Why Comprehensive Historical Equity Data Requires Multiple Sources
Summary
The document explains why a retail trader seeking long-run, survivorship-bias-free US security data may need to combine several vendors. The requested package spans equities, ETFs, ADRs and indices, split- and dividend-adjusted prices, delisted securities, minute bars, fundamentals and an ongoing update feed. The response says these categories are commonly sold separately, with high-frequency data treated as a distinct product, so one provider may not cover the whole requirement.
It points to major commercial data vendors and suggests comparing available sources against the buyer’s cost and convenience needs. For a large-scale setup, it names fundamental and market-data providers as possible starting points and indicates that enterprise access can be very expensive. The answer offers no vendor-by-vendor feature comparison, coverage verification or practical implementation steps, and its quoted cost estimates are contextual to the original discussion rather than a current price guide. The main takeaway is to define the required coverage and budget, then expect to assemble a solution from separate datasets.
Key ideas
- Historical equity, ETF, fundamental and minute-bar data are often provided as separate products.
- Survivorship-bias-free history requires including securities that later became delisted.
- A single vendor may not satisfy broad coverage, adjustment, frequency and update requirements.
- Commercial data packages can carry substantial costs, especially for institutional-scale coverage.
Tags
Full text
# DATA for Backtest # DATA for Backtest I am a retail trader and i am looking for a data set that comprehends all us equities, ETF's, ADR's and indices going back at least to January 1st 2007. Data would need to be split + dividend adjusted and be survivorship bias free (i.e. include all the above mentioned asset clases that have been active at some point for trading during the time period between 01/01/2007 until present day and now are delisted). Another requirement would be 1 min bar frequency for the data set. In addition i am looking for fundamental data on the same asset clases and time period mentioned above. (i.e. market cap, P/E, EPS, etc. Finally a data feed to keep the both above mentioned types of data sets updated would also be of my interest. Could anyone help me to find good sources? i am willing to pay for the data. Thanks for your time and response! ## Answer by Chris (score 2) https://quant.stackexchange.com/a/45242 I don't know of any single dataset that encompasses everything you're looking for, even including the standard heavyweight commercial sources (which cost $250k+ per year). Equity data is typically distinct from ETF data, and fundamental data is typically separate still. High-frequency data is a different animal beyond those (eg, 1-min bars). It kind of sounds like you're looking for a data unicorn without really even knowing what you're looking for, practical costs aside. It's not uncommon for an individual to purchase a Factset or Bloomberg subscription for a personal book (which typically run ~$30k a year for the standard seats), but I don't get the impression that's the kind of cost you're prepared to pay if you're asking a general question for data sources here. With that, I'd suggest you take a look at the data thread here and figure out what is available. You're almost certainly going to need to piece it together, even using commercial sources, so your best bet will be to figure out your personal sweet spot for cost and convenience. I will add since I don't see it mentioned in the thread I linked to, Thomson Reuters, Bloomberg and Factset are among the standard commercial sources available. If I misunderstood your intentions and you're planning to go full-bore, Worldscope and either Bloomberg or Factset are a good access point for building something to scale. You're likely looking at +/-$300k for data per year going that direction though.
Shown in full with attribution under the source's licence. Licence: CC BY-SA 4.0 (Stack Exchange)
This summary was written by Stratmill's research agent from the original; it is not a copy of the source.