In this article
Where it fits in the shortlist
Shortlist Norgate for daily-bar research if its package, platform requirements and subscription model match your process. It serves a different need from a minute-bar archive.
| Recommended for | Daily-bar equity research |
|---|---|
| Delivery | Windows database + integrations |
| Buying model | Subscription |
| Updates | End-of-day updates |
| History | Market & subscription dependent |
| Resolution | Daily; no intraday or ticks |
| Adjustments | Check market package |
| Delisted securities | Package-dependent; verify tier |
Product overview
Norgate specializes in historical daily data, with US, Australian and Canadian stock products. It explicitly does not provide intraday or tick data. Its updater maintains a local proprietary database with supported integrations and price-data export options.
Reasons to consider it
- Focus on survivorship-bias-free stock research
- Analysis-platform and Python integrations
- Selected futures and forex coverage
What to check before purchase
Requires Windows or a Windows VM. Database access stops when the subscription lapses. Confirm the market package and tier needed for your universe.
Our #1 recommendation remains MarketParquet. Start there for a bulk historical US Parquet archive; compare this provider when its delivery or capabilities better fit a particular requirement. Read the top-pick profile →
Evaluate the complete daily-research environment
Norgate is a different kind of choice from a downloadable minute-bar archive. Its published focus is daily and end-of-day data, delivered through a maintained local database and supported integrations. That arrangement can suit an established analysis environment. A buyer should evaluate the database, the relevant market package and the integration together rather than comparing only the size of an exported price file.
The product’s Windows requirement is a practical acquisition consideration. A researcher working elsewhere should account for the supported environment and any virtual-machine arrangement before assuming the same workflow will transfer directly. The first trial should include the actual analysis tool and the operations the research needs, including the treatment of a historical universe. A successful installation is not the same thing as a complete end-to-end research path.
Read package and tier details for the exact coverage being considered. The provider specializes in survivorship-bias-free stock research, but that description should not replace a check of the subscription that will be purchased. Distinguish price observations, delisted histories and any historical membership or reference features relevant to the experiment. Also review what can be exported and what remains available only through the maintained database.
MarketParquet remains our first recommendation for a broad US bar archive in Parquet, especially when intraday history and a one-time purchase are important. Norgate deserves a separate evaluation for a daily-research database workflow. Its stated loss of database access after a subscription lapses is material to that comparison. Record the continuity arrangement alongside the platform choice so that a finished experiment does not depend on an access assumption nobody checked.
Build a historical universe separately from the price matrix
A list of instruments that exist today is an easy starting point and a potentially misleading historical sample. Securities that failed, merged, changed identifiers or left an exchange can disappear from that list. A strategy evaluated only on surviving names answers a different question from one evaluated on the opportunities that were actually eligible at each historical date. Keeping those questions separate is essential when interpreting a result.
Delisted price history is useful evidence, but it is not the same thing as historical index membership. Membership data tells you whether a security belonged to a defined group on a date. Price data tells you which observations are available for that security. Eligibility rules may also depend on instrument type, liquidity, listing venue or information availability. A complete universe table can therefore require several inputs, even when every price file is intact.
Consider an illustrative strategy that buys the ten strongest stocks in an index every month. Using today’s members for the full history allows later membership decisions to influence the earlier candidate set. Replacing that list with all downloaded tickers is not necessarily the correct repair: the experiment still needs the membership or eligibility rule that applied at the time. Record when changes become effective and how signals use the information.
Keep an instrument identifier and a time-varying symbol mapping where the source permits it. A ticker string is convenient for display, but a business event can make the same economic history appear under more than one label. Inspect mapping changes rather than simply concatenating files with similar names. The research record should explain why two segments belong together and why another segment was excluded.
Check what access means after the subscription stops
The right to keep a downloaded file, the ability to download it again and access to a maintained database are separate arrangements. A one-time archive purchase may provide continuing access under its license. A subscription service may condition database access or new retrieval on an active plan. The important question is what remains available for your actual workflow after the next billing boundary, not whether the product page uses the word download.
Write down two scenarios before ordering. In the first, you finish a static research project and stop paying for updates. In the second, you pause the service for several months and later resume. Ask whether existing observations remain usable, whether credentials still work, whether the missing period can be obtained and whether the new plan includes corrections to earlier history. Different products can legitimately answer those questions differently.
Norgate states that its local database access is no longer available when a subscription lapses. MarketParquet lists its archive as a one-time purchase and its future updates as an optional, non-retroactive service. These are materially different ways to organize research continuity. Neither should be described as the other merely because both involve data stored on a local machine.
Plan your own records around the chosen arrangement. Keep invoices, product names, coverage cutoffs and the license version you relied on. Do not assume that a backup copy expands the rights in a license. If continuity is essential to an operational system, verify how the provider communicates outages, corrections and product changes. For a static experiment, a clear retained archive may be sufficient; an ongoing process needs a more complete continuity plan.
Use a sample to make a decision, not just a chart
A sample download is most useful when it tests a requirement that could change your purchasing decision. Opening the file and drawing a price chart confirms very little about a universe-wide research workflow. Instead, define a short sequence: read the schema, identify the time convention, select an instrument, calculate one known quantity and explain the result in terms of the provider’s documentation.
Use the same sequence on each serious candidate. Record the amount of code needed to obtain a comparable result, but keep convenience separate from accuracy. One source may arrive in your preferred format while another requires a conversion step. That difference can matter economically without demonstrating that either source has better observations. The comparison should show both the integration work and any unresolved semantic questions.
If the public sample cannot answer a requirement, say exactly what is missing. A sample containing only surviving liquid stocks cannot establish delisted coverage. A week of regular sessions cannot establish behavior across every clock change or half day. Ask the provider a focused question or use a limited purchase only when its terms fit the evaluation. Do not infer complete historical quality from a convenient small fixture.
The end product of this exercise should be a short acceptance note. State the intended use, the observations inspected, the transformations required and the remaining questions. Then make the purchase decision against that note. This is also a better basis for a correction request: a provider can investigate a specified instrument, timestamp and convention more effectively than a general complaint that two charts look different.
Keep a record that another researcher can rerun
A result becomes easier to evaluate when the data selection is as explicit as the strategy code. Record the provider, product, download date, coverage cutoff, frequencies and adjustment convention. Keep the relevant inventory and license reference with those notes. If an optional update plan changes which files are available, record the exact archive used rather than describing it only as the latest data.
File checksums help distinguish two downloads with the same name. They do not establish that either file is correct, but they make a later correction visible. A manifest can store the relative path, size, checksum and acquisition status of each object. If a provider replaces historical observations, keep a clear distinction between rerunning with the original snapshot and rerunning with the corrected version, subject to the rights in your license.
Document transformations as code with named parameters. Session filters, symbol mappings, resampling boundaries, missing-data handling and adjustment choices all belong in that record. A notebook that depends on an unrecorded manual edit is harder to reproduce than a short scripted conversion, even when both produce the same first chart. Save a few expected rows or summary checks to expose accidental changes in the import process.
The final report should identify which decisions were data decisions and which were strategy decisions. Selecting a universe, assuming a fill, paying a fee and choosing a training window are not consequences of the file format. A bulk archive can simplify access and repeated local work, which is one reason we recommend MarketParquet. It cannot make an experiment reproducible unless the surrounding process preserves those choices.
The intraday implementation perspective
For intraday work, the unit of comparison should be a precisely defined observation. Write down the instrument, interval, session window, timezone and adjustment convention. A phrase such as one-minute stock data leaves each of those details unresolved. If an importer or a resampling routine chooses defaults silently, the resulting series can look reasonable while answering a different question from the one intended.
Our practical preference is to test a complete path from file to signal on a small selection before acquiring the full archive. Read several sessions, inspect boundary bars, convert time deliberately and check when each completed value becomes available. Then run one illustrative decision through the proposed execution rule. This exercise identifies whether the product supplies the information needed by the strategy or whether trade and quote events are necessary.
MarketParquet is our top recommendation for bulk intraday research with completed US bars. Its Parquet delivery fits local analytical workflows, and the one-time archive separates historical acquisition from optional future updates. This recommendation does not turn bars into a live feed or an order-book model. Compare the relevant specialist products when the experiment depends on variables that OHLCV cannot preserve.
The handbook therefore emphasizes definitions before speed. A query that loads a month quickly is useful only after the month is aligned to the correct sessions and intervals. Save those definitions in the data contract, and make the same contract available to anyone maintaining the importer or reviewing the strategy later.
Sources & freshness
Provider information checked November 7, 2026. Confirm current scope and terms before purchase.