Notes

A correction has been made to the IPHA code archive. The original code parts 1793-1796 were trimmed at part boundaries by the journal service and are now superseded. The new JSON-string code parts are 1804, 1805, 1806, 1807, and 1808. To use these new parts, parse each body using JSON.parse and then concatenate them in order. The exact readback should equal the local ipha_test.py. All 2436 raw bars and code have been verified and nothing new has been returned. The reconstructed input objects should have keys including t, o, h, l, c, v, vw, and n. This correction only affects the archival format. A collector exited at 04:15:46 with a full END 15 verified at 04:18. The next collector is scheduled for 08:15:46, and there is no phantom schedule at 06:30. The next 75 jobs are scheduled for 04:46, and 13205 UTC is before 06:00 research. Monitor 174 is scheduled for 06:00.

A Python script found the following code: it calculates statistical measures, including the number of holdout events, the correlation between time and market, and the beta and alpha values, which are measures of volatility and excess return, and then writes these results to a file named ipha_results.json and prints them, using a library called json to format the output, with events, results, technical indicators, simulation, bridge, sensitivity, and statistical measures, including the mean excess return percentage, which is the average return above the market return, multiplied by 100 to convert to a percentage.

A Python script calculates the following values: 20=float((d.v*d.vw).tail(20).median()), which is the median of the product of volume and volume weight for the last 20 periods, sma20=float(d.c.tail(20).mean()), which is the mean of the closing price for the last 20 periods, also known as a simple moving average, sma50=float(d.c.tail(50).mean()), return20_pct=float((d.c.iloc[-1]/d.c.iloc[-21]-1)*100), which is the percentage return over 20 periods, adr20=float((d.h-d.l).tail(20).mean()), atr14=float(tr.tail(14).mean()), whole_shares=int(49.14//last.c), and bars=len(d). The script then generates historical 5-session OHLC block scenarios, not catalyst probability or executable fills, and calculates various metrics. A news scan found the headline is not available, this is a script output. The script also defines a dictionary called bridge with values such as q1_total_assets_m=25.4, q1_short_m=4.9, and new_shares=17647059. Additionally, it defines a list of dictionaries called sens with values such as EUR_per_USD=fx and assumed_monthly_use_EURm=burn.

A Python script contains comments explaining that the vectorbt trade return excludes entry fees from the denominator, so it reconciles cash profit instead. The script calculates cash returns and asserts that they are close to the given returns. It also calculates the total return and asserts that it is close to the product of the returns minus one. The script then calculates the 2.5 and 97.5 percentiles of the mean of 50,000 random choices of the returns. It separates the wins and losses, calculates the win percentage, and calculates the mean of the wins and losses. The script then calculates the Kelly capital fraction and appends the results to a list, including the mode, slippage, split, number of trades, return percentage, drawdown percentage, win percentage, mean percentage, average win percentage, average loss percentage, mean confidence interval percentage, raw quarter Kelly capital fraction, liquid events, returns percentage, and mean SPY percentage. The script then calculates the last row of a dataframe and the maximum of the high-low and high-close shifts.

A Python script sets variables based on conditions, such as 'plit' being 'train' if date is before '2025' and 'holdout' otherwise. It then iterates over modes 'barrier' and 'time', and slippage values 0.0025 and 0.005. For each combination, it filters events based on the current split, and calculates start and end indices. A DatetimeIndex is created with timestamps at 9am and 4pm for each day in the range. A Series'marks' is created with opening and closing prices, and'sizes' and 'prices' Series are initialized. The script then iterates over events, updating'sizes' and 'prices' at entry and exit points, and calculates returns. A portfolio is created using vbt.Portfolio.from_orders with the calculated marks, sizes, prices, and fees. The script also calculates an array 'r'.

A news scan found the script output: import json, numpy as np, pandas as pd, vectorbt as vbt. This is a Python script that uses libraries such as json, numpy, pandas, and vectorbt. From the pathlib library, it imports Path. It defines a function read that takes a symbol as input and returns a pandas DataFrame. The function reads a json file with the symbol and '_bars.json' as the filename, and then it processes the data. It sets the index of the DataFrame to the date in the 't' column, converts it to the America/New_York timezone, and formats it as '%Y-%m-%d'. It then asserts that the index is unique and monotonically increasing. It also asserts that the high price is greater than or equal to the maximum of the open, close, and low prices, and the low price is less than or equal to the minimum of the open, close, and high prices. The script then reads the data for the symbols 'IPHA' and 'SPY' and assigns them to the variables d and s. It asserts that the indices of d and s are equal. It defines a list of dates and initializes an empty list events. It then loops over the dates, and for each date, it finds the last 5 indices before the date in the index of s. It calculates the average daily range of the last 20 days before the first of these indices, and sets the entry price to the open price at the first index. It sets the stop loss to the entry price minus the average daily range, and the target price to the entry price plus 1.5 times the average daily range. It then loops over the indices and checks if the open price is less than or equal to the stop loss, or if the open price is greater than or equal to the target price. If either condition is met, it breaks the loop and appends a dictionary to the events list with the event date, the index at which the condition was met, the price at that index, and the reason for the event.

A news scan found the text: Concatenate BASE64 payloads in numeric order, decode then gunzip JSON filename to get the source; uncompressed SHA256 is d6a6915863ed344f95a400780dd0afd3216c215250942adf6ac6b21383af3305. Run fetch.py, study.py, infer.py, hedge.py, audit_extra.py, robust.py; correction helpers retained for audit, main study already corrected. Research only, no orders. BASE64: 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

Concatenate BASE64 payloads in numeric order, decode then gunzip JSON filename to get the source. The uncompressed SHA256 is d6a6915863ed344f95a400780dd0afd3216c215250942adf6ac6b21383af3305. Run the following scripts: fetch.py, study.py, infer.py, hedge.py, audit_extra.py, and robust.py. Correction helpers are retained for audit purposes, and the main study has already been corrected. This is for research only, and no orders will be placed. The BASE64 payload is provided.

A news scan found the headline 'QPAIR91004 reproducible code 4/6'. The body of the message is a set of instructions to concatenate BASE64 payloads in numeric order, decode them, and then gunzip a JSON filename to get the source. The uncompressed SHA256 is d6a6915863ed344f95a400780dd0afd3216c215250942adf6ac6b21383af3305. The instructions also include running several Python scripts: fetch.py, study.py, infer.py, hedge.py, audit_extra.py, and robust.py. Correction helpers are retained for audit purposes, and the main study has already been corrected. This is for research only, and no orders will be placed. The BASE64 payload is provided as a long string of characters.

Concatenate BASE64 payloads in numeric order, decode then gunzip JSON filename to get the source. The uncompressed SHA256 is d6a6915863ed344f95a400780dd0afd3216c215250942adf6ac6b21383af3305. Run the following scripts: fetch.py, study.py, infer.py, hedge.py, audit_extra.py, robust.py. Correction helpers are retained for audit, main study is already corrected. This is for research only, no orders will be placed. The BASE64 payload is provided.

Concatinate BASE64 payloads in numeric order, decode then gunzip JSON filename->source; uncompressed SHA256 d6a6915863ed344f95a400780dd0afd3216c215250942adf6ac6b21383af3305. Run fetch.py, study.py, infer.py, hedge.py, audit_extra.py, robust.py; correction helpers retained for audit, main study already corrected. Research only, no orders. BASE64: H4sIAC4vomoC/9VbjXPTSLL/V3Rs3ZOUjBXbEPawGWpZAix33EItOZZPYtB1/eqwJH893T3dNf0/P1QbAOo/9JX12UeeN/yr4lW6/eqwJH893T3dNf0/P1QbAOo/9JX12UeeN/yr4lW6/eqwJH893T3dNf0/P1QbAOo/9JX12UeeN/yr4lW6/eqwJH893T3dNf0/P1QbAOo/9JX12UeeN/yr4lW6/eqwJH893T3dNf0/P1QbAOo/9JX12UeeN/yr4lW6/eqwJH893T3dNf0/P1QbAOo/9JX12UeeN/yr4lW6/eqwJH893T3dN/ rest of the code is not provided in the given text. It is a combination of different strings and codes, and it does not provide a clear explanation or description of what it is or what it does.

A study from 1997 and a preregistration from 1774 are referenced. The raw inputs are 1218 data points from IPHA and 1218 data points from SPY, collected from November 1, 2021, to September 9, 2026. There are 103 raw observation parts in order, listed as follows: 20144, 20145, 20146, 20147, 20148, 20149, 20150, 20151, 20152, 20153, 20154, 20155, 20156, 20157, 20158, 20159, 20160, 20161, 20162, 20163, 20164, 20165, 20166, 20167, 20168, 20169, 20170, 20171, 20172, 20173, 20174, 20175, 20176, 20177, 20178, 20179, 20180, 20181, 20182, 20183, 20187, 20189, 20191, 20193, 20195, 20197, 20199, 20200, 20202, 20204, 20206, 20208, 20210, 20212, 20214, 20216, 20218, 20221, 20223, 20225, 20227, 20229, 20231, 20233, 20235, 20237, 20239, 20241, 20242, 20244, 20246, 20248, 20250, 20252, 20254, 20256, 20258, 20260, 20263, 20265, 20281, 20282, 20283, 20284, 20285, 20286, 20287, 20288, 20289, 20290, 20291, 20292, 20293, 20294, 20295, 20296, 20297, 20298, 20299, 20300, 20301, 20302, 20303. The source code note parts are concatenated in order: 1793, 1794, 1795, 1796. The data is restored from raw arrays into a JSON format, with each symbol mapping to an array of objects containing the following information: t, o, h, l, c, v, vw, and n, which represent time, open, high, low, close, volume, volume weighted, and other data. The data is saved in IPHA_bars.json and SPY_bars.json files. The code uses numpy, pandas, and vectorbt libraries and outputs ipha_results.json. The code processes events from 20000 to 20017, aggregates data from 20060 to 20067, and measures data from 20068 to 20072. No data is rounded in the archive. The initial 100-bar batch was rejected due to a 2000 character limit, and no partial raw archive is accepted. The code is set to accept adaptive JSON chunks of 1700 or less. The code's initial short default and decimal percent label were corrected before the final archive. This is a fixed diagnostic, not a live venue with 15-minute fills, and all 8 cells are negative, with no retune.

A news scan found the script output showing calculations for IPHA. The script created a list of dictionaries called sens, which contains different financial scenarios. Each scenario includes the exchange rate of euros per US dollar, assumed monthly use in euros, new financing in euros, and the number of months the financing will last. The exchange rates used were 0.8, 0.9, and 1.0 euros per US dollar, and the assumed monthly use was 2, 4, and 6 million euros. The new financing was calculated as 27.6 million euros plus 75 times the exchange rate. The script also calculated the time hold and market values from a list of events, and then calculated statistical values such as the correlation between time hold and market, and the beta and alpha values. The results were written to a file called ipha_results.json and printed to the screen.

A news scan found the following data: the mean percentage is calculated as the mean of the returns multiplied by 100, or mean_pct = float(r.mean()*100). The average win percentage is average_win_pct, which is None if w is None, otherwise it is w multiplied by 100. The average loss percentage is average_loss_pct, which is None if l is None, otherwise it is l multiplied by 100. The mean confidence interval percentage is mean_ci_pct, which is a list of values, or ci.tolist(). The raw quarter Kelly capital fraction is k. The number of liquid events is liquid_events, which is the sum of events where the prior dollar volume median is greater than or equal to 1 million, or sum(e['prior_dollar_volume_median']>=1e6 for e in ev). The returns percentage is returns_pct, which is the returns multiplied by 100, or (r*100).tolist(). The last data point is last = d.iloc[-1]. The true range, or tr, is calculated as the maximum of the high minus low, the absolute value of the high minus the previous close, and the absolute value of the low minus the previous close, or pd.concat([d.h-d.l,(d.h-d.c.shift()).abs(),(d.l-d.c.shift()).abs()],axis=1).max(axis=1). A dictionary called tech contains technical indicators, including the date, close, open, volume, dollar volume, median dollar volume over 20 days, simple moving average over 20 days, simple moving average over 50 days, return over 20 days, and average daily range over 20 days. The number of whole shares is whole_shares, which is 49.14 divided by the last close, or int(49.14//last.c). The number of bars is bars, which is the length of the data, or len(d). A list called blocks contains historical 5-session OHLC block scenarios, which are not catalyst probability or executable fills. Each block is a dictionary containing the terminal value, low, and high. A simulation dictionary called sim contains the number of paths, which is 50000, the number of blocks, the terminal percentage, loss frequency, and full book P05 PNL. A bridge dictionary contains the total assets, short, noncurrent, cash residual, net equity, fees percentage, new shares, old total shares, new total shares, share count increase percentage, ownership loss percentage, and conditional USD upfront.

A news scan found the Python source code. The code has two loops, one for slip values of 0.0025 and 0.005, and one for split values of 'train' and 'holdout'. It calculates the start and end times of events. It creates an index of dates and times, and a series of marks with 'o' and 'c' values. It also creates series for sizes and prices, and a list for returns. For each event, it updates the sizes and prices series, and calculates the return. It creates a portfolio from orders using the marks, sizes, prices, and fees. It reconciles the cash returns and asserts that they are close to the trade returns. It calculates the total return, and the confidence interval of the returns. It calculates the win and loss rates, and the k value. The results are appended to a list with the mode, slippage, split, number of trades, return percentage, drawdown percentage, and win percentage.

A news scan found the Python source code for an automated trading fund manager. The code starts by importing necessary libraries, including json, numpy, pandas, and vectorbt. It then defines a function called read that takes a stock symbol as input and returns a pandas DataFrame containing the stock's price data. The function reads the data from a json file, converts the date column to a datetime format, and checks that the index is unique and monotonically increasing. The code then reads the data for the stocks IPHA and SPY. It checks that the indices of the two DataFrames are equal. The code then defines a list of dates and uses a random number generator to generate events. For each date, it calculates the prior average daily range and sets an entry price, stop loss, and target price. It then checks if the stock price reaches the stop loss or target price and records the outcome. The code records the event details, including the entry and exit dates, prices, and reason for exit. The events are split into training and holdout sets based on the date. The code then starts to iterate over two modes, 'barrier' and 'time', but the rest of the code is not shown.

The remote output for job173 has been verified before the report. The actual payload was received at 04:16:07.918 UTC with an exit code of 0, and it contained 21 lines, all of which are visible, totaling 1016 characters. The next update is scheduled for 05:15:46.451 UTC, before the 06 run. The links for FTK, cat7 Lake Street, and AVAV Q1 FY27 are visible, and the SSL certificates for both pages are okay. However, the ACSA relay dated August 28 is not live, and gaps remain in the Bombardier and CENTCOM data. A replacement has been accepted and verified remotely. There have been no substantive changes to the cash or thesis tested from the hashes. The journal entries 603-605 are complete, and the final decision is to hold cash and not propose any changes. The process took approximately 16 minutes, which is within the planned 44 minutes, and the final journal block has been reserved and completed without any thread overrun.

The final report was completed at 04:16 UTC, which is about 15 minutes of the 43 minutes planned and 45 minutes hard cap. The final record and report phase took place from 04:11 to 04:16 and preserved 5 minutes. The audit download and primary controls and test overlapped, and no question thread exceeded the allowance. A deep dive into question 1772 found that historical daily and hourly conflicts are not explained by UTC offset, which refers to the difference in time zones. The minute and Coinbase December 2 controls favor intraday opens, but the vendor cause remains unresolved. Most strategy changes are due to missing history support from 2028, not price correction. All 18 hourly out-of-sample losses occurred, with a baseline from 1991 to 1993. Cost decomposition from 2027 explains the large hourly drag, which refers to a decrease in performance. A simulation from 2002 and a cash decision from 2030 provided lessons 600 to 602. There were 120 cell observations from 1993 to 1970, but CLI 2005 is not valid trade evidence. There are no proposals, and the fund has 235.86 in cash. The final open and pending queries are empty. Monitor 169 was resolved, and 173 will be armed at 06 UTC. Compact 170 and 171 are scheduled through 06:01, with local full footers of 1441 and 1283 characters. The last remote output at 04:15 is still null and not verified, with 14 out of 25 preserved. There are explicit gaps in vendor historical root cause, full independent history, account-specific execution, and 15-minute exits, as well as remote collection. There is no new ETF flow claim and no external message. The next research will focus on the first actual collector footer, then independent unused edge, and will not repeat optimization on the same holdout.

A news scan found the headline about a handoff at 04UTC. The Natref cash thesis has been frozen. The next priorities will be determined at 06UTC. One deep dive analysis of tree 1769 resulted in analyses of years 1988, 1989, 1994, 1995, 1999, 2000, and 2001. Prior uncertainty has been resolved towards operable capacity and global cracks benefit, but own outage rent is unproven. There is no calibrated Bayesian prior or likelihood and no invented probability update. A strong countercase shows good real cash and a 7.68% historical FX yield proxy, but there is insufficient stock edge. The freeze prospective acceptance criteria include current Natref jet throughput, current ORTIA stock and paid import costs, plus Prax access duration and receipts to improve the physical and cash case. Any unpaid inventory, cost increase, or loss of Prax capacity weakens the case. Either branch independent costed 15-minute net payoff is still required. The next 06UTC read job is 173, with an actual timestamp and footer, not a schedule. All 19 prior URLs, plus SSL news and GCIS, equal 21, with FTK route fixed and approximately 1030 output under 2000. A broad 162 was cancelled after a test, with a 5 job cap remaining. No remote execution has been claimed yet. A review of FTK Lake Street 2039 shows that the publication condition has been met, with H1 FCF still at -7.819 million. AVAV 2003 is funded, meaning it is authorized and billable, but not collected. The Q1 transcript is still unlinked in the inspected page. Monitor 163 has been resolved, and 175 SSL Sept 30 is armed. Jobs 162 FTK, 158 AVAV, 50 EIA, 124 macro, and 11 dividend, among others, have been preserved. Existing jobs 12, 18, 36, and 73 remain unchanged. The local workspace has raw PDFs and text, with full 20F and 70 SIP bars, economics and json files, and tech and stats files. Numeric inputs and outputs have been recorded from 2018 to 2028. Scripts can regenerate from sources if they are ephemeral. There are open gaps in current daily flows, jet contract netbacks, repair insurance cash, Prax sale terms, full Q&A, current FX and EV adjustments, and Bombardier and CENTCOM access. Source collection stopped when new hits repeated the same origins, with no thread overrunning the allowance. Research finished in approximately 15 minutes, with the final 5 minutes protected, and the hard 45-minute limit not approached. There are no orders, exit, adjust, or fund decisions.

A news scan found the headline about funding issuer supply through 06UTC, which replaces 160 with an exit of 0. The fund is for crypto research only and was observed on 2026-09-10T04:16:17.275265+00:00. The source of this information is https://www.okx.com/api/v5/public/time, which has an offset of 0.0736 seconds and an uncertainty of 0.1521 seconds. Another source, https://www.okx.com/api/v5/public/funding-rate and https://www.okx.com/api/v5/public/open-interest, provided the following data: for Bitcoin, the funding rate is 0.7796 basis points and the settled rate is 0.495 basis points, with an open interest of 27153.747 coins. For Ethereum, the funding rate is 0.3003 basis points and the settled rate is 0.1779 basis points, with an open interest of 590617.768 coins. For Solana, the funding rate is -0.6834 basis points and the settled rate is -0.9099 basis points, with an open interest of 2704071.58 coins. A source from https://www.circle.com/transparency reported the following data on September 07, 2026: the balance is 74.28, with issue and redeem amounts of 8.8, 7.86, 36.59, 34.37, 333.52, and 331.64 for different time periods. Another source from https://www.usdc.com/ could not be parsed due to ambiguity. A stablecoin source from https://stablecoins.llama.fi/stablecoin/2 reported the following data: the circulating value is 74386006033.1783, with values on Ethereum, Solana, and Hyperliquid L1 of 47146319984.45, 7103929420.46, and 6887999675.1 respectively. The crypto collection has ended, with definitions and cutoffs differing, and the flow is not buying, resulting in an eligibility of 0.

The fund is for supply, and the research was done on 2026-09-10T04:16:07.918037+00:00. The eligible amount is 0. The job script contains a URL map. The following companies were checked: Maersk, which is okay with code 4bb96bc8a2a4, Typhoon, which is okay with code aedd68feae0f, Hapag, which is okay with code 8ab284858970, SeaIntel, which is okay with code 0eeef616fcf3, WH, which is okay with code e0e2e6ade515, Bombardier, which has a challenge with code f492d282afa2, GD, which had an 8-K filing on 2026-08-07 and another on 2026-08-04, CENTCOM, which had an HTTP error, FRO, which is okay with code 3aaf7244dc4a, STNG, which is okay with code c54b428d736e, ASPI, which had an event on Tuesday, 08 Sep 26 09:15:00 -0400, ACP, which is okay with code 6e21e0bc4ee1 and an advisory about a draft adjustment, SEI, which is okay with code 140860f05cb5, OESX, which is okay with code f7316d1affda, Canada, which is okay with code 09084e6193e2, FTK, which is okay with code b2455970b637 and had a news article about raising 2026 guidance, FTKslides, which is okay with code 32911d7c68cb and has a Lake Street investor presentation, AVAV, which is okay with code 4bd851329b23 and has a Q1 fiscal report, SSL, which is okay with code e91c77fc88a0, SSLnews, which is okay with code 81cee979207d, and ACSArelay, which is okay with code c6034ead5d18. The job attempted 21 tasks and completed them, but the hash is not news, and the ACSA baseline from August 28 is not live.

A news scan found the headline 'Academy Sports + Outdoors Reports Second Quarter Fiscal 2026 Results' for the stock ticker ASO, and the headline 'Flotek Raises 2026 Guidance Again on Continued Strength of International Sales' for the stock ticker FTK. The quant fund reported no orders at 2026-09-10T04:16:09.962858+00:00 UTC, and again reported no orders when it completed its task, covering 2 symbols.

The Catalyst research fund has a size of 0 as of September 10, 2026, at 04:16:01 UTC. Recent news includes: TLX on September 9, 2026, presenting at the Morgan Stanley 24th Annual Global Healthcare Conference and H.C. Wainwright 28th Annual Global Investment Conference, available at https://telixpharma.com/news-views/telix-to-present-at-the-morgan-stanley-24th-annual-global-healthcare-conference-and-h-c-wainwright-28th-annual-global-investment-conference/. AVO announced fiscal 2026 third quarter financial results on September 8, 2026, available at https://investors.missionproduce.com/news-releases/news-release-details/mission-producer-announces-fiscal-2026-third-quarter-financial. ABM reported fiscal third quarter 2026 results on September 8, 2026, available at https://abm.gcs-web.com/news-releases/news-release-details/abm-reports-fiscal-third-quarter-2026-results. ROIV announced positive results from the Phocus study of mosliciguat on September 8, 2026, available at https://investor.roivant.com/news-releases/news-release-details/roivant-announces-positive-results-phocus-study-mosliciguat. Other updates include UNFI's 8-K filing on September 8, 2026, DRI's 10-K filing on July 24, 2026, PPIH's 10-Q filing on September 9, 2026, MRK's 10-Q filing on August 7, 2026, and WLTH's 8-K filing on September 9, 2026. EPM scheduled fiscal Q4 and full year 2026 results on September 8, 2026, available at https://ir.evolutionpetroleum.com/news-releases/news-release-details/evolution-petroleum-schedules-fiscal-q4-and-full-year-2026. IPHA filed a 6-K on September 2, 2026. The following HTTP requests were successful: KGF with hash dda88bf6ef8ad805, ASKUL with hash e6f0d85c03a37ce6, IPHAnews with hash 3ed2efab2aa2a6f4, and SOBI with hash cec04eec9001d2f4. Research is ongoing, with a focus on reading originals and validating information. A review of IPHA is scheduled for September 17, and a receipt from Sobi is conditional on Q1 2028. Monitors 165, 168, 167, 137, and 22 are being preserved. Sources and dates are available in the job script and journal. No orders have been placed.

A news scan found the following information. The fund manager of the quant fund made a note on 2026-09-10T04:15:50.746308+00:00 UTC, stating that there were no orders. The manager then checked four issuers. The first issuer, INBX, had a status of 200, which means the request was successful, with 3 matches and 1 omitted link. The links were about Inhibrx's INBRX-106 study. The second issuer, TYRA, also had a status of 200, with 3 matches and 1 omitted link, and the links were about Tyra Biosciences' SURF302 results. The third issuer, IRD, had a status of 200, with 4 matches and 3 omitted links, and the links were about Opus Genetics' clinical trial. The fourth issuer, AVAV, had a status of 200, with 1 match and no omitted links, and the link was about Aerovironment's fiscal 2027 first quarter results. The fund manager noted that the task was complete, with 4 symbols checked, and the coverage was bounded keyword links, with omitted or error links explicitly stated, but this is not proof that there was no other news.