Corrada
FeaturesDownloadBlogSign inGet free API

Corrada

FeaturesDownloadBlogTerms & ConditionsPrivacy PolicyIDE for quant strategies development

Blog

English繁體中文All
  • Aug 13, 2026English

    Recommended usage: treat the graph as research, the backtest as proof

    Corrada is a research copilot, not a strategy factory. Find tickers and relations that actually exist on the graph, write one testable hypothesis, then walk-forward — do not treat a snapshot z as a trade signal.

  • Aug 18, 2026EnglishFree+

    LIT, REMX, XME: PCA residual sleeve, Sharpe 1.18

    Long-only residual mean-reversion on LIT, REMX, XME (battery metals vs diversified miners). Live rule e=−0.5 / x=+0.5: Sharpe 1.18, +482%, max DD -18.5%. Walk-forward OOS Sharpe 0.97.

  • Aug 18, 2026繁體中文Free+

    LIT、REMX、XME:PCA 殘差 sleeve,Sharpe 1.18

    只做多的殘差均值回歸,標的為 LIT、REMX、XME(電池金屬對上綜合礦業)。實盤規則 e=−0.5 / x=+0.5:Sharpe 1.18、+482%、最大回撤 -18.5%。Walk-forward 樣本外 Sharpe 0.97。

  • Aug 18, 2026EnglishFree+

    CPER, COPX, XME: PCA residual sleeve, Sharpe 1.06

    Long-only residual mean-reversion on CPER, COPX, XME (copper futures, copper miners, and metals). Live rule e=−0.5 / x=+0.5: Sharpe 1.06, +487%, max DD -36.4%. Walk-forward OOS Sharpe 0.72.

  • Aug 18, 2026繁體中文Free+

    CPER、COPX、XME:PCA 殘差 sleeve,Sharpe 1.06

    只做多的殘差均值回歸,標的為 CPER、COPX、XME(銅期貨、銅礦與金屬)。實盤規則 e=−0.5 / x=+0.5:Sharpe 1.06、+487%、最大回撤 -36.4%。Walk-forward 樣本外 Sharpe 0.72。

  • Aug 18, 2026EnglishFree+

    CPER, COPX: PCA residual sleeve, Sharpe 0.92

    Long-only residual mean-reversion on CPER, COPX (copper futures versus copper miners). Live rule e=−0.5 / x=+0.5: Sharpe 0.92, +277%, max DD -21.4%. Walk-forward OOS Sharpe 0.38.

  • Aug 18, 2026繁體中文Free+

    CPER、COPX:PCA 殘差 sleeve,Sharpe 0.92

    只做多的殘差均值回歸,標的為 CPER、COPX(銅期貨對上銅礦股)。實盤規則 e=−0.5 / x=+0.5:Sharpe 0.92、+277%、最大回撤 -21.4%。Walk-forward 樣本外 Sharpe 0.38。

  • Aug 18, 2026EnglishFree+

    USO, XLE: PCA residual sleeve, Sharpe 0.91

    Long-only residual mean-reversion on USO, XLE (crude futures versus energy equities). Live rule e=−0.5 / x=+0.5: Sharpe 0.91, +291%, max DD -18.8%. Walk-forward OOS Sharpe 0.63.

  • Aug 18, 2026繁體中文Free+

    USO、XLE:PCA 殘差 sleeve,Sharpe 0.91

    只做多的殘差均值回歸,標的為 USO、XLE(原油期貨對上能源股)。實盤規則 e=−0.5 / x=+0.5:Sharpe 0.91、+291%、最大回撤 -18.8%。Walk-forward 樣本外 Sharpe 0.63。

  • Aug 18, 2026EnglishFree+

    XLE, XOP: PCA residual sleeve, +239%

    Long-only residual mean-reversion on XLE, XOP (integrated energy versus E&P). Live rule e=−0.5 / x=+0.5: Sharpe 0.76, +239%, max DD -30.9%. Walk-forward OOS Sharpe 0.69.

  • Aug 18, 2026繁體中文Free+

    XLE、XOP:PCA 殘差 sleeve,報酬 +239%

    只做多的殘差均值回歸,標的為 XLE、XOP(綜合能源對上上游 E&P)。實盤規則 e=−0.5 / x=+0.5:Sharpe 0.76、+239%、最大回撤 -30.9%。Walk-forward 樣本外 Sharpe 0.69。

  • Aug 18, 2026EnglishFree+

    USO, UGA, XES: PCA residual sleeve, +275%

    Long-only residual mean-reversion on USO, UGA, XES (crude, gasoline, and oilfield services). Live rule e=−0.5 / x=+0.5: Sharpe 0.72, +275%, max DD -38.4%. Walk-forward OOS Sharpe 0.49.

  • Aug 18, 2026繁體中文Free+

    USO、UGA、XES:PCA 殘差 sleeve,報酬 +275%

    只做多的殘差均值回歸,標的為 USO、UGA、XES(原油、汽油與油田服務)。實盤規則 e=−0.5 / x=+0.5:Sharpe 0.72、+275%、最大回撤 -38.4%。Walk-forward 樣本外 Sharpe 0.49。

  • Aug 18, 2026EnglishFree+

    AMLP, XLE: PCA residual sleeve, +116%

    Long-only residual mean-reversion on AMLP, XLE (midstream MLPs versus broad energy). Live rule e=−0.5 / x=+0.5: Sharpe 0.71, +116%, max DD -23.3%. Walk-forward OOS Sharpe 0.60.

  • Aug 18, 2026繁體中文Free+

    AMLP、XLE:PCA 殘差 sleeve,報酬 +116%

    只做多的殘差均值回歸,標的為 AMLP、XLE(中游 MLP 對上廣泛能源)。實盤規則 e=−0.5 / x=+0.5:Sharpe 0.71、+116%、最大回撤 -23.3%。Walk-forward 樣本外 Sharpe 0.60。

  • Aug 18, 2026EnglishFree+

    LIT, REMX: PCA residual sleeve, +139%

    Long-only residual mean-reversion on LIT, REMX (lithium versus rare earths). Live rule e=−0.5 / x=+0.5: Sharpe 0.65, +139%, max DD -23.9%. Walk-forward OOS Sharpe 0.67.

  • Aug 18, 2026繁體中文Free+

    LIT、REMX:PCA 殘差 sleeve,報酬 +139%

    只做多的殘差均值回歸,標的為 LIT、REMX(鋰對上稀土)。實盤規則 e=−0.5 / x=+0.5:Sharpe 0.65、+139%、最大回撤 -23.9%。Walk-forward 樣本外 Sharpe 0.67。

  • Aug 18, 2026EnglishFree+

    PCA residual method: from PCA to OLS, z-score, and the entry signal

    Step-by-step calculation used by the Commodity book: log returns, 126-day PCA, OLS residual, z-score, regime gate, and the −0.5 / +0.5 entry.

  • Aug 18, 2026繁體中文Free+

    PCA 殘差方法:從主成分、OLS、z-score 到進場訊號

    商品簿實際在算的步驟:對數報酬、126 日 PCA、OLS 殘差、z-score、體制閘門,以及 −0.5 / +0.5 進場。

  • Aug 18, 2026English

    Use Hermes as your trading research AI: install, register Corrada, connect MCP

    Install Hermes, create a free Corrada account, copy the MCP command, then ask about strategy, pairs, filings, and market structure.

  • Aug 18, 2026繁體中文

    用 Hermes 當交易研究 AI:安裝、註冊 Corrada、接上 MCP

    安裝 Hermes、註冊免費 Corrada 帳號、複製 MCP 指令接上,就能開始問策略、配對、財報與市場結構。

  • Aug 17, 2026English

    If You Don't Know Who Is Losing Money To You, You Are The Yield

    Every profitable trade has a counterparty error. Name it — or you are someone else's coupon.

  • Aug 10, 2026English

    The Expiration Date of Alpha: Why Momentum Dies After 12 Months

    Long-horizon winners become tomorrow's exit liquidity. We trade 10-day relative impulses — not 12-month trophies.

  • Aug 10, 2026EnglishFree+

    Gold Still Held Up — Miners Already Refused to Confirm

    Your miners sleeve just flashed a long — and equity investors were already refusing to pay for gold exposure through stocks.

  • Aug 10, 2026繁體中文Free+

    黃金還撐著——礦商已經拒絕確認

    你等黃金圖表。GLD 還撐著,避險頭條還在流,於是買礦商——GDX 卻已相對金屬跑輸到最低十分位好幾週。

  • Aug 9, 2026EnglishFree+

    Factories Still Looked Busy — Ships Were Already Empty

    Your metals sleeve just printed a long — and dry-bulk freight had already collapsed relative to equities.

  • Aug 9, 2026繁體中文Free+

    工廠看來還忙——船已經空了

    你等生產圖表。工廠頭條聽起來還忙,銅現貨圖表還撐著,於是進場——波羅的海乾散貨運價卻已墜崖好幾週。

  • Aug 8, 2026EnglishFree+

    Retail Looked Resilient — Groceries Were Already Squeezing

    Your discretionary sleeve just flashed a buy — and the grocery bill had already started stealing the marginal dollar.

  • Aug 8, 2026繁體中文Free+

    零售看來有韌性——菜籃子已經在擠壓

    你等零售圖表。同店銷售標題還算好看,消費者「看來有韌性」,於是進場——農產品期貨卻已相對股市跑贏到最高十分位好幾天。

  • Aug 7, 2026EnglishFree+

    The Story Still Sounded Good — Lottery Tickets Were Being Sold

    Your speculative sleeve just printed a long — and the market was already liquidating lottery tickets.

  • Aug 7, 2026繁體中文Free+

    故事聽起來還好——彩券部位正在被賣

    你等圖表。論點聽起來仍有力,散戶選擇權活動在頭條股上依舊熱絡,於是進場——無營收生技相對防禦性醫療卻已經失血好幾週。

  • Aug 6, 2026EnglishFree+

    Growth Still Looked Cheap — Inflation Was Already Repricing

    Your long-duration sleeve just flashed a buy — and the bond market had already repriced inflation upward.

  • Aug 6, 2026繁體中文Free+

    成長看來還便宜——通膨已經在重定價

    你等股票圖表。營收還在印、本益比在上次回檔後仍顯得「合理」,於是進場——損益兩平通膨率(breakeven inflation)卻已飆升好幾天。

  • Aug 5, 2026EnglishFree+

    Risk Looked Crowded — The Funding Currency Snapped Back

    The yen can jump 3% in a week and every high-beta sleeve you own suddenly trades like one trade.

  • Aug 5, 2026繁體中文Free+

    風險看來很擁擠——融資貨幣已經彈回

    日圓一週可以跳 3%,你持有的每條高 beta sleeve 突然像同一筆交易在跌。

  • Aug 4, 2026EnglishFree+

    The Index Made Highs — Most Stocks Already Didn't

    SPY can print a fresh high while the average stock is already falling — and your sleeve trades the average, not the headline.

  • Aug 4, 2026繁體中文Free+

    指數創新高——多數股票早已沒份

    SPY 可以創新高,平均股票卻已在跌——而你的 sleeve 交易的是平均參與者,不是頭條指數。

  • Aug 3, 2026English

    Your Backtest Looks Perfect. That’s Exactly Why You Shouldn’t Trust It.

    Most strategies don’t die in live markets because the idea was stupid. They die because the backtest was a flattering mirror — and nobody checked whether the reflection survived contact with the future.

  • Aug 3, 2026EnglishFree+

    Travel Demand Looked Fine — Fuel Was Eating Margins

    Airline momentum can look perfect right before jet fuel turns the P&L inside out for two quarters.

  • Aug 3, 2026繁體中文Free+

    旅遊需求看來還好——燃油正在吃掉利潤

    旅遊與航空 sleeve 感覺像純需求交易:訂位強、重啟順風、圖表叫你上車。但航空燃油佔航空公司營運成本 20–30%——美國多數業者已放棄大規模避險。因提前訂票,票價調整落後油價數月。

  • Aug 2, 2026EnglishFree+

    The Cycle Looked Intact — Builders Already Knew

    Housing breaks the business cycle first — and homebuilder equities confess before the permits data prints.

  • Aug 2, 2026繁體中文Free+

    景氣看來還完整——建商已經知道

    綁定房市的利率敏感 sleeve,在房貸利率看似穩定、大盤指數還在磨高時,感覺很合理。但房市不是普通週期板塊。Leamer 那句話仍然成立:房市就是景氣循環——實體經濟中對利率最敏感、經信用管道傳導最強的板塊。

  • Aug 1, 2026EnglishFree+

    Copper Still Looked Bid — China Was Already Selling

    The marginal buyer of half the world's copper can go quiet weeks before the chart confesses.

  • Aug 1, 2026繁體中文Free+

    銅看來還有買盤——中國已經在賣

    銅的 sleeve 在現貨看起來還強、頭條還在談電網資本支出時,感覺最安全——也正是最貴的錯誤藏得最深的地方:你做多一種金屬,而決定邊際價格的買家,早已開始賣掉那條「折現需求」的股票。

  • Jul 31, 2026EnglishFree+

    The Core Held — The Periphery Was Already Sold

    Your commodity-producer long can break out on domestic momentum while global capital is already fleeing the riskiest geography bucket.

  • Jul 31, 2026繁體中文Free+

    核心還撐著——外圍已經被賣

    你的大宗商品生產商多單可以因本土動能突破,全球資金卻已經在撤出風險最高的地理籃子。

  • Jul 30, 2026EnglishFree+

    Big Banks Looked Fine — Regionals Were Already Cracking

    Your financials long can still pass every chart filter while the domestic credit system's canary is already dead.

  • Jul 30, 2026繁體中文Free+

    大銀行看來還好——區域銀行已經在裂

    金融 sleeve——XLF、KBE——以及往來銀行是區域銀行的小型股(IWM 類)靠相對強度訊號進場。商用不動產(VNQ)看起來像另一個資產類別。但區域銀行一裂,傷害是透過信用收縮擴散,不只是情緒。XLF 裡的大型銀行、保險、交易所還撐著,KRE 卻在流血——你的進場訊號看不到這道分裂。

  • Jul 29, 2026EnglishFree+

    Energy Equities Signaled Buy — Crude Was Collapsing

    Your E&P sleeve can flash a perfect relative-strength entry while the commodity that pays the bills is in freefall.

  • Jul 29, 2026繁體中文Free+

    能源股發出買訊——原油正在崩跌

    你的 E&P sleeve 可以閃出完美的相對強度進場,付帳單的大宗商品卻已在自由落體。

  • Jul 28, 2026EnglishFree+

    Industrials Printed Gains — Freight Didn't Confirm

    Your infrastructure long can break out on earnings momentum while the trucks that move the goods are already pricing a slowdown.

  • Jul 28, 2026繁體中文Free+

    工業股印出漲幅——貨運沒有確認

    基建與工業 sleeve——PAVE、XLI——在資本支出敘事轉熱時進場。CARZ 這類機械相關標的吃交付週期動能。航空(JETS)吃旅遊需求。圖表說工業板塊很強。但如果搬貨的公司不確認,工業股的上漲就可疑。

  • Jul 27, 2026English

    Hysteresis Bands: An Engineering Weapon Against Market Noise

    Hysteresis stops threshold chattering with dual entry/exit bands that kill commission bleed.

  • Jul 27, 2026EnglishFree+

    Tech Still Looked Strong — Semis Already Quit

    Your cybersecurity breakout can print green while the sector's own engine is already stalling.

  • Jul 27, 2026繁體中文Free+

    科技看來還強——半導體已經先退

    PANW 的一根紅 K 不是供應鏈投下的信任票——是半導體早已定價的支出週期落後讀數。

  • Jul 26, 2026EnglishFree+

    The Dip Looked Buyable — Long Rates Were Spiking

    The utility chart offered a clean pullback — while long-duration bonds were having one of their worst two-week stretches in years.

  • Jul 26, 2026繁體中文Free+

    回檔看起來可買——長端利率正在急升

    你在 XLU、VNQ、TAN 或建商上買回檔,因為股票回撤看起來有序。基本面聽起來還好,股息率也誘人。

  • Jul 25, 2026EnglishFree+

    Risk Assets Held Up — The Dollar Was Squeezing

    Your commodity sleeve still looked tradeable — while a dollar squeeze was already cutting the price of everything it sells.

  • Jul 25, 2026繁體中文Free+

    風險資產還撐著——美元已在擠壓

    你的大宗商品 sleeve 看起來還能交易——美元擠壓卻已在砍它賣的每一樣東西的價格。

  • Jul 24, 2026EnglishFree+

    The Human-Activity Indicator You Must Know

    PMI is still a month away — and industrial demand already rolled over in the copper/gold ratio.

  • Jul 24, 2026繁體中文Free+

    人類活動指標——你一定要知

    你等數據。下一季資本支出、調查、GDP 標題。同時因為股票走勢還算正面,就進場工業、礦業或新興市場出口國 sleeve。

  • Jul 23, 2026EnglishFree+

    The Index Looked Calm — Money Was Already Hiding

    The S&P still felt fine — while institutional money was already selling your sleeve to buy toothpaste.

  • Jul 23, 2026繁體中文Free+

    指數看來平靜——資金已經在躲

    你看指數。走平,甚至收紅。你的非必需消費或主題 sleeve 出現教科書級做多形態,於是進場。

  • Jul 22, 2026EnglishFree+

    Equities Still Hoped — Credit Already Knew

    Your sleeve just printed a fresh long signal — and the junk-bond market had already priced the pain.

  • Jul 22, 2026繁體中文Free+

    股票還在希望——信用已經知道

    你等股票圖表確認。財報標題還算好看,指數撐得住,於是進場——高收益利差卻已經擴大好幾天。

  • Jul 20, 2026English

    The Log-Ratio Scale: How to Calculate Outperformance

    In relative-value or statistical-arbitrage work, the first core problem is this: how do you measure, precisely and without bias, whether asset A is truly outperforming asset B?

  • Jul 13, 2026English

    The Chart Showed a Breakout. You Were Providing Liquidity.

    Discretionary trading makes it dangerously easy to get lost in flickering candles: a big green bar above a moving average triggers FOMO; a long lower wick looks like iron-clad proof.

  • Jul 6, 2026English

    Most Quant Dreams Die in the Wrong Lane

    You can write perfect code in a dead business. Infrastructure and capital decide which games a solo desk can survive.