Data Feeds/Streams: extended hours data is now multi-provider

src/content/data-feeds/llms-full.txt

@@ -7854,10 +7854,10 @@ Please note that while intelligent smoothing is designed to mitigate these issue
 Tokenized equity feeds combine multiple input sources:
 
 - **Regular hours data**: Multi-sourced from consolidated tape data with deep liquidity
-- **Extended hours data**: Pre-market and post-market venue data
+- **Extended hours data**: Multi-sourced from pre-market and post-market venue data
 - **Overnight data**: Data from overnight trading venues
 
-*Extended and overnight session data may be sourced from fewer providers than regular hours data, reflecting the reduced number of venues operating during these periods. See the [Limited Provider Coverage](#limited-provider-coverage-during-extended-and-overnight-sessions) risk section for details.*
+*Overnight session data may be sourced from fewer providers than regular and extended hours data, reflecting the reduced number of venues operating during this period. See the [Limited Provider Coverage](#limited-provider-coverage-during-overnight-sessions) risk section for details.*
 
 ## Key differences from standard feeds
 
@@ -7884,7 +7884,7 @@ The reliability and accuracy of tokenized equity feeds varies by trading session
 | Session   | Data Quality | Provider Coverage      | Update Frequency |
 | :-------- | :----------- | :--------------------- | :--------------- |
 | Regular   | Highest      | Multiple providers     | Highest          |
-| Extended  | Moderate     | Limited providers      | Moderate         |
+| Extended  | Moderate     | Multiple providers     | Moderate         |
 | Overnight | Lower        | Very limited           | Lower            |
 | Weekend   | N/A          | No traditional trading | Stale            |
 
@@ -7898,13 +7898,13 @@ Tokenized equity feeds for assets from different issuers may differ in terms of
 
 Tokenized equity feeds introduce specific risks related to the nature of traditional equity markets, multi-session data aggregation, smoothing algorithms, and data sourcing. Users must understand these factors and implement appropriate safeguards.
 
-### Limited provider coverage during extended and overnight sessions
+### Limited provider coverage during overnight sessions
 
-Extended and overnight session price data is sourced from a more limited number of data providers compared to regular hours, making these sessions less reliable than regular hours feeds, which are aggregated from a broader set of providers.
+Overnight session price data is sourced from a more limited number of data providers compared to regular and extended hours, making this session less reliable than regular and extended hours feeds, which are aggregated from a broader set of providers.
 
 If providers experience downtime, technical failures, or connectivity disruption, the feed may flatline or report stale values. Such issues may lead to mispricing, failed liquidations, and potential bad debt accumulation.
 
-**Overnight and extended sessions** carry additional nuance: liquidity is thinner and fewer venues contribute data. In those conditions the feed may continue to update while surfacing **atypical numeric values**—including, in edge cases, **a reported price of zero** or other answers that are not economically meaningful as a reference price for the token. That behavior is distinct from staleness (fresh timestamps can accompany unsafe numbers). **Treat unexpected zeros, abrupt discontinuities, or values outside plausible bands as invalid for automated risk decisions** unless you independently confirm them against issuer methodology, recent observations, or supplementary market data.
+**Overnight and extended sessions** carry additional nuance: liquidity is thinner than during regular hours, and overnight fewer venues contribute data. In those conditions the feed may continue to update while surfacing **atypical numeric values**—including, in edge cases, **a reported price of zero** or other answers that are not economically meaningful as a reference price for the token. That behavior is distinct from staleness (fresh timestamps can accompany unsafe numbers). **Treat unexpected zeros, abrupt discontinuities, or values outside plausible bands as invalid for automated risk decisions** unless you independently confirm them against issuer methodology, recent observations, or supplementary market data.
 
 #### Mitigation
 

src/content/data-feeds/tokenized-equity-feeds/index.mdx

@@ -100,10 +100,10 @@ Please note that while intelligent smoothing is designed to mitigate these issue
 Tokenized equity feeds combine multiple input sources:
 
 - **Regular hours data**: Multi-sourced from consolidated tape data with deep liquidity
-- **Extended hours data**: Pre-market and post-market venue data
+- **Extended hours data**: Multi-sourced from pre-market and post-market venue data
 - **Overnight data**: Data from overnight trading venues
 
-_Extended and overnight session data may be sourced from fewer providers than regular hours data, reflecting the reduced number of venues operating during these periods. See the [Limited Provider Coverage](#limited-provider-coverage-during-extended-and-overnight-sessions) risk section for details._
+_Overnight session data may be sourced from fewer providers than regular and extended hours data, reflecting the reduced number of venues operating during this period. See the [Limited Provider Coverage](#limited-provider-coverage-during-overnight-sessions) risk section for details._
 
 ## Key differences from standard feeds
 
@@ -130,7 +130,7 @@ The reliability and accuracy of tokenized equity feeds varies by trading session
 | Session   | Data Quality | Provider Coverage      | Update Frequency |
 | :-------- | :----------- | :--------------------- | :--------------- |
 | Regular   | Highest      | Multiple providers     | Highest          |
-| Extended  | Moderate     | Limited providers      | Moderate         |
+| Extended  | Moderate     | Multiple providers     | Moderate         |
 | Overnight | Lower        | Very limited           | Lower            |
 | Weekend   | N/A          | No traditional trading | Stale            |
 
@@ -144,13 +144,13 @@ Tokenized equity feeds for assets from different issuers may differ in terms of
 
 Tokenized equity feeds introduce specific risks related to the nature of traditional equity markets, multi-session data aggregation, smoothing algorithms, and data sourcing. Users must understand these factors and implement appropriate safeguards.
 
-### Limited provider coverage during extended and overnight sessions
+### Limited provider coverage during overnight sessions
 
-Extended and overnight session price data is sourced from a more limited number of data providers compared to regular hours, making these sessions less reliable than regular hours feeds, which are aggregated from a broader set of providers.
+Overnight session price data is sourced from a more limited number of data providers compared to regular and extended hours, making this session less reliable than regular and extended hours feeds, which are aggregated from a broader set of providers.
 
 If providers experience downtime, technical failures, or connectivity disruption, the feed may flatline or report stale values. Such issues may lead to mispricing, failed liquidations, and potential bad debt accumulation.
 
-**Overnight and extended sessions** carry additional nuance: liquidity is thinner and fewer venues contribute data. In those conditions the feed may continue to update while surfacing **atypical numeric values**—including, in edge cases, **a reported price of zero** or other answers that are not economically meaningful as a reference price for the token. That behavior is distinct from staleness (fresh timestamps can accompany unsafe numbers). **Treat unexpected zeros, abrupt discontinuities, or values outside plausible bands as invalid for automated risk decisions** unless you independently confirm them against issuer methodology, recent observations, or supplementary market data.
+**Overnight and extended sessions** carry additional nuance: liquidity is thinner than during regular hours, and overnight fewer venues contribute data. In those conditions the feed may continue to update while surfacing **atypical numeric values**—including, in edge cases, **a reported price of zero** or other answers that are not economically meaningful as a reference price for the token. That behavior is distinct from staleness (fresh timestamps can accompany unsafe numbers). **Treat unexpected zeros, abrupt discontinuities, or values outside plausible bands as invalid for automated risk decisions** unless you independently confirm them against issuer methodology, recent observations, or supplementary market data.
 
 #### Mitigation
 

src/content/data-streams/llms-full.txt

@@ -7150,7 +7150,7 @@ Once the overnight venue starts publishing its own quotes, the data will update
 
 Users relying on bid/ask data during early overnight hours should be aware of this carryover behavior and implement appropriate safeguards, such as checking `lastSeenTimestampNs` for staleness or widening risk parameters during overnight session transitions.
 
-Separately, the overnight stream can occasionally surface **atypical `mid` values**—including **zero**—while timestamps still update. Treat unexpected zeros or discontinuities as unsafe inputs unless confirmed against your reference logic; see [Single Provider for Extended & Overnight Data](#single-provider-for-extended--overnight-data).
+Separately, the overnight stream can occasionally surface **atypical `mid` values**—including **zero**—while timestamps still update. Treat unexpected zeros or discontinuities as unsafe inputs unless confirmed against your reference logic; see [Single Provider for Overnight Data](#single-provider-for-overnight-data).
 
 ##### Mitigation
 
@@ -7327,11 +7327,11 @@ While 24/5 streams provide extended market coverage, they introduce specific ris
 
 In the following sections, we outline key risks and recommended mitigation strategies.
 
-### Single Provider for Extended & Overnight Data
+### Single Provider for Overnight Data
 
-Extended and overnight session price feeds are currently sourced from a single data provider, making these sessions less reliable than the regular hours price feed which is multi-sourced.
+Overnight session price feeds are currently sourced from a single data provider, making this session less reliable than the regular and extended hours price feeds, which are multi-sourced.
 
-Unlike regular hours where multiple independent sources cross-validate each other, a single provider means there is no redundant check on individual quote quality. The following risks are closely related:
+Unlike regular and extended hours, where multiple independent sources cross-validate each other, a single provider means there is no redundant check on individual quote quality. The following risks are closely related:
 
 - **Provider outage or disruption:** If the provider experiences downtime, technical failures, or connectivity disruption, the feed may flatline or report highly inaccurate figures, which can prevent users from reacting to real price movements. Such issues may lead to mispricing, failed liquidations, and potential bad debt accumulation.
 - **Erroneous individual quotes:** Even when the provider is operational, it may publish a single bid or ask quote that is significantly off from the prevailing market price — for example, a bid price printing far below the previous in-session price. Because there is no second source to cross-validate against, this outlier quote directly distorts the mid-price calculation, creating a large artificial price move.
@@ -7343,7 +7343,7 @@ Unlike regular hours where multiple independent sources cross-validate each othe
 Consumers should not rely on the staleness indicator alone. A feed can be actively updating with a technically fresh timestamp while still publishing an incorrect mid price derived from a single outlier quote. Recommended safeguards include:
 
 - **Monitor bid/ask spread:** A sudden, significant widening of the spread between `bid` and `ask` is a strong indicator that a single outlier quote has distorted the mid-price. Implement a maximum spread threshold — for example, if `(ask - bid) / mid` exceeds a percentage beyond a configured threshold, treat the price as unreliable and restrict or pause exposure accordingly.
-- **Implement value-level checks on `mid`:** Reject or treat as invalid **zero** and other out-of-range answers; compare against your previous accepted tick and configured bands. Apply these checks especially to **Overnight Hours** (and extended-hours) reports, where single-provider risk is highest.
+- **Implement value-level checks on `mid`:** Reject or treat as invalid **zero** and other out-of-range answers; compare against your previous accepted tick and configured bands. Apply these checks to all sessions, and especially to **Overnight Hours** reports, where single-provider risk applies.
 - **Implement price deviation checks:** Compare the current mid-price against a reference (e.g., the previous valid price or a short-term moving average) and trigger a circuit breaker if the move exceeds a configured threshold.
 - **Do not use data when `marketStatus` is `closed`:** When the market is closed, data providers may perform maintenance, which can result in stale or incorrect prices being published. Consuming or acting on feed data during a closed market status is not recommended — pause or restrict all activity until the market reopens and a valid `open` status is confirmed.
 - **Check staleness via `lastSeenTimestampNs`:** Detect when data stops updating entirely, which is a separate but complementary signal to the above.

src/content/data-streams/rwa-streams/24-5-us-equities-user-guide.mdx

@@ -179,7 +179,7 @@ Once the overnight venue starts publishing its own quotes, the data will update
 
 Users relying on bid/ask data during early overnight hours should be aware of this carryover behavior and implement appropriate safeguards, such as checking `lastSeenTimestampNs` for staleness or widening risk parameters during overnight session transitions.
 
-Separately, the overnight stream can occasionally surface **atypical `mid` values**—including **zero**—while timestamps still update. Treat unexpected zeros or discontinuities as unsafe inputs unless confirmed against your reference logic; see [Single Provider for Extended & Overnight Data](#single-provider-for-extended--overnight-data).
+Separately, the overnight stream can occasionally surface **atypical `mid` values**—including **zero**—while timestamps still update. Treat unexpected zeros or discontinuities as unsafe inputs unless confirmed against your reference logic; see [Single Provider for Overnight Data](#single-provider-for-overnight-data).
 
 ##### Mitigation
 
@@ -365,11 +365,11 @@ While 24/5 streams provide extended market coverage, they introduce specific ris
 
 In the following sections, we outline key risks and recommended mitigation strategies.
 
-### Single Provider for Extended & Overnight Data
+### Single Provider for Overnight Data
 
-Extended and overnight session price feeds are currently sourced from a single data provider, making these sessions less reliable than the regular hours price feed which is multi-sourced.
+Overnight session price feeds are currently sourced from a single data provider, making this session less reliable than the regular and extended hours price feeds, which are multi-sourced.
 
-Unlike regular hours where multiple independent sources cross-validate each other, a single provider means there is no redundant check on individual quote quality. The following risks are closely related:
+Unlike regular and extended hours, where multiple independent sources cross-validate each other, a single provider means there is no redundant check on individual quote quality. The following risks are closely related:
 
 - **Provider outage or disruption:** If the provider experiences downtime, technical failures, or connectivity disruption, the feed may flatline or report highly inaccurate figures, which can prevent users from reacting to real price movements. Such issues may lead to mispricing, failed liquidations, and potential bad debt accumulation.
 - **Erroneous individual quotes:** Even when the provider is operational, it may publish a single bid or ask quote that is significantly off from the prevailing market price — for example, a bid price printing far below the previous in-session price. Because there is no second source to cross-validate against, this outlier quote directly distorts the mid-price calculation, creating a large artificial price move.
@@ -381,7 +381,7 @@ Unlike regular hours where multiple independent sources cross-validate each othe
 Consumers should not rely on the staleness indicator alone. A feed can be actively updating with a technically fresh timestamp while still publishing an incorrect mid price derived from a single outlier quote. Recommended safeguards include:
 
 - **Monitor bid/ask spread:** A sudden, significant widening of the spread between `bid` and `ask` is a strong indicator that a single outlier quote has distorted the mid-price. Implement a maximum spread threshold — for example, if `(ask - bid) / mid` exceeds a percentage beyond a configured threshold, treat the price as unreliable and restrict or pause exposure accordingly.
-- **Implement value-level checks on `mid`:** Reject or treat as invalid **zero** and other out-of-range answers; compare against your previous accepted tick and configured bands. Apply these checks especially to **Overnight Hours** (and extended-hours) reports, where single-provider risk is highest.
+- **Implement value-level checks on `mid`:** Reject or treat as invalid **zero** and other out-of-range answers; compare against your previous accepted tick and configured bands. Apply these checks to all sessions, and especially to **Overnight Hours** reports, where single-provider risk applies.
 - **Implement price deviation checks:** Compare the current mid-price against a reference (e.g., the previous valid price or a short-term moving average) and trigger a circuit breaker if the move exceeds a configured threshold.
 - **Do not use data when `marketStatus` is `closed`:** When the market is closed, data providers may perform maintenance, which can result in stale or incorrect prices being published. Consuming or acting on feed data during a closed market status is not recommended — pause or restrict all activity until the market reopens and a valid `open` status is confirmed.
 - **Check staleness via `lastSeenTimestampNs`:** Detect when data stops updating entirely, which is a separate but complementary signal to the above.