategy you can recommend?

Ilan

You will need to use Option 1. IQFeed does not allow users to cycle through pulling historical data on an automated basis. IQfeed is meant to be a streaming feed primarily. While IQFeed has significant historical data, it is there to fill in gaps or do end of day downloading/backfilling. During market hours, you need to set a watch on the symbols for which you need data, then build your bars intraday. If you need to refresh data on the stocks you watch, you can do it afterhours.

---

Jay Froscheiser
DTN - Trading Markets

What would be the typical daily bandwidth requirements for getting every tick for 1000 shares?

I watch ticks for the most active 980 or so Nasdaq symbols. Lately this has amounted to 3.6 GB daily expanded tick data (what your program has to process). The internet traffic is considerably less, as the data is transmitted compressed. On Monday I'll watch my connection manager to see what the compression factor is and report back.

JKD
Edited by kdresser on Jun 24, 2006 at 10:00 AM

The ratio of data out of the connection manager to the data coming in via the net is roughly 8:1 (8.3 today).

Thank you, this is very very helpful!