Asset Class: US Stocks and US Equity Options

Q1. What is the largest number of Days for which Historical Tick Data that can be requested during US trading hours?

Q2. Lets say Answer to Q1 = 10 days of Tick/Symbol - Can I request a 10 days from 30 days ago (which is part of the 180 Days Tick Historical that is stored)

---

AK786

I believe that during trading hours IQFeed only provides tick data for the past 8 calendar days. In other words, even if you request just a single minute of tick data that is older than 8 days ago, IQFeed will return a !NO_DATA! response.

At the moment, the generic NO_DATA message is quite confusing, and prevents users from taking appropriate action depending on the actual error cause. I would like to suggest that in such a case, IQFeed will return a more informative error message, that will inform the users for the reason of NO_DATA, i.e. is there really no data for this request, or perhaps no data because of trading-hours limitations, or perhaps due to the 180-day limitation.

---

- Yair
IQML - IQFeed-Matlab connector
https://UndocumentedMatlab.com/IQML

Good morning, I just wanted to confirm the answer above is correct. As to the other, I will certainly let the server team know and place an enhancement request with management.

Tim

May I ask why this limitation exists? Also what defines trading hours? I'm working with futures and they trade around the clock.

Also is this noted in the documentation somewhere? I don't recall seeing it mentioned.

Hello, This is documented in our API FAQ document and I believe in other places as well, but I would have to look around to see what the exact links would be. We limit things just during the main day trading session, 9:30 AM to 16:30 PM. It is done simply to ensure we keep a high performance level during the highest user loads. Outside of these hours, the number of active users drops off considerably which allows us to perform the longer requests with no impact to others.

Tim

That's going to make backtesting/simulation difficult.

Understood. Generally I find that just storing those larger requests is the only way to go. The old data hardly ever changes and having to wait for 6 months of @ES tick data to download to see any results can be frustrating, so I would certainly recommend using some local storage options.

We are always looking at increasing our offering though and we revisit this subject from time to time, increasing it as new hardware and technology changes allow, so if our engineers and management determine we can start doing this in the future we will certainly do so.

Our apologies for the unexpected inconvenience.

Tim

Yeah I am storing it locally. I'll just have to do bulk downloads off hours.

One question though. Is there a way to change the timeout value for history requests? I cannot pull more than about three days worth of tick data before the connection closes. I suspect its timing related.

There is nothing there on our end to stop a download from going for however long it needs. Are you getting a socket error or anything when it closes?

Tim

Yes I believe it was a socket error, but I'll run a quick test to verify.

You could also turn on the lookup logging in the diagnostics app (via start menu) and you should be able to confirm what we sent to your socket vs what you received.

Also, if you can send the request to me that you tested with, I should be able to confirm if the connection to the server completed successfully.

Tim

@bigjake - if you are writing code in C#, try using Mathieu P C# Service.
https://github.com/mathpaquette/IQFeed.CSharpApiClient

To Avoid Time Out,

1. Alternatively, request 1 day of tick data and loop -

2. Alternatively, have upto 15 requests simulataneously to request Tick Data

IMO IQfeed is a great product, i had to figure things out through trial and error.

---

AK786

I've developed my own C# iqfeed client that I am considering productizing.