FTAlphavilleQuantinvestingFT阿爾法城量化投資It’sbeenquitetheyearforsystematicinvestors對(duì)于系統(tǒng)化投資者而言，這是相當(dāng)不平凡的—年RobinWigglesworth2025年"滾雷式"量化危機(jī)內(nèi)幕羅賓·威格爾斯沃思Published發(fā)布UnlocktheEditor’sDigestforfree免費(fèi)解鎖編輯精選weeklynewsletter.Signup注冊(cè)英國(guó)《金融時(shí)報(bào)》主編魯拉·Signup注冊(cè)EnteryouremailaddressInOctober,RenaissanceTechnologiessufferedanightmare,withitstwopublichedgefundsabruptlylosingabout15percent,beforesoaringbackinNovember.Yetforalotofotherbigquantitativeinvestors,theautumnwasmostlyplainsailing.Why?他大型量化投資者來說，這個(gè)秋季大體上風(fēng)平浪靜。為何會(huì)出現(xiàn)這種差異？Wedon’tknowhowRenTech’sfabledemployee-onlyMedallionfundhasfared,butforitsRIEFandRIDAvehicles—whichcollectivelymanagemorethan$20bn—Octoberwasamongtheirworstmonthsever.Bothfundsreboundedbydoubledigitslastmonth,butarestillheadingforanotherpooryear.我們無從得知文藝復(fù)興旗下傳奇的員工專屬基金——大獎(jiǎng)?wù)禄鹬换鹕显聦?shí)現(xiàn)兩位數(shù)反彈，但今年仍可能以慘淡收?qǐng)?。Thishasbeenatopsy-turvyyearformanyinvestors,butRenTech’sautumnalwoesencapsulatehowithasbeenanexceptionallyturbulentyearforquantitativeones.Hedgefundsthatrelyonsophisticatedmodellingandsystematictradinghavebeenrattledbyaseriesofmini-crisesthatinsiderssayhavebeenmildlyreminiscentoftheviolent“quantquake”thatrattledtheentireindustryinAugust2007.2025年對(duì)許多投資者而言堪稱波詭云譎的—年，而文藝復(fù)興的秋日困局恰恰折射出量化投資領(lǐng)域遭遇的空前動(dòng)蕩。依賴復(fù)雜模Tobeclear,thishasonthewholenotbeenabadyearformostquants.Ithasjustbeenriddledwithanunusuallylargenumberofthesemysterious,oftenunsettling,quanttremors.Andtothebafflementofevenmanyindustryinsiders,theimpacthasoftenbeenextremelyvaried,withsomefirmshardlynoticingcertainquiversbutgettingseverelyrattledbyotherones.必須明確指出，對(duì)多數(shù)量化從業(yè)者而言，2025年整體并非糟糕年份。只是這—年異常頻繁地爆發(fā)了諸多神秘且令人不安的量化震蕩。更令業(yè)內(nèi)人士困惑的是，這些震蕩帶來的影響往往天差地別——某些公司對(duì)特定波動(dòng)毫無察覺，卻會(huì)因其他震蕩遭受重Theresulthasbeenanunusuallylargedivergenceintheperformanceofmanyquantstrategies,accordingtoindustryexecutives.AsPhilippeJordan,presidentofCFM,toldAlphaville:據(jù)行業(yè)高管透露，這導(dǎo)致眾多量化策略的業(yè)績(jī)表現(xiàn)出現(xiàn)罕見分化。正如CFM總裁菲利普·喬丹向Alphaville所述：It’snotbeenofthemagnitudeofthe2007quantquake,butit’sbeenlikerollingthunderthroughouttheyear.It’srolledthroughdifferentpeopleatdifferenttimesandindifferentways.雖然規(guī)模不及2007年的量化地震，但全年都像持續(xù)滾動(dòng)的雷暴。它以不同方式在不同時(shí)段沖擊著不同群體。Intriguingly,Alphavillehasheardmyriadpossibleexplanationsforwhythesetremorsseemtohavehadsuchadisparateimpactevenonsimilarquantstrategies,butnooneseemstoknowforcertain.有趣的是，阿爾法城聽聞了無數(shù)種解釋，試圖說明這些震蕩為何甚至對(duì)相似的量化策略造成如此迥異的影響，但似乎無人能給出確鑿答案。Theobviousquestioniswhethertheseidiosyncratic,briefquanttremorsarejustrandomorsignifysomethingbigbuildingupunderthesurface.SoAlphavillethoughtitmightbeworthtakingalookattheissue.Wedon’thaveanygoodanswers,unfortunately,buthere’swhatwefoundout.顯而易見的問題是，這些獨(dú)特而短暫的數(shù)量化市場(chǎng)震蕩究竟是隨機(jī)波動(dòng)，還是預(yù)示著潛藏的更大危機(jī)。因此Alphaville認(rèn)為有必是我們的發(fā)現(xiàn)。FromDeepSeekmageddontoMinVolpocalyse從深度求索末日到最小波動(dòng)率災(zāi)難Whiletheexactnatureofthe2025quanttremorsissomewhatopaque,thecausesaremostlyobviousandinseveralcaseshavealsorattledtraditionalinvestors.盡管2025年量化震蕩的確切性質(zhì)尚不透明，但其誘因大多顯而易見，且在多個(gè)案例中同樣沖擊了傳統(tǒng)投資者。ThefirstonecameinJanuary,whenaChinesecompanycalledDeepSeekunveiledanAIchatbotthatwasfarcheapertobuildandmoreefficientthanmanyUSrivals.ThisbrieflyhammeredthesharesofmanyUStechnologycompaniesandfizzledcertainsystematictradingmodelsintheprocess.首輪沖擊出現(xiàn)在—月份，當(dāng)時(shí)中國(guó)企業(yè)深度求索(D的人工智能聊天機(jī)器人，其開發(fā)成本遠(yuǎn)低于多數(shù)美國(guó)競(jìng)爭(zhēng)對(duì)手且過程中導(dǎo)致某些系統(tǒng)性交易策略失效。Somequanthedgefunds—suchasAQR’slong-shortDelphifund,RenTech’sRIDAandRIEF,andCapitalFundManagement’sStratus—enjoyedhealthyreturnsinJanuary,butotherswerehurtbytheturbulence.TheDeepSeekshockseemedtohavestartedaparticularlybadspellfortrend-followingquanthedgefunds,suchasSystematica’sBlueTrendandManAHL’sAlpha.部分量化對(duì)沖基金——如AQR的多空德爾斐基金、文藝復(fù)興收獲了穩(wěn)健收益，但其他基金則因市場(chǎng)動(dòng)蕩受到?jīng)_擊。深尋沖擊似乎為趨勢(shì)跟蹤型量化對(duì)沖基金開啟了特別糟糕的時(shí)期，例如系統(tǒng)化資本的藍(lán)趨勢(shì)和曼氏AHL的阿爾法基金。InFebruary,manyquanthedgefundswererattledbyasuddenburstofamoremysterious,under-the-surfacestockmarketchaosthatwaslargelyunnoticedbytraditionalinvestors.股市混亂突然爆發(fā)所震動(dòng)，而傳統(tǒng)投資者大多未能察覺這場(chǎng)動(dòng)ItlateremergedthattwoinvestmentteamsatMillenniumthatspecialiseinindexrebalancing—bettingonandagainststocksthatareexpectedtobepromotedorrelegatedfrommajorbenchmarks—hadlostabout$900mn.Thesubsequentdeleveragingrippledthroughseveralcornersofthestockmarketanddingedahostofquantstrategiesintheprocess.后續(xù)調(diào)查顯示，千禧年資本旗下專攻指數(shù)再平衡策略的兩個(gè)投資團(tuán)隊(duì)——通過押注預(yù)計(jì)將納入或剔除出主要基準(zhǔn)指數(shù)的股票進(jìn)行多空操作——損失了約9億美元。隨之而來的去杠桿化浪潮了股市多個(gè)領(lǐng)域，并在此過程中沖擊了大量量化策略。ThencamePresidentDonaldTrump’s“l(fā)iberationday”tarifftantrum.Thiswasabigdealforinvestorsofallstripes,buttheabruptnessofthecrash—andtherapidrecoverywhentheUSadministrationpauseditstariffs—wreakedhavoconalotofquantmodels.Onceagain,itwasparticularlypainfulforquantsthatspecialisedinsurfingtrends,buttovaryingdegrees.是重大事件，但市場(chǎng)暴跌的突然性——以及美國(guó)政府暫停關(guān)稅后的迅速復(fù)蘇——給大量量化模型造成了嚴(yán)重破壞。專注于趨勢(shì)跟蹤的量化策略再度遭受重創(chuàng)，只是程度各有不同。Systematica’sBlueTrendandManAHL’sAlphafundslost8.8percentand4.3percentrespectivelyinApril,butAQR’sManagedFuturesfundonlydropped1.3percent,andAQR’sHelixfund—whichsystematicallyfollowstrendsinmoreexoticmarkets—returned3.9percent.Someotherquantstrategiesdidfine.BothofRenaissance’stwopublichedgefundswerehitbytheimmediateturmoilbutendedthemonthwithpositivegains,forexample.了3.9%的正收益。部分其他量化策略表現(xiàn)尚可：例如文藝復(fù)興旗下的兩只公開對(duì)沖基金雖受短期震蕩波及，但月末仍以正收益收Overthesummer,manyquantswerestruckbyanotherpuzzlingboutofturbulencethatlargelyseemedtopasstraditionalinvestorsby.夏季期間，許多量化機(jī)構(gòu)遭遇了另—波令傳統(tǒng)投資者幾乎毫無察覺的詭異市場(chǎng)波動(dòng)。ThisparticulartremorappearstohavestartedinJuneasaretail-drivenrallyin“garbagestocks”,whichsqueezedshortpositionsatsomequanthedgefunds.Thatinturnappearstohaveledtoabroaderdeleveraginginadjacentcornersofthestockmarketthathurteveryonewithsimilarstrategies.AsMSCIconcludedinitspostmortem:而擠壓了部分量化對(duì)沖基金的空頭倉位。這似乎又引發(fā)了股市相鄰領(lǐng)域更大范圍的去杠桿化，使采用相似策略的機(jī)構(gòu)普遍受損。正如MSCI在事后分析報(bào)告中總結(jié)：Giventhebullishmarket,itisn’tsurprisingthatthestockswithhighbeta,volatility,tradingactivityorlargemarket-capperformedwell.However,anomalously,momentumstocksdidnotdowell.Heavilyshortedstocksalsooutperformed,aclearheadwindfortheshortlegofthesesystematicfunds,and,consistentwiththeideaofa“junk”rally,profitablestocksalsounderperformed.躍度或大市值股票表現(xiàn)優(yōu)異并不令人意外。但反常的是，動(dòng)量股卻表現(xiàn)不佳。被大量做空的股票反而跑贏大盤，這顯然對(duì)這些系統(tǒng)化基金的空頭盈利股票也表現(xiàn)遜色。...Wefoundclearevidencethatthisperformancedragwasgreatlymagnifiedbylargeinteractioneffectsamongthesesamefactorsandbyapartialcrowdingunwind.......我們發(fā)現(xiàn)確鑿證據(jù)表明，這些因子間強(qiáng)烈的交互效應(yīng)，加上部分擁擠交易的平倉，極大放大了這次業(yè)績(jī)拖累的幅度。Whatstartedasaseriesofsmallbutalmostdaily,grindinglosseseventuallymorphedintosomethingabitmoreunsettling.Somequantssaidthatbymid-Julyitwaseerilyreminiscentofamajorinvestorhurriedlyliquidatingalargeportfolio,inanechooftheproximatecauseofthe2007quantquake.“Itfeltlikeabodywashittingthemarket,”onequantfundmanagerrecalls.起初只是—連串微小但幾乎每天發(fā)生的漸進(jìn)虧損，最終演變成了詭異的熟悉感讓人想起2007年量化地震的直接誘因——某位大投資者倉促拋售巨額投資組合的情形。"仿佛有具尸體砸向了市Nobodiessubsequentlyemerged,butthepainwasfairlywidespread.AQR’sDelphiandApexfundssufferedtheirworstmonthsofwhatwasotherwiseastrongyearinJuly,andRenTech’stwopublicfundssufferedtheirfirstseveresetbacksof2025.OtherfundshurtbythesummerquantquakereportedlyincludedPoint72’sCubistunitandhigh-flyingQubeResearch&Technologies.Aurum,ahedgefundinvestor,estimatesthat“statisticalarbitrage”fundshadaparticularlybadJuly,losing2.9percentonaverage.雖未出現(xiàn)重大損失，但傷痛卻相當(dāng)普遍。AQR旗下德爾斐基金和藝復(fù)興科技的兩支公開基金也迎來了年內(nèi)首次重挫。知情人士透露，在這場(chǎng)夏季量化地震中受傷的還包括Point72旗下的立方體部門與風(fēng)頭正勁的Qube研究技術(shù)公司。對(duì)沖基金投資機(jī)構(gòu)奧勒Interestingly,itlookslikethesummermini-quakemarkedanadirfortrend-followingquanthedgefunds.MosttooksomehitsinJuly,buttheyweremostlyveryminor,andbothAugustandSeptemberwereexceptionallygoodforahostofthem.Systematica’sBlueTrendreturned9percentinSeptemberalone,andManAHL’sAlphafundhasnoweraseditsearlierlosses.有趣的是，2025年夏季的小型量化地震似乎標(biāo)志著趨勢(shì)跟蹤型ManAHL的Alpha基金也已收復(fù)了此前的虧損。GregBond,ManGroup’schiefinvestmentofficer,toldAlphavillethattheyearhadbeenaturbulentoneforquants,butstressedthatithadmostlynotbeenaviolentone.化基金而言可謂波瀾起伏，但他強(qiáng)調(diào)大多數(shù)情況下并非劇烈動(dòng)Therehavebeenfourtofiveofthesesmallquantquakesthisyear,withpocketsofdeleveraging.Butmarketshavelargelyabsorbedthem.Idiosyncraticriskhasjustbecomeanewfactor.今年以來已發(fā)生四至五次這類小型量化地震，伴隨著局部去杠桿化現(xiàn)象。但市場(chǎng)基本消化了這些沖擊，特異性風(fēng)險(xiǎn)已然成為新的影響因素。ThelatestquanttremorcameinOctober.Thistimethecausewassomewhatmystifying,thenaturehardtonaildownandtheimpactonvariousquanthedgefundsextremelyvaried.ButRenTech’sRIEFandRIDAfundswereveryclearlythebiggestvictims.界定，對(duì)不同量化對(duì)沖基金的影響也差異懸殊。但文藝復(fù)興科技的RIEF和RIDA基金確鑿無疑成為最大輸家。The14.4percentlosssufferedbytheroughly$18bnRenaissanceInstitutionalEquitiesFundinOctoberwasitsbiggestmonthlylossinmorethanadecade,surpassingeventheblowsdealtbytheCovid-19pandemicin2020.14.4%的虧損，創(chuàng)下十余年來最大單月跌新冠疫情造成的沖擊。RivalquantsthathavestudiedthepatternofRenTech’sreturnssaythatRIEFappearstobemostlytiltedtowardstwostockmarketcharacteristicsthatwerebothpunishedinOctober:“quality”and“l(fā)owvolatility”.研究過文藝復(fù)興科技回報(bào)模式的競(jìng)爭(zhēng)對(duì)手們表示，RIEF基金似量"和"低波動(dòng)率"。Bothareso-calledinvestmentfactorsthatovertimearesupposedtodelivermodestlymarket-beatingreturns.Inhedgefundtermsitinvolvesgoinglongstocksofcompaniesthatscorehighlyonvariousmeasuresofsteadinessandshortthosethatscorebadlyonthesamemetrics.Renaissancethenprobablyusesleveragetobringuptheportfolio’svolatility(sincegoinglonglow-volatilitystocksandshorthigh-volstockswouldpushthefund’soverallchoppinessbelowanoptimallevel).這兩個(gè)所謂的投資因子長(zhǎng)期來看應(yīng)能提供略勝市場(chǎng)的回報(bào)。在對(duì)沖基金領(lǐng)域，這種策略意味著做空各項(xiàng)穩(wěn)定性指標(biāo)得分較低的企業(yè)股票。文藝復(fù)興科技很可能是通過使用杠桿來提高投資組合的波動(dòng)性（因?yàn)樽龆嗟筒▌?dòng)率股票同時(shí)做空高波動(dòng)率股票會(huì)導(dǎo)致基However,neitherqualitynorlow-volatilityfactorsdidthatbadlyinOctober,atleastjudgingbyJPMorgan’smonthlyround-upoftheperformanceofvariousfactors.WolfeResearch’smeasuresofqualityandlow-volatility—or“l(fā)owbeta”asitcallsthefactor—weregrimmer,butMaytoJunewereactuallyfarworsethanSeptembertoOctober.和低波動(dòng)率因子的表現(xiàn)并非特別糟糕。沃爾夫研究公司追蹤的質(zhì)ThesuddencollapseandsubsequentreboundofRenTech’sfundshavethereforebaffledalotofotherquantsandcometoexemplifywhataweirdyearithasbeen.AsatopexecutiveatalargeUSquantitativehedgefundtoldAlphaville:的情況令許多量化同行困惑不已，堪稱今年詭異行情的最佳寫照。正如某美國(guó)大型量化對(duì)沖基金高管向Alphaville透露的：Thingshavebeenupanddownallyear...We’vehadaseriesofquantcrises,butwithverydifferentoutcomesfordifferentpeopleatdifferenttimes.全年行情猶如過山車……我們經(jīng)歷了—連串量化危機(jī)，但不同公司在不同時(shí)點(diǎn)的結(jié)局天差地別。Renaissancedeclinedtocomment.文藝復(fù)興科技公司拒絕置評(píng)。So,umm,what’sup?Firstofall,weshouldnotethatnoteveryoneagreesthat2025hasbeenayearofunusuallyhighdispersionfortheperformanceofvariousquantitativeinvestingstrategies.首先需要指出的是，并非所有人都認(rèn)同2025年量化投資策略出現(xiàn)了異常嚴(yán)重的業(yè)績(jī)分化現(xiàn)象。Somearguethatreturndifferencesarestillmostlyexplainedbyhowdifferentstrategieshaveonthewholeperformed—forexample,systematictrend-followinghadaterriblefirst-halfyearbuthassinceenjoyedabounce—andsubtlenuanceswithinstrategies.有觀點(diǎn)認(rèn)為，收益差異主要還是源于不同策略的整體表現(xiàn)（例如系統(tǒng)性趨勢(shì)跟蹤策略上半年表現(xiàn)極差但隨后反彈）以及策略內(nèi)部Take“statistical-arbitrage”.Somestat-arbstrategiesareveryfocusedonlargeUSstocks,whileothersventureintosmallstocksorevenemergingmarkets.Withinsystematictrend-following,somefirmsspecialiseinlonger-termmomentum,whileothersmostlyexploitmorefleetingtrends.以"統(tǒng)計(jì)套利"策略為例。部分統(tǒng)計(jì)套利策略高度聚焦美國(guó)大盤域，有些公司專攻長(zhǎng)期動(dòng)量交易，而其他機(jī)構(gòu)則主要捕捉短期趨However,everyoneAlphavillespoketoagreedthat2025hasbeencharacterisedbyanunusuallylargenumberofmodestbutunpleasantandunmistakablequanttremors.AsaseniorquantexecutiveatamajorUShedgefundtoldus:然而所有受訪者都承認(rèn)，2025年的顯著特征在于接連出現(xiàn)了多場(chǎng)規(guī)模不大卻令人困擾、可明確識(shí)別的"量化震顫大型對(duì)沖基金資深量化高管向我們透露的那樣：Asynchronouslossesaregreat.Ifeveryoneisdoingthesamethenit’saworry.Butconditionalonstrategiesthereisn’tmuchdispersion...There’sbeenperformancedispersionbetweendifferentquantclusters,butifyoulookateachofthemthereisn’tactuallymuch.非同步虧損是件好事。若所有人都在做同樣的事，那才令人擔(dān)憂。但就具體策略而言，差異其實(shí)并不顯著……不同量化集群間的業(yè)績(jī)確實(shí)存在分化，但細(xì)看每個(gè)集群內(nèi)部其實(shí)差異不大。Weprobablygotspoiltinthelasttwotothreeyears,whenyoucouldspitonthegroundandmakemoney,butthisyearithasbeenharderforalotofquantstrategies.過去兩三年我們或許過得太順?biāo)炝?，那時(shí)隨便往地上吐口唾沫都能賺錢，但今年許多量化策略都舉步維艱。Investmentfactorsareanimperfectproxyforabroadanddiverseindustry,butit’stellingthatnineoutofWolfeResearch’s16mostwidelyfollowedquantfactorshavelostmoneythisyear,andalmostallofthemhavebeenfarmorevolatilethanusual.投資因子雖不能完美代表這個(gè)廣闊多元的行業(yè)，但頗具說服力的是：Wolfe研究機(jī)構(gòu)最受關(guān)注的16出現(xiàn)虧損，且?guī)缀跛幸蜃拥牟▌?dòng)幅度都遠(yuǎn)超往常。YinLuo,headofquantitativeresearchatWolfeResearch,pointsoutthatvalue,short-termreversal—adecentproxyforstat-arb—qualityandlow-vol/lowbetawerecomfortablytheworstperformersin2025.“Thisyearhasbeendominatedbyperiodicrisk-onandrisk-off,withmultiplesell-offsand‘junk’rallies,whichtriggeredthesemini-quantcrises,”hesays.年顯然表現(xiàn)墊底。"今年市場(chǎng)呈現(xiàn)周期性風(fēng)險(xiǎn)偏好與避險(xiǎn)情緒交替，多次拋售與'垃圾股'反彈并行，從而引發(fā)了這些小型量化危However,ifweacceptthat2025didleadtoabizarreamountofdivergenceintheperformanceofevensimilarquantstrategies—asmostindustryinsiderssay—whatcausedit?Whatcouldpossiblymaketrend-following,statisticalarbitrage,multi-factorquantorsystematiclong-shortmarket-neutralinvestingtogohaywireatdifferenttimesandindifferentwaysfordifferentpeople?但如果我們認(rèn)可2025年確實(shí)導(dǎo)致連相似量化策略都出現(xiàn)異常業(yè)績(jī)分化——正如多數(shù)業(yè)內(nèi)人士所言——那成因何在？究竟是什么能略，在不同時(shí)期以不同方式對(duì)不同人群集體失控？FTAlphavillemostlyheardtwopossibleexplanations:FTAlphaville聽到的主要有兩種可能的解釋：—Artificialintelligence.DifferentquanthedgefundsareimplementingAIinverydifferentways,indifferentareasandatdifferentspeeds,whichisnaturallyleadingtoverydifferentoutcomes.式和速度上存在顯著差異，這自然導(dǎo)致了截然不同的結(jié)果。Thismayseemlikeavague,faddishexplanation,butitmakesalotofsense.Thereisanimmenseamountofcomplexitywithmodernquantinvesting,andevenextremelysubtledifferencesinhowyouresearchsignals,constructaportfolioorimplementtradescanhaveabigimpact.Andwhilesomequanthedgefundshavebeenusingmachinelearningornaturallanguageprocessingforwelloveradecade,othersmightonlyhaverecentlyjumpedintothefield.投資涉及高度復(fù)雜性，即便在信號(hào)研究、組合構(gòu)建或交易執(zhí)環(huán)節(jié)存在極其細(xì)微的方法差異，都可能產(chǎn)生巨大影響。雖然部分量化對(duì)沖基金采用機(jī)器學(xué)習(xí)或自然語言處理技術(shù)已逾十年，但其他基金可能剛剛涉足這—領(lǐng)域。That’snottosaythattheAIvetshavenecessarilydonebetterthanthenewbies,merelythatverydifferentapproachestoAI—andareaswhereyouimplementit—couldplausiblyproducelarger-than-usualdispersioninreturns.這并不意味著AI領(lǐng)域的先行者必然比新入行者表現(xiàn)更優(yōu)，只是說明：不同的AI應(yīng)用路徑及其實(shí)施領(lǐng)域，確實(shí)可能引發(fā)較平常更為顯著的收益差異。Asaresult,hedgefundAandBthatmighthistoricallyhaveminedmanyofthesametradingsignalsandasaresultoftensufferedfromcorrelatedperformancemightnowincreasinglyseetheirfortunesdivergeattimesofstress.相關(guān)性的對(duì)沖基金A與B，如今壓力時(shí)期業(yè)績(jī)分化的現(xiàn)象可能愈—Crowding.AsAlphavillehaswrittenaboutbefore,severallargeproprietarytradingfirmshaveinrecentyearsexpandedaggressivelyinto“mid-frequency”trading.Thisinvolvesholdingpositionsforseveraldaysorevenoccasionallyweeks,andmeansagrowingdegreeofoverlapwithsystematichedgefunds—especiallythoseactiveinstatisticalarbitrage.——策略擁擠。正如Alphaville此前報(bào)道，近年來多家大型自營(yíng)Somequantssuggestedthatthismighthavebeenafactorbehindsomeofthesuddendrawdownsthatmanyquantstrategiessufferedatvarioustimesin2025,especiallyaroundthesummer.AsonefundmanagertoldAV:“Thecommonriskfactoriscrowding.Yougetpaidforcrowding,butalsohurtbyit,andit’softenveryhardtoidentifyaheadoftime.”Crowdingcanhurtallquantstrategies,butit’salsopossiblethatthearrivalandexpansionofanewbreedofsystematictraderaffectedcertainquantsmorethanothers.Forexample,whiletrend-followerswouldsimplyhavebeenwhipsawedbymarketevents,stat-arbfundsmighthaveseensomesignalsdecaymuchmorequicklythaninthepast.市場(chǎng)擁擠可能損害所有量化策略，但新型系統(tǒng)交易者