SimpleQuant 1.6.0

下载

首次安装请点击下载（已安装 0.1.1 及以后版本的程序会自动检测并下载更新）：

系统	安装包
Windows 10 / 11（64 位，无需管理员权限）	SimpleQuant-1.6.0-Setup.exe
macOS 11 及以上（Apple 芯片：M1 / M2 / M3 / M4…）	SimpleQuant-1.6.0-macos-arm64.dmg
Linux（x86_64）	SimpleQuant-1.6.0-linux-x86_64.tar.gz

Intel 芯片的 Mac 最后支持的版本是 0.2.5，已安装的可以继续使用，此后的版本不再提供。

国内下载慢可以用 123 云盘：https://1829763294.share.123pan.cn/123pan/mv2ejv-wbhtA?pwd=ZfDO

macOS 版未经 Apple 公证，第一次打开会被拦下：请在「系统设置 → 隐私与安全性」中点「仍要打开」，或在终端运行 xattr -dr com.apple.quarantine /Applications/SimpleQuant.app。

下方「Assets」中的其余文件（manifest、-from- 增量包）供程序自动更新使用，无需下载。

更新内容

• 新增：定投。回测页「资金与交易成本」中可填写每期追加金额（每月或每周第一个交易日到账，初始资金视为第一期），并新增两个策略模板：「定期定额定投」每期资金到账即买入；「估值定投」按市盈率历史百分位调整每期投入（默认低于 30% 投 2 倍、高于 70% 投 0.5 倍，少投的资金留作储备、低估时使用，储备按设定的年化收益计息，默认 2%，相当于放在货币基金）。设置追加金额后，总收益按累计投入计算，并显示资金加权年化收益率（XIRR）、累计投入与时间加权收益率；基准为同样的资金、同样的日期买入持有；最大回撤、波动率与夏普比率按扣除追加资金后的净值计算，追加的资金不会被当成收益或掩盖亏损。资金曲线增加累计投入线。参数优化、滚动优化（「延续持仓」方式）、单标的模拟账户与导出 Python 脚本也支持追加资金；模拟盘在追加日的前一天提醒转入资金，收益按累计投入计算；AI 解读按定投的口径（累计投入、XIRR）评价。初始资金最低可填 1000 元（原为 1 万元），方便小额起步的定投。暂不支持：多因子选股、选股与组合模拟账户、导出到聚宽 / 掘金 / QMT。用到定投的方案文件，旧版软件导入时会提示更新
• 新增：多因子选股的股票池「全 A 股」：沪深两市全部 A 股，按上市、退市日期每月确定当时在市的股票（含此后退市的，避免幸存者偏差）；交易日历用上证指数，基准为中证全指全收益指数（H00985）。数据量较大（2020 年起约 5400 只），受 BaoStock 每天 5 万次请求的限制，首次下载日线、财务与分红约需 3 天：每天下载到上限自动停止，次日再点下载从断点继续。暂不支持导出到聚宽 / 掘金 / QMT 在平台上选股，可用「按名单调仓」导出
• 改进：多因子选股处理股票退市（所有股票池）。最后交易日事先可知的（退市整理期结束、吸收合并停牌换股，即最后交易日距退市日不超过 15 天），持有的在最后一个交易日开盘卖出；停牌后才宣布退市的（交易类退市没有整理期）当时无法预知，不提前卖出。跌停或停牌卖不出、一直持有到退市日的，退市前为 ST、股价低于 1 元或简称带「退」的（困境退市）按 0 计价，其余（吸收合并、主动退市）按最后价格折为现金。之前退市股按最后价格一直挂在持仓里卖不出，资金无法再投入，困境退市的损失也被低估。持有过退市股的回测结果会变化：例如中证500 中市值最小的 10 只、每月调仓，2021 年起由 +0.6% 变为 −5.0%（持有过的 *ST中天、鹏都农牧停牌后退市，按 0 计价）；沪深300、中证500 按市盈率倒数或低波动选股的结果不变
• 改进：选股数据（BaoStock）改为逐只下载，并按天统计请求次数。BaoStock 规定每天请求不超过 5 万次、不能多个连接同时访问，超过后 IP 会被暂停访问（当年第一次 6 小时，之后按次数累加）。当天请求接近上限时下载自动停止，已下载的数据保留，次日再次下载从断点继续；IP 已被暂停时立即停止并说明原因，不再反复重试。选股页的「并行连接数」设置随之去掉
• 修复：多因子选股按 IC / ICIR 加权、回看期超过 252 个交易日时，回测开始前只多取约 400 天数据，开头一段的权重实际用到的历史比设定短（如设为 504 天，第一年的调仓日只用到约 250 天）。现在按回看期多取数据；本地数据不够早时，在结果中提示受影响的调仓日数。回看期不超过 252 天的结果不变；导入的方案文件仍按其中记录的数据起点回测

本软件仅供学习与研究使用，不构成任何投资建议。

---

Download

For a fresh install: SimpleQuant-1.6.0-Setup.exe (Windows), SimpleQuant-1.6.0-macos-arm64.dmg (macOS, Apple silicon) or SimpleQuant-1.6.0-linux-x86_64.tar.gz (Linux). Intel Macs are supported up to 0.2.5. Installed copies (0.1.1 or later) update themselves. The macOS app is not notarized: on first launch, allow it under System Settings → Privacy & Security → Open Anyway. The other assets are used by the auto-updater.

Changes

• New: dollar-cost averaging. On the backtest page, Capital & trading costs now takes a contribution per period (added on the first trading day of each month or week; the initial capital counts as the first installment), with two new templates: "Dollar-cost averaging" buys whenever a contribution arrives, and "Valuation-based DCA" scales each contribution by the P/E percentile (by default 2x below 30% and 0.5x above 70%; money held back is kept as a reserve for cheap periods and earns a set annual rate, 2% by default, like a money market fund). With contributions, total return is measured against the total invested, and the money-weighted annual return (XIRR), total invested and time-weighted return are shown; the benchmark invests the same amounts on the same dates and holds; maximum drawdown, volatility and Sharpe ratio use the unit value with contributions removed, so new money is neither counted as gains nor hides losses. The equity chart adds a total-invested line. Parameter optimization, walk-forward optimization (with "Keep positions"), single-asset paper accounts and exported Python scripts also support contributions; paper trading reminds you to transfer the money the day before and measures returns against the total invested; AI explanations judge DCA results by total invested and XIRR. Initial capital can now be as low as 1,000 (was 10,000), for DCA plans that start small. Not yet supported: multi-factor selection, selection and portfolio paper accounts, and exporting to JoinQuant / MyQuant / QMT. Older versions ask to be updated when importing a plan file that uses DCA
• New: an "All A-shares" universe for multi-factor selection: every A-share on the Shanghai and Shenzhen exchanges, with membership set monthly from listing and delisting dates (stocks delisted later are included, avoiding survivorship bias); the trading calendar is the SSE Composite and the benchmark is the CSI All Share total return index (H00985). The data is large (about 5,400 stocks from 2020); with BaoStock's limit of 50,000 requests a day, the first download of daily bars, financials and dividends takes about 3 days: it stops at the limit each day and resumes from there when you download again the next day. Selecting on JoinQuant / MyQuant / QMT is not supported for it yet; use the "follow picks" export
• Improved: multi-factor selection now handles delistings (all universes). When the last trading day is known in advance (the end of a delisting period, or a merger's share-swap suspension, i.e. delisting within 15 days of the last trade), a held stock is sold at the open of its last trading day; stocks delisted only after a suspension (trading-based delistings without a delisting period) could not have been foreseen and are not sold early. If it is limit-down or suspended and still held on the delisting date, it is written off when it was ST, below 1 yuan or renamed with 退 beforehand (distressed delisting), and otherwise (mergers, voluntary delistings) settled in cash at the last price. Previously a delisted stock stayed in the portfolio at its last price, unsellable, locking up the cash and understating losses from distressed delistings. Backtests that held delisted stocks will change: for example, the 10 smallest CSI 500 stocks rebalanced monthly from 2021 go from +0.6% to −5.0% (*ST Zhongtian and Pengdu Agriculture were delisted after suspensions and are written off); CSI 300 and CSI 500 earnings-yield and low-volatility selections are unchanged
• Improved: stock-selection data (BaoStock) is now downloaded one stock at a time, with requests counted per day. BaoStock allows 50,000 requests a day and no concurrent connections, and suspends the IP beyond that (6 hours the first time in a year, longer after that). The download stops when today's requests near the limit, keeping what was downloaded, and resumes from there the next day; if the IP is already suspended it stops at once with an explanation instead of retrying. The "parallel connections" setting on the selection page is removed
• Fix: with IC / ICIR weighting and a lookback longer than 252 trading days, only about 400 days of data before the start were loaded, so early rebalances used less history than configured (with 504 days, the first year used only about 250). Enough data for the lookback is now loaded; when local data does not reach back far enough, the result notes how many rebalance dates are affected. Results with lookbacks up to 252 days are unchanged; imported plan files still use the data start they recorded

This software is for learning and research only and is not investment advice.

SimpleQuant 1.5.1

下载

首次安装请点击下载（已安装 0.1.1 及以后版本的程序会自动检测并下载更新）：

系统	安装包
Windows 10 / 11（64 位，无需管理员权限）	SimpleQuant-1.5.1-Setup.exe
macOS 11 及以上（Apple 芯片：M1 / M2 / M3 / M4…）	SimpleQuant-1.5.1-macos-arm64.dmg
Linux（x86_64）	SimpleQuant-1.5.1-linux-x86_64.tar.gz

Intel 芯片的 Mac 最后支持的版本是 0.2.5，已安装的可以继续使用，此后的版本不再提供。

国内下载慢可以用 123 云盘：https://1829763294.share.123pan.cn/123pan/mv2ejv-wbhtA?pwd=ZfDO

macOS 版未经 Apple 公证，第一次打开会被拦下：请在「系统设置 → 隐私与安全性」中点「仍要打开」，或在终端运行 xattr -dr com.apple.quarantine /Applications/SimpleQuant.app。

下方「Assets」中的其余文件（manifest、-from- 增量包）供程序自动更新使用，无需下载。

更新内容

• 新增：完成 10 次回测（择时或选股）后，询问一次是否愿意在 GitHub 上点星或支持开发；选「以后再说」时过一段时间再询问一次，选「不再提示」或打开了链接后不再询问。设置页「关于」中也有这两个链接
• 修复：多因子选股「载入已保存的选股策略」没有恢复保存时的初始资金与费率。现在保存策略时一并记下，载入时恢复；此前保存的策略没有记录，载入时保持当前设置，重新保存一次即可
• 修复：按加法复权提示重新下载数据后，回测页仍选着旧数据。现在下载同一标的（同一数据源、周期、复权）且区间覆盖旧数据时，旧数据被替换，回测页等处自动改用新数据；数据选择框中加法复权的数据标为「后复权（加法）」

本软件仅供学习与研究使用，不构成任何投资建议。

---

Download

For a fresh install: SimpleQuant-1.5.1-Setup.exe (Windows), SimpleQuant-1.5.1-macos-arm64.dmg (macOS, Apple silicon) or SimpleQuant-1.5.1-linux-x86_64.tar.gz (Linux). Intel Macs are supported up to 0.2.5. Installed copies (0.1.1 or later) update themselves. The macOS app is not notarized: on first launch, allow it under System Settings → Privacy & Security → Open Anyway. The other assets are used by the auto-updater.

Changes

• New: after 10 backtests (timing or stock selection), SimpleQuant asks once whether you would like to star it on GitHub or support its development. "Maybe later" asks again much later; "Don't ask again" or opening a link stops the prompt for good. The links are also in Settings → About
• Fix: loading a saved stock selection strategy did not restore its starting capital and costs. They are now saved with the strategy and restored on load; strategies saved earlier keep the current settings until saved again
• Fix: after downloading again as the additive-adjustment notice suggests, the backtest page still used the old data. A new download of the same symbol (same source, frequency and adjustment) whose range covers an older dataset now replaces it, and pages that had the old data selected switch to the new one; additively adjusted data is labelled "Back-adjusted (additive)" in data pickers

This software is for learning and research only and is not investment advice.

SimpleQuant 1.5.0

下载

首次安装请点击下载（已安装 0.1.1 及以后版本的程序会自动检测并下载更新）：

系统	安装包
Windows 10 / 11（64 位，无需管理员权限）	SimpleQuant-1.5.0-Setup.exe
macOS 11 及以上（Apple 芯片：M1 / M2 / M3 / M4…）	SimpleQuant-1.5.0-macos-arm64.dmg
Linux（x86_64）	SimpleQuant-1.5.0-linux-x86_64.tar.gz

Intel 芯片的 Mac 最后支持的版本是 0.2.5，已安装的可以继续使用，此后的版本不再提供。

国内下载慢可以用 123 云盘：https://1829763294.share.123pan.cn/123pan/mv2ejv-wbhtA?pwd=ZfDO

macOS 版未经 Apple 公证，第一次打开会被拦下：请在「系统设置 → 隐私与安全性」中点「仍要打开」，或在终端运行 xattr -dr com.apple.quarantine /Applications/SimpleQuant.app。

下方「Assets」中的其余文件（manifest、-from- 增量包）供程序自动更新使用，无需下载。

更新内容

• 改进：多因子选股（沪深300、中证500、上证50）的基准改用全收益指数（H00300、H00905、H00016，中证指数官网，分红再投资）。之前用的是价格指数，不含分红，而策略收益含分红再投资，超额收益每年约多算一个股息率（沪深300 约 2%～3%）。超额收益会变小：例如沪深300 中按市盈率倒数选 10 只、每月调仓，2022 年起基准由 −11.4% 变为 +0.4%，超额收益由 +69.3% 变为 +57.5%；中证500 同样做法 2020 年起超额收益由 +13.4% 变为 −2.1%。策略本身的收益不变。全收益指数在更新数据时一起下载，已有数据第一次载入时自动补下载；没下载到时退回价格指数并提示。模拟盘资产曲线中已记录的基准一并改为全收益指数。可转债的基准仍为中证转债指数
• 改进：AKShare 数据源的股票与 ETF 复权一律改为等比复权（分红再投资，与 BaoStock、多因子选股一致）：股票为东方财富不复权价 × BaoStock 复权因子，ETF 由不复权价与新浪累计分红计算。之前用的是东方财富自带的复权价，属于加法复权（复权价 = 不复权价 + 上市以来的累计分红），每天的涨跌幅被压缩，分红越多越明显：红利 ETF 510880 的单日涨跌只有真实的约三分之二，2015 年起持有收益由 +68.6% 变为 +101.5%；510300 2013 年起由 +99.6% 变为 +109.2%；159915 基本不分红，不变。按百分比的止损、持仓收益率、波动率因此更准确。择时回测的结果会变化。已下载的数据保持不变，回测时提示「加法复权」，在「数据」页勾选「忽略本地已有数据，重新下载」即可更新；取不到复权因子时（如北交所股票）仍用东方财富的复权价并提示。已有的择时模拟盘账户保持原来的加法复权（不改写已有成交），新开的账户用等比复权。导出的 Python 脚本同样改为等比复权。方案文件记录复权算法，导入时与本机数据不同会提示
• 改进：择时回测按真实股价凑 100 股一手（与多因子选股相同）。复权数据下载时附带不复权收盘价（AKShare、BaoStock 日线），下单时按真实股价计算能买几手，成交明细、期末持仓、模拟盘信号显示真实股数和成交价（K 线图上的买卖点仍按复权价）。之前按复权价凑整手，后复权价远高于实际的标的一手所需资金被高估：例如贵州茅台后复权价约为实际的 6.5 倍，30 万元资金按复权价一手要 73 万元、无法买入，现在按实际约 1100 元一股可以买入 200 股；中国银行的收益基本不变，但成交明细由「15000 股 × 6.34 元」变为实际的「27100 股 × 3.51 元」。调仓时持仓与目标相差不足一手不买卖，避免为零头支付最低佣金。较早下载、没有不复权价的数据仍按复权价计算，重新下载即可；已有的模拟盘账户保持原来的方式。导出的 Python 脚本同样按真实股价计算

本软件仅供学习与研究使用，不构成任何投资建议。

---

Download

For a fresh install: SimpleQuant-1.5.0-Setup.exe (Windows), SimpleQuant-1.5.0-macos-arm64.dmg (macOS, Apple silicon) or SimpleQuant-1.5.0-linux-x86_64.tar.gz (Linux). Intel Macs are supported up to 0.2.5. Installed copies (0.1.1 or later) update themselves. The macOS app is not notarized: on first launch, allow it under System Settings → Privacy & Security → Open Anyway. The other assets are used by the auto-updater.

Changes

• Improved: the benchmark for multi-factor selection (CSI 300, CSI 500, SSE 50) is now the total return index (H00300, H00905, H00016 from the CSI website, dividends reinvested). It used to be the price index, which excludes dividends while the strategy reinvests them, so excess return was overstated by roughly the dividend yield each year (about 2–3% for CSI 300). Excess returns get smaller: for example, the top 10 CSI 300 stocks by earnings yield, rebalanced monthly from 2022, now have a benchmark of +0.4% instead of −11.4% and an excess return of +57.5% instead of +69.3%; the same strategy on CSI 500 from 2020 goes from +13.4% to −2.1% excess. Strategy returns themselves are unchanged. The total return index is downloaded with data updates, and fetched automatically the first time existing data is loaded; if it cannot be downloaded, the price index is used with a note. Benchmarks already recorded in paper-trading equity curves are switched to the total return index as well. Convertible bonds still use the CSI Convertible Bond Index
• Improved: stock and ETF adjustment from the AKShare source is now always proportional (dividends reinvested, matching BaoStock and stock selection): stocks are Eastmoney raw prices × BaoStock adjustment factors, and ETFs are rebuilt from raw prices and Sina cumulative dividends. Eastmoney's own adjusted prices, used until now, are additive (adjusted price = raw price + cumulative dividends since listing), which compresses daily percentage moves, more so the more dividends were paid: the dividend ETF 510880 showed only about two thirds of its real daily moves, and buy-and-hold from 2015 goes from +68.6% to +101.5%; 510300 from 2013 goes from +99.6% to +109.2%; 159915 pays almost no dividends and is unchanged. Percentage stop-losses, position returns and volatility are therefore more accurate. Timing backtest results will change. Data already downloaded is kept as it is, with an "additive" note in backtests; re-download it on the Data page (tick "Ignore local data and re-download") to update. When the factors are unavailable (e.g. Beijing Stock Exchange stocks), Eastmoney's adjusted prices are still used, with a note. Existing timing paper-trading accounts keep additive adjustment (their recorded fills are not rewritten); new accounts use proportional adjustment. Exported Python scripts use proportional adjustment as well. Plan files record the adjustment method, and importing one onto data with a different method shows a note
• Improved: timing backtests size 100-share lots at the real share price (as stock selection already does). Adjusted data now comes with the unadjusted close (AKShare and BaoStock daily bars), orders are sized by how many lots the real price allows, and the fill list, final positions and paper-trading signals show real share counts and prices (trade markers on the price chart stay on the adjusted price). Lots used to be sized at the adjusted price, which overstated the cost of a lot when back-adjusted prices are far above the real ones: Kweichow Moutai's back-adjusted price is about 6.5 times the real price, so with 300,000 CNY one lot "cost" 730,000 and nothing was bought, whereas now 200 shares at about 1,100 CNY are bought; Bank of China returns barely change, but the fill list shows the real "27,100 shares × 3.51" instead of "15,000 × 6.34". When rebalancing, differences smaller than one lot are left alone to avoid paying the minimum commission on odd shares. Data downloaded earlier without unadjusted prices still uses the adjusted price until re-downloaded; existing paper-trading accounts keep the old behaviour. Exported Python scripts size lots the same way

This software is for learning and research only and is not investment advice.

SimpleQuant 1.4.0

下载

首次安装请点击下载（已安装 0.1.1 及以后版本的程序会自动检测并下载更新）：

系统	安装包
Windows 10 / 11（64 位，无需管理员权限）	SimpleQuant-1.4.0-Setup.exe
macOS 11 及以上（Apple 芯片：M1 / M2 / M3 / M4…）	SimpleQuant-1.4.0-macos-arm64.dmg
Linux（x86_64）	SimpleQuant-1.4.0-linux-x86_64.tar.gz

Intel 芯片的 Mac 最后支持的版本是 0.2.5，已安装的可以继续使用，此后的版本不再提供。

国内下载慢可以用 123 云盘：https://1829763294.share.123pan.cn/123pan/mv2ejv-wbhtA?pwd=ZfDO

macOS 版未经 Apple 公证，第一次打开会被拦下：请在「系统设置 → 隐私与安全性」中点「仍要打开」，或在终端运行 xattr -dr com.apple.quarantine /Applications/SimpleQuant.app。

下方「Assets」中的其余文件（manifest、-from- 增量包）供程序自动更新使用，无需下载。

更新内容

• 改进：择时规则用到「市盈率 / 市净率历史百分位」时，所选回测区间开始前的数据只用于计算指标、不交易，收益与交易仍从区间开始日计算。之前百分位只用区间内的数据计算，第 1 年没有百分位，回看年数也凑不满；例如 510300「百分位低于 20 买入、高于 80 卖出」，2020 年起的收益由 +12.9% 变为 +48.8%。回测、批量回测、参数优化（含样本外检验）与滚动优化都按此计算；区间开始前的估值数据不足回看年数时会提示。没有用到估值百分位的策略，结果不变
• 新增：可信度检查加入「调仓日检验」。定期调仓的策略（动量轮动、定比再平衡、多因子选股）结果可能取决于调仓恰好落在哪一天：规则与区间不变，把调仓日依次错开 1 ~ N−1 个交易日重新回测（N 为调仓间隔，最多 20 种），列出总收益与最大回撤的范围、中位数和当前结果的排名；当前结果明显高于中位数时警告。例如 510300 + 510500 动量轮动（回看 20、每 5 天调仓），2014 年起当前结果 +267%，错开调仓日后为 +131%～+224%，中位数 +141%。择时回测在报告显示后自动进行；选股回测较慢，在可信度检查中点「调仓日检验」进行，可随时停止

本软件仅供学习与研究使用，不构成任何投资建议。

---

Download

For a fresh install: SimpleQuant-1.4.0-Setup.exe (Windows), SimpleQuant-1.4.0-macos-arm64.dmg (macOS, Apple silicon) or SimpleQuant-1.4.0-linux-x86_64.tar.gz (Linux). Intel Macs are supported up to 0.2.5. Installed copies (0.1.1 or later) update themselves. The macOS app is not notarized: on first launch, allow it under System Settings → Privacy & Security → Open Anyway. The other assets are used by the auto-updater.

Changes

• Improved: when a timing rule uses the P/E or P/B percentile, data before the chosen backtest start is used only to compute indicators, with no trading; returns and trades still count from the start date. Previously the percentile used only data inside the range, so there was none in the first year and the lookback was never full; for example, 510300 "buy below the 20th percentile, sell above the 80th" from 2020 goes from +12.9% to +48.8%. Single and batch backtests, optimization (including the out-of-sample check) and walk-forward all work this way, with a note when the valuation data before the start is shorter than the lookback. Results of strategies that do not use valuation percentiles are unchanged
• New: a "rebalance-day test" in the credibility assessment. Strategies that rebalance on a schedule (momentum rotation, fixed-weight rebalancing, multi-factor selection) can depend on which day the rebalance happens to fall: with the same rules and period, the rebalance days are shifted by 1 to N−1 trading days (N = the rebalance interval, at most 20 variants) and re-run, showing the range and median of total return and maximum drawdown and where the current result ranks, with a warning when it is well above the median. For example, 510300 + 510500 momentum rotation (20-bar lookback, every 5 bars) from 2014 returns +267%, versus +131% to +224% (median +141%) with shifted rebalance days. Timing backtests run the test automatically after the report appears; selection backtests are slower, so click "Rebalance-day test" in the credibility assessment (it can be stopped at any time)

This software is for learning and research only and is not investment advice.

SimpleQuant 1.3.0

下载

首次安装请点击下载（已安装 0.1.1 及以后版本的程序会自动检测并下载更新）：

系统	安装包
Windows 10 / 11（64 位，无需管理员权限）	SimpleQuant-1.3.0-Setup.exe
macOS 11 及以上（Apple 芯片：M1 / M2 / M3 / M4…）	SimpleQuant-1.3.0-macos-arm64.dmg
Linux（x86_64）	SimpleQuant-1.3.0-linux-x86_64.tar.gz

Intel 芯片的 Mac 最后支持的版本是 0.2.5，已安装的可以继续使用，此后的版本不再提供。

国内下载慢可以用 123 云盘：https://1829763294.share.123pan.cn/123pan/mv2ejv-wbhtA?pwd=ZfDO

macOS 版未经 Apple 公证，第一次打开会被拦下：请在「系统设置 → 隐私与安全性」中点「仍要打开」，或在终端运行 xattr -dr com.apple.quarantine /Applications/SimpleQuant.app。

下方「Assets」中的其余文件（manifest、-from- 增量包）供程序自动更新使用，无需下载。

更新内容

• 新增：可转债「强赎触发天数」——近 30 个交易日中转股价值不低于 130（正股收盘价不低于转股价的 130%）的天数。强赎提示性公告没有历史数据，以此近似：多数转债满 15 天即可强赎，2018–2026 年的强赎公告中约 97% 在公告日已达到 15 天。可作为因子（「强赎触发天数（近 30 日）」），也可作为选股条件「强赎触发天数上限」（默认不限，原有结果不变）；自定义因子可用字段 call_days
• 用到强赎触发天数的选股方案文件，1.2.1 及以前的版本导入时会提示先更新软件

本软件仅供学习与研究使用，不构成任何投资建议。

---

Download

For a fresh install: SimpleQuant-1.3.0-Setup.exe (Windows), SimpleQuant-1.3.0-macos-arm64.dmg (macOS, Apple silicon) or SimpleQuant-1.3.0-linux-x86_64.tar.gz (Linux). Intel Macs are supported up to 0.2.5. Installed copies (0.1.1 or later) update themselves. The macOS app is not notarized: on first launch, allow it under System Settings → Privacy & Security → Open Anyway. The other assets are used by the auto-updater.

Changes

• New: "call-trigger days" for convertible bonds — the number of the last 30 trading days on which the conversion value was 130 or more (stock close at least 130% of the conversion price). Call warnings have no history, so this approximates them: most bonds can be called after 15 such days, and about 97% of 2018–2026 call notices had reached 15 on the notice date. Available as a factor ("Call-trigger days (last 30)") and as the filter "Max call-trigger days" (no limit by default, so existing results are unchanged); custom factors can use the field call_days
• Selection plan files that use call-trigger days ask version 1.2.1 and earlier to update first

This software is for learning and research only and is not investment advice.

SimpleQuant 1.2.1

下载

首次安装请点击下载（已安装 0.1.1 及以后版本的程序会自动检测并下载更新）：

系统	安装包
Windows 10 / 11（64 位，无需管理员权限）	SimpleQuant-1.2.1-Setup.exe
macOS 11 及以上（Apple 芯片：M1 / M2 / M3 / M4…）	SimpleQuant-1.2.1-macos-arm64.dmg
Linux（x86_64）	SimpleQuant-1.2.1-linux-x86_64.tar.gz

Intel 芯片的 Mac 最后支持的版本是 0.2.5，已安装的可以继续使用，此后的版本不再提供。

国内下载慢可以用 123 云盘：https://1829763294.share.123pan.cn/123pan/mv2ejv-wbhtA?pwd=ZfDO

macOS 版未经 Apple 公证，第一次打开会被拦下：请在「系统设置 → 隐私与安全性」中点「仍要打开」，或在终端运行 xattr -dr com.apple.quarantine /Applications/SimpleQuant.app。

下方「Assets」中的其余文件（manifest、-from- 增量包）供程序自动更新使用，无需下载。

更新内容

• 改进：多因子选股页载入更快。建立选股数据（股票池全部成分股的价格、财务与分红）的计算加快一倍以上（沪深300 2022 年起：约 27 秒 → 约 11 秒）；建好的数据保存在数据目录的 data_cache/panel_cache 中，之后重新打开软件、进入选股页约 1 秒即可载入。数据更新或软件升级后自动重建；最多保留最近使用的 3 份（每份约 40～80 MB）
• 改进：择时规则用到「市盈率 / 市净率历史百分位」时，如果所选回测区间开始前还有估值数据，回测页与参数优化页会提示：百分位仅使用区间内的数据计算（第 1 年没有百分位），并列出区间内的年数、回看年数与未参与计算的年数；将开始日期提前即可按完整的回看年数计算

本软件仅供学习与研究使用，不构成任何投资建议。

---

Download

For a fresh install: SimpleQuant-1.2.1-Setup.exe (Windows), SimpleQuant-1.2.1-macos-arm64.dmg (macOS, Apple silicon) or SimpleQuant-1.2.1-linux-x86_64.tar.gz (Linux). Intel Macs are supported up to 0.2.5. Installed copies (0.1.1 or later) update themselves. The macOS app is not notarized: on first launch, allow it under System Settings → Privacy & Security → Open Anyway. The other assets are used by the auto-updater.

Changes

• Improved: the stock selection page loads faster. Building the selection data (prices, financials and dividends of every constituent) is more than twice as fast (CSI 300 from 2022: about 27 s → about 11 s), and the result is saved in data_cache/panel_cache in the data folder, so after restarting the app the selection page loads in about a second. It is rebuilt automatically after a data update or an upgrade; the 3 most recently used copies are kept (about 40–80 MB each)
• Improved: when a timing rule uses the P/E or P/B percentile and valuation data exists before the chosen backtest start, the backtest and optimization pages note that the percentile is computed only from data inside the range (none in the first year), listing the years inside the range, the lookback and the years left unused; move the start date earlier to use the full lookback

This software is for learning and research only and is not investment advice.

SimpleQuant 1.2.0

下载

首次安装请点击下载（已安装 0.1.1 及以后版本的程序会自动检测并下载更新）：

系统	安装包
Windows 10 / 11（64 位，无需管理员权限）	SimpleQuant-1.2.0-Setup.exe
macOS 11 及以上（Apple 芯片：M1 / M2 / M3 / M4…）	SimpleQuant-1.2.0-macos-arm64.dmg
Linux（x86_64）	SimpleQuant-1.2.0-linux-x86_64.tar.gz

Intel 芯片的 Mac 最后支持的版本是 0.2.5，已安装的可以继续使用，此后的版本不再提供。

国内下载慢可以用 123 云盘：https://1829763294.share.123pan.cn/123pan/mv2ejv-wbhtA?pwd=ZfDO

macOS 版未经 Apple 公证，第一次打开会被拦下：请在「系统设置 → 隐私与安全性」中点「仍要打开」，或在终端运行 xattr -dr com.apple.quarantine /Applications/SimpleQuant.app。

下方「Assets」中的其余文件（manifest、-from- 增量包）供程序自动更新使用，无需下载。

更新内容

• 新增：多因子选股的「AI 因子挖掘」标签页：AI 每轮提出一批因子，软件只用区间前段的「挖掘期」评估（IC、t 值、与已有因子的相关性），把结果反馈给 AI 改进或换方向；全部轮次结束后，才对入选因子计算一次区间末段「检验期」的表现，并与现有最好的几个因子、两者的等权合成并排比较。可设轮数、每轮因子数、检验期比例、入选门槛和挖掘方向；入选因子可一键保存为自定义因子
• 新增：AI 因子挖掘的可信度检查：列出一共试了多少个因子及相应的多重检验门槛，入选因子是否通过检验期（IC 同号且不低于挖掘期的一半），以及是否胜过现有因子
• AI 因子挖掘中 AI 写的代码会在本机自动运行：运行前做严格检查（不允许 import、读写文件等），并在截到较早日期的数据上重算一遍，用到未来数据的因子不会入选
• 新增：AI 因子挖掘的「导出挖掘记录」（Markdown 文件：设置、两段区间、可信度检查、全部候选的结果与代码、模型每轮的说明）；入选因子表增加「达到门槛」（挖掘期 |t| 达到多重检验门槛）和「通过检验期」两列；判为「覆盖不足」的因子会写明实际覆盖率，下一轮 AI 据此修改代码
• 改进：桌面窗口里的文字（说明、表格、代码等）可以选中复制
• 改进：界面字体统一：标题仍为宋体，正文、表格与普通数字统一为无衬线字体，等宽字体仅用于代码；部分说明与提示改为更正式的表述
• 修复：DeepSeek 等服务繁忙时会一直保持连接而不回复，AI 功能可能卡住很久。现在每次调用 AI 最多等 5 分钟（设置里的超时更长时按设置），超时提示稍后再试；AI 因子挖掘在等待时显示已等秒数，点「停止」立即生效

本软件仅供学习与研究使用，不构成任何投资建议。

---

Download

For a fresh install: SimpleQuant-1.2.0-Setup.exe (Windows), SimpleQuant-1.2.0-macos-arm64.dmg (macOS, Apple silicon) or SimpleQuant-1.2.0-linux-x86_64.tar.gz (Linux). Intel Macs are supported up to 0.2.5. Installed copies (0.1.1 or later) update themselves. The macOS app is not notarized: on first launch, allow it under System Settings → Privacy & Security → Open Anyway. The other assets are used by the auto-updater.

Changes

• New: an "AI factor mining" tab in multi-factor selection: each round the AI proposes a batch of factors, the software evaluates them on the early "mining period" only (IC, t-statistic, correlation with existing factors) and feeds the results back so the AI can refine them or change direction. Only after the last round are the accepted factors measured once on the held-out "test period" at the end of the range, side by side with the best existing factors and equal-weight composites of both. Rounds, factors per round, test share, acceptance thresholds and a search direction can be set; accepted factors can be saved as custom factors with one click
• New: a credibility assessment for factor mining: how many factors were tried and the matching multiple-testing threshold, whether the accepted factors held up in the test period (same sign and at least half of the mining IC), and whether they beat the existing factors
• Factor code written by the AI during mining runs automatically on this computer: it is strictly checked first (no imports, no file access, etc.) and recomputed on data cut off at earlier dates; factors that use future data are rejected
• New: "Export mining record" for factor mining (a Markdown file with the settings, both periods, the credibility assessment, every candidate's results and code, and the model's notes); the accepted-factor table gains "Passes threshold" (mining |t| reaches the multiple-testing threshold) and "Held up in test" columns; factors rejected for low coverage now state their actual coverage, so the AI can fix the code in the next round
• Improved: unified typography: headings stay serif; body text, tables and ordinary numbers use one sans-serif font, and monospace is used only for code. Some descriptions and hints were reworded more formally
• Improved: text in the desktop window (notes, tables, code) can now be selected and copied
• Fixed: services such as DeepSeek keep the connection open without replying when busy, so AI features could hang for a long time. Each AI call now waits at most 5 minutes (or the timeout in Settings if longer) and then asks you to try again later; factor mining shows how long it has been waiting, and Stop takes effect immediately

This software is for learning and research only and is not investment advice.

SimpleQuant 1.1.0

下载

首次安装请点击下载（已安装 0.1.1 及以后版本的程序会自动检测并下载更新）：

系统	安装包
Windows 10 / 11（64 位，无需管理员权限）	SimpleQuant-1.1.0-Setup.exe
macOS 11 及以上（Apple 芯片：M1 / M2 / M3 / M4…）	SimpleQuant-1.1.0-macos-arm64.dmg
Linux（x86_64）	SimpleQuant-1.1.0-linux-x86_64.tar.gz

Intel 芯片的 Mac 最后支持的版本是 0.2.5，已安装的可以继续使用，此后的版本不再提供。

国内下载慢可以用 123 云盘：https://1829763294.share.123pan.cn/123pan/mv2ejv-wbhtA?pwd=ZfDO

macOS 版未经 Apple 公证，第一次打开会被拦下：请在「系统设置 → 隐私与安全性」中点「仍要打开」，或在终端运行 xattr -dr com.apple.quarantine /Applications/SimpleQuant.app。

下方「Assets」中的其余文件（manifest、-from- 增量包）供程序自动更新使用，无需下载。

更新内容

• 新增：条件组件的「交易节奏」：每个大于 / 小于条件可以设「连续 K 线数」（连续 N 根 K 线都满足才成立，可过滤一两天的假突破）；可设「卖出后多少根 K 线内不再买入」；可开启「分批建仓」：仓位分成几份，满足买入条件时先买一份，之后在比上次买入时上涨或下跌一定幅度、或买入条件再次满足时再加一份。这些设置可用于参数优化，回测、模拟盘、方案文件、Python 脚本以及导出到聚宽 / 掘金 / QMT 的结果逐笔一致；用一句话生成策略（AI）时也能识别「连续 3 天」「卖出后 10 天内不买」「越跌越买」等说法。用到这些设置的方案文件需要本版本或更新的版本才能导入
• 新增：估值百分位条件「市盈率历史百分位」「市净率历史百分位」：当前估值在过去 N 年里的位置（0 = 最低，100 = 最高）。AKShare 下载沪深300、中证500、上证50 等常见宽基 ETF / 指数时，附带所跟踪指数的市盈率（TTM）和市净率（乐咕乐股，每月更新），数据库中标为「含指数估值」；BaoStock 股票日线逐日提供。已下载的 ETF / 指数需重新下载才会带上估值
• 新增：回测「可信度检查」的随机买卖对照：把已平仓交易的买卖时机随机打乱 1000 次（笔数、每笔持有时间不变），看策略好于多少比例的随机结果，判断收益来自买卖时机还是持有期间的整体涨跌
• 新增：参数优化结果的「最优参数逐年排名」：最优参数在每个自然年的收益排在所有组合的第几名、当年中位数和最好的组合；可信度检查在最优参数多数年份排不进前一半时警告
• 新增：用一句话生成策略（AI）的结果逐条标出「状态」条件（大于 / 小于，每根 K 线都检查）和「事件」条件（上穿 / 下穿，只在穿越当天成立）并说明区别；描述里说「上穿 / 金叉」但生成的是大于 / 小于（或反过来）时提醒；同一段描述再次生成的规则与上一次不同时，提示描述可能有歧义并列出上一次的规则
• 改进：回测报告注明数据来源（如「东方财富 · 后复权」）；下载时东方财富无法连接、改用新浪财经的，数据页会说明；股票数据改用新浪（等比复权，与东方财富的加法复权口径不同）时，回测与参数优化页会提示
• 改进：设置页不再直接显示本机路径（AI 设置文件、数据目录），改为「打开所在文件夹」按钮，鼠标悬停时显示路径，录屏和截图时不会露出
• 修复：点击「运行回测」后，新结果出来前会闪现上一份报告
• 修复：参数优化图的悬停提示把参数值显示成「-8.00000」这样的 6 位有效数字

本软件仅供学习与研究使用，不构成任何投资建议。

---

Download

For a fresh install: SimpleQuant-1.1.0-Setup.exe (Windows), SimpleQuant-1.1.0-macos-arm64.dmg (macOS, Apple silicon) or SimpleQuant-1.1.0-linux-x86_64.tar.gz (Linux). Intel Macs are supported up to 0.2.5. Installed copies (0.1.1 or later) update themselves. The macOS app is not notarized: on first launch, allow it under System Settings → Privacy & Security → Open Anyway. The other assets are used by the auto-updater.

Changes

• New: "Trade pacing" in condition blocks: each > / < condition can require several bars in a row (filters out one- or two-day false breakouts); a cooldown blocks new buys for N bars after a sell; and "Scale in" splits the position into parts, buying one part when the buy conditions are met and adding one part after the price rises or falls by a set % from the last buy, or when the buy conditions are met again. These settings can be optimized, and backtests, paper trading, plan files, Python scripts and JoinQuant / MyQuant / QMT exports match trade for trade; the AI rule generator understands phrases such as "3 days in a row", "no buying for 10 days after a sell" and "buy more as it falls". Plan files using these settings need this version or later
• New: valuation percentile conditions "P/E percentile" and "P/B percentile": where the current valuation sits within the past N years (0 = lowest, 100 = highest). AKShare downloads of common broad-index ETFs / indices such as CSI 300, CSI 500 and SSE 50 now include the tracked index's P/E (TTM) and P/B (Legulegu, updated monthly), marked "with index valuation" in the library; BaoStock stock data has them daily. Re-download existing ETF / index data to get the valuation
• New: random-timing comparison in the backtest credibility assessment: the timing of the closed trades is shuffled 1,000 times (same number of trades and holding periods) to show what share of random results the strategy beats, i.e. whether returns come from timing or simply from the market move while holding
• New: "Best parameters, year by year" in optimization results: the rank of the best parameters among all combinations in each calendar year, with that year's median and best; the credibility assessment warns when they miss the top half in most years
• New: the AI rule generator now marks each condition as a "state" (> / <, checked on every bar) or an "event" (crosses above / below, true only on the day of the cross) and explains the difference; it warns when the description says "crosses" but the rule uses > / < (or the reverse), and when the same description produces a different rule than last time, showing the previous rule
• Improved: backtest reports state the data source (e.g. "Eastmoney · back-adjusted"); the Data page notes when Eastmoney was unreachable and Sina Finance was used; backtest and optimization pages warn when stock data came from Sina (proportional adjustment, unlike Eastmoney's additive adjustment)
• Improved: Settings no longer shows local paths (AI settings file, data folder) directly; an "Open folder" button shows the path on hover, so it stays out of screen recordings and screenshots
• Fixed: the previous report flashed briefly after clicking Run backtest, before the new results appeared
• Fixed: optimization chart tooltips showed parameter values padded to six significant digits, such as "-8.00000"

This software is for learning and research only and is not investment advice.

SimpleQuant 1.0.0

下载

首次安装请点击下载（已安装 0.1.1 及以后版本的程序会自动检测并下载更新）：

系统	安装包
Windows 10 / 11（64 位，无需管理员权限）	SimpleQuant-1.0.0-Setup.exe
macOS 11 及以上（Apple 芯片：M1 / M2 / M3 / M4…）	SimpleQuant-1.0.0-macos-arm64.dmg
Linux（x86_64）	SimpleQuant-1.0.0-linux-x86_64.tar.gz

Intel 芯片的 Mac 最后支持的版本是 0.2.5，已安装的可以继续使用，此后的版本不再提供。

国内下载慢可以用 123 云盘：https://1829763294.share.123pan.cn/123pan/mv2ejv-wbhtA?pwd=ZfDO

macOS 版未经 Apple 公证，第一次打开会被拦下：请在「系统设置 → 隐私与安全性」中点「仍要打开」，或在终端运行 xattr -dr com.apple.quarantine /Applications/SimpleQuant.app。

下方「Assets」中的其余文件（manifest、-from- 增量包）供程序自动更新使用，无需下载。

更新内容

• 新增：多因子选股的「因子加权方式」增加「机器学习」：以所选因子为特征、未来持有期收益的排名为目标，在每个调仓日只用当时已知的数据训练模型（默认 3 年滚动窗口，最多每 20 个交易日重新训练），以模型的预测值选股。可选岭回归、K 近邻、随机森林、梯度提升树四种模型，均在本机计算，无需额外安装，结果可复现。结果页新增样本外 IC 对比（模型、ICIR 加权与各单因子）和因子重要性；可信度检查提示样本外 IC 不为正、训练集 IC 远高于样本外 IC（过拟合）、模型不如 ICIR 加权等情况。机器学习的参数可用于滚动优化，策略可开设模拟账户、导出方案文件与 Python 脚本（导出到聚宽 / 掘金 / QMT 时请用「按名单调仓」）。含机器学习设置的方案文件需要本版本或更新的版本才能导入
• 修复：东方财富无法连接、ETF 改用新浪财经下载时，复权价没有处理份额拆分与折算（如 510500 在 2015-04-15 单日 +249%、2022-08-29 单日 -13%）。现在按新浪复权表标出的日期和当天价格跳空推算比例；仍有无法解释的跳空时不使用新浪数据并提示稍后重试。此前用新浪下载过拆分过的 ETF 复权数据的，请在数据页重新下载
• 修复：自定义因子或代码策略运行出错时，错误信息里附带的那一行代码有时是另一段代码的（先后试算过几段代码后，较早那段出错时显示了后来那段的同一行号）；行号本身是对的

本软件仅供学习与研究使用，不构成任何投资建议。

---

Download

For a fresh install: SimpleQuant-1.0.0-Setup.exe (Windows), SimpleQuant-1.0.0-macos-arm64.dmg (macOS, Apple silicon) or SimpleQuant-1.0.0-linux-x86_64.tar.gz (Linux). Intel Macs are supported up to 0.2.5. Installed copies (0.1.1 or later) update themselves. The macOS app is not notarized: on first launch, allow it under System Settings → Privacy & Security → Open Anyway. The other assets are used by the auto-updater.

Changes

• New: "Machine learning" as a factor weighting method in stock selection. The chosen factors are features and the rank of the next holding-period return is the target; on every rebalance date a model is trained only on data known at the time (3-year rolling window by default, retrained at most every 20 trading days) and its predictions pick the stocks. Choose ridge regression, k-nearest neighbors, random forest or gradient boosting; everything runs locally with no extra install and is reproducible. Results add an out-of-sample IC comparison (model, ICIR weighting, single factors) and factor importance; the credibility check flags non-positive out-of-sample IC, overfitting (training IC far above out-of-sample IC) and models that do worse than ICIR weighting. Model parameters can be walk-forward optimized, and such strategies can be paper traded and exported as plan files or Python scripts (use "follow picks" for JoinQuant / MyQuant / QMT). Plan files with machine-learning settings need this version or later
• Fixed: when Eastmoney is unreachable and ETF data comes from Sina Finance, adjusted prices ignored share splits (e.g. 510500 showed +249% on 2015-04-15 and -13% on 2022-08-29). Split ratios are now inferred from the price gap on the dates in Sina's adjustment table; if an unexplained gap remains, the Sina data is rejected with a message to retry later. Re-download adjusted ETF data previously fetched from Sina for funds that had splits
• Fixed: when a custom factor or code strategy failed, the source line quoted in the error could come from different code (after several snippets had been run, an error in an earlier one showed the same line number of a later one); the line number itself was right

This software is for learning and research only and is not investment advice.

SimpleQuant 0.3.0

下载

首次安装请点击下载（已安装 0.1.1 及以后版本的程序会自动检测并下载更新）：

系统	安装包
Windows 10 / 11（64 位，无需管理员权限）	SimpleQuant-0.3.0-Setup.exe
macOS 11 及以上（Apple 芯片：M1 / M2 / M3 / M4…）	SimpleQuant-0.3.0-macos-arm64.dmg
Linux（x86_64）	SimpleQuant-0.3.0-linux-x86_64.tar.gz

Intel 芯片的 Mac 最后支持的版本是 0.2.5，已安装的可以继续使用，此后的版本不再提供。

国内下载慢可以用 123 云盘：https://1829763294.share.123pan.cn/123pan/mv2ejv-wbhtA?pwd=ZfDO

macOS 版未经 Apple 公证，第一次打开会被拦下：请在「系统设置 → 隐私与安全性」中点「仍要打开」，或在终端运行 xattr -dr com.apple.quarantine /Applications/SimpleQuant.app。

下方「Assets」中的其余文件（manifest、-from- 增量包）供程序自动更新使用，无需下载。

更新内容

• 不再提供 Intel 芯片 Mac 的安装包：0.2.5 是最后一个支持 Intel Mac 的版本，已安装的可以继续使用。以后的版本在新版本没有本系统安装包时，设置页的「检查更新」会说明情况，不再显示为出错
• 改进：多因子选股回测的「运行日志」显示股票或转债名称（如「搜特转债(128100) 停牌或开盘跌停，无法卖出」），与成交记录一致；此前只显示代码
• 新增：用一句话描述自定义因子（AI）：在选股页「自定义因子」中描述计算思路（如「近 20 天上涨日成交额占总成交额的比例」），由大模型写出 factor(p) 代码，并列出无法用现有数据计算的部分；可勾选「在编辑器中的代码基础上修改」。生成的代码不会自动运行，填入编辑器、阅读后点击「试算」才在本机运行；生成时会检查偷看未来（shift(-n)、bfill 等）、不存在的字段、import 与读写文件，发现问题先让模型修正
• 改进：自定义因子编辑器在代码用到未来数据（shift(-n)、bfill、rolling(center=True)）或不存在的字段时，在编辑器下方提示
• 改进：AI 解读回测结果（择时回测与多因子选股回测）时，除汇总指标外还参考交易明细：最好与最差的几笔交易、按标的的盈亏、去掉最好几笔后的收益与重抽样区间，以及期末未平仓持仓的浮动盈亏（胜率、盈亏比只统计已平仓交易，期末浮亏较大时会指出）；并说明回测已扣除所设的手续费、印花税和滑点
• 改进：回测「可信度检查」新增期末未平仓提示（择时回测、多因子选股回测、参数优化结果）：期末仍持有的仓位按最后收盘价计浮亏达到初始资金的 3% 时警告，并列出已平仓交易的合计收益（胜率与盈亏比只统计已平仓交易）；收益一半以上是尚未兑现的浮盈时也会说明

本软件仅供学习与研究使用，不构成任何投资建议。

---

Download

For a fresh install: SimpleQuant-0.3.0-Setup.exe (Windows), SimpleQuant-0.3.0-macos-arm64.dmg (macOS, Apple silicon) or SimpleQuant-0.3.0-linux-x86_64.tar.gz (Linux). Intel Macs are supported up to 0.2.5. Installed copies (0.1.1 or later) update themselves. The macOS app is not notarized: on first launch, allow it under System Settings → Privacy & Security → Open Anyway. The other assets are used by the auto-updater.

Changes

• Intel Macs are no longer supported: 0.2.5 is the last version built for them, and installed copies keep working. From now on, when a newer release has no package for your system, "Check for updates" in Settings explains this instead of reporting an error
• Improved: the Log tab of the selection backtest shows stock and bond names (as in Fills) instead of codes only
• New: describe a custom factor in plain language (AI): on the Custom factors tab of the Selection page, describe how it is computed (e.g. "the share of the last 20 days' trading value that came on up days") and the model writes the factor(p) code and lists anything the available data cannot compute; tick "Modify the code in the editor" to change existing code. Generated code never runs by itself: it is placed in the editor and runs on this computer only when you click Trial run. Generation checks for look-ahead (shift(-n), bfill and similar), unknown fields, imports and file access, and asks the model to fix them first
• Improved: the custom factor editor warns below the code when it uses future data (shift(-n), bfill, rolling(center=True)) or a field that does not exist
• Improved: the AI explanation of backtest results (timing and multi-factor selection) now looks beyond the summary metrics at the trades: the best and worst trades, P&L by security, the return without the top trades and the bootstrap range, and the unrealized P&L of positions still open at the end (win rate and profit factor cover closed trades only, so a large open loss is pointed out); it also knows the results are already net of the commission, stamp duty and slippage you set
• Improved: the credibility assessment (timing and selection backtests, optimization results) now covers positions still open at the end: it warns when their unrealized loss at the last close reaches 3% of starting capital and shows what the closed trades made (win rate and profit factor cover closed trades only); it also points out when more than half of the return is an unrealized gain

This software is for learning and research only and is not investment advice.