{"id":13320,"date":"2026-06-22T19:46:52","date_gmt":"2026-06-22T19:46:52","guid":{"rendered":"https:\/\/bjftradinggroup.com\/?p=13320"},"modified":"2026-10-17T22:37:49","modified_gmt":"2026-10-17T22:37:49","slug":"latency-arbitrage-optimization-live-test","status":"publish","type":"post","link":"https:\/\/bjftradinggroup.com\/tr\/latency-arbitrage-optimization-live-test\/","title":{"rendered":"Ayn\u0131 Strateji, \u0130ki Hesap: XAUUSD’de Optimize ve Varsay\u0131lan Gecikme Arbitraj\u0131"},"content":{"rendered":"

<\/p>\n

\n


\n
\n<\/p>\n

\n

BJF TRADING GROUP \u00b7 SHARPTRADER OPTIMIZER<\/span><\/p>\n

Ayn\u0131 Strateji, \u0130ki Hesap: Optimize Edilmi\u015f Gecikme Arbitraj\u0131 Parametreleri XAUUSD<\/span>‘de Ger\u00e7ekte Neyi De\u011fi\u015ftirdi<\/h2>\n

Ayn\u0131 gecikme arbitraj\u0131 mant\u0131\u011f\u0131n\u0131 iki canl\u0131 hesapta ayn\u0131 anda \u00e7al\u0131\u015ft\u0131rd\u0131k: ayn\u0131 0,1 lot b\u00fcy\u00fckl\u00fc\u011f\u00fc, ayn\u0131 alt\u0131n paritesi, ayn\u0131 ba\u015flang\u0131\u00e7 bakiyesi. Hesaplardan biri SharpTrader Optimizer<\/strong>‘da ayarlanm\u0131\u015f parametreleri kulland\u0131. Di\u011feri varsay\u0131lan ayarlarla \u00e7al\u0131\u015ft\u0131. Yakla\u015f\u0131k 5.700 i\u015flem boyunca optimize edilmi\u015f hesap yaln\u0131zca daha fazla kazanmad\u0131. Ortalama zarar\u0131 \u00fc\u00e7te iki oran\u0131nda azaltt\u0131 ve k\u00e2r fakt\u00f6r\u00fcn\u00fc iki kat\u0131na \u00e7\u0131kard\u0131. \u0130\u015fte ayr\u0131nt\u0131l\u0131 d\u00f6k\u00fcm.<\/p>\n

Enstr\u00fcman:<\/strong> XAUUSD
\nD\u00f6nem:<\/strong> 19\u201322 Haziran 2026
\nLot b\u00fcy\u00fckl\u00fc\u011f\u00fc:<\/strong> 0,1 (sabit)
\nBa\u015flang\u0131\u00e7 bakiyesi:<\/strong> her biri 5.000 $<\/div>\n<\/div>\n


\n
\n<\/p>\n

\n

K\u0131sa cevap:<\/strong> Gecikme arbitraj\u0131 parametrelerini optimize etmek, i\u015flem ba\u015f\u0131na daha b\u00fcy\u00fck k\u00e2r pe\u015finde ko\u015fmak de\u011fildi. Zararl\u0131 i\u015flemlerin maliyetini k\u00fc\u00e7\u00fcltt\u00fc. Optimize edilmi\u015f hesab\u0131n ortalama zarar\u0131 \u22126,20 $’dan \u22122,04 $’a<\/strong> d\u00fc\u015ft\u00fc, ortalama zararl\u0131 i\u015flemi \u22126,2 pip’ten \u22122,0 pip’e<\/strong> geriledi ve k\u00e2r fakt\u00f6r\u00fc 5,41’den 11,3’e<\/strong> y\u00fckseldi. Test boyunca net sonu\u00e7: 5.000 $’l\u0131k taban \u00fczerinde kapat\u0131lm\u0131\u015f k\u00e2r olarak 14.576 $’a kar\u015f\u0131 11.041 $<\/strong> ve daha d\u00fc\u015f\u00fck maksimum d\u00fc\u015f\u00fc\u015f (drawdown).<\/p>\n<\/div>\n


\n
\n<\/p>\n

Tek sat\u0131rda sonu\u00e7<\/h2>\n

Her iki hesap da 5.000 $ ile ba\u015flad\u0131 ve alt\u0131nda birebir ayn\u0131 gecikme arbitraj\u0131 giri\u015f mant\u0131\u011f\u0131yla i\u015flem yapt\u0131. Tek fark parametre setiydi: biri SharpTrader Optimizer’daki bir optimizasyon \u0131zgaras\u0131ndan \u00e7\u0131kt\u0131, di\u011feri standart varsay\u0131lanlar\u0131 kulland\u0131. \u0130\u015fte ana fark.<\/p>\n

\n
+$14,576<\/span>Optimize net<\/span><\/div>\n
+$11,041<\/span>Varsay\u0131lan net<\/span><\/div>\n
11.3<\/span>Optimize PF<\/span><\/div>\n
5.41<\/span>Varsay\u0131lan PF<\/span><\/div>\n<\/div>\n

K\u00e2r fark\u0131 ger\u00e7ek, ancak sayfadaki en az ilgin\u00e7 rakam bu. \u00dc\u00e7 seans boyunca %32 daha y\u00fcksek net toplam g\u00fcr\u00fclt\u00fc olabilir. G\u00fcr\u00fclt\u00fc olmayan \u015fey, optimize edilmi\u015f hesab\u0131n oraya nas\u0131l<\/em> ula\u015ft\u0131\u011f\u0131d\u0131r ve okumaya de\u011fer k\u0131s\u0131m da budur.<\/p>\n


\n
\n<\/p>\n

Yan yana iki sermaye e\u011frisi<\/h2>\n

Test ilerlemesine g\u00f6re \u00e7izildi\u011finde (kapat\u0131lm\u0131\u015f i\u015flemlerin %0’\u0131ndan %100’\u00fcne), her iki e\u011fri de istikrarl\u0131 \u015fekilde y\u00fckselir; bu, \u00e7al\u0131\u015fan bir arbitraj avantaj\u0131ndan beklenen \u015feydir. Optimize edilmi\u015f \u00e7izgi kademeli olarak \u00f6ne ge\u00e7er ve yol boyunca daha az ve daha s\u0131\u011f d\u00fc\u015f\u00fc\u015flerle daha y\u00fcksek bir seviyede biter.<\/p>\n


\n\"0,1<\/p>\n

5.000 $’l\u0131k tabandan kapat\u0131lm\u0131\u015f i\u015flem sermayesi; iki farkl\u0131 i\u015flem say\u0131s\u0131n\u0131n (3.588’e kar\u015f\u0131 2.149) kar\u015f\u0131la\u015ft\u0131r\u0131labilmesi i\u00e7in test ilerlemesine g\u00f6re normalize edilmi\u015ftir.<\/p>\n<\/div>\n


\n
\n<\/p>\n

Rakamlar\u0131n tamam\u0131<\/h2>\n

Bu tabloyu yukar\u0131dan a\u015fa\u011f\u0131ya okuyun. Hik\u00e2ye k\u00e2r sat\u0131r\u0131nda de\u011fil. Zarar sat\u0131rlar\u0131nda.<\/p>\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n
Metrik<\/th>\nOptimize<\/th>\nVarsay\u0131lan<\/th>\n<\/tr>\n<\/thead>\n
Kapat\u0131lan pozisyonlar<\/td>\n3,588<\/td>\n2,149<\/td>\n<\/tr>\n
Net kapat\u0131lm\u0131\u015f k\u00e2r<\/td>\n+$14,576<\/td>\n+$11,041<\/td>\n<\/tr>\n
Kazanma oran\u0131<\/td>\n79.4%<\/td>\n81.2%<\/td>\n<\/tr>\n
K\u00e2r fakt\u00f6r\u00fc<\/td>\n11.3<\/td>\n5.41<\/td>\n<\/tr>\n
Ortalama kazan\u00e7<\/td>\n+$5.61<\/td>\n+$7.76<\/td>\n<\/tr>\n
Ortalama zarar<\/td>\n\u2212$2.04<\/td>\n\u2212$6.20<\/td>\n<\/tr>\n
Ortalama zararl\u0131 i\u015flem (pip)<\/td>\n\u22122.0<\/td>\n\u22126.2<\/td>\n<\/tr>\n
Maksimum d\u00fc\u015f\u00fc\u015f (drawdown)<\/td>\n0.42%<\/td>\n0.62%<\/td>\n<\/tr>\n
Maksimum ard\u0131\u015f\u0131k zarar<\/td>\n4<\/td>\n5<\/td>\n<\/tr>\n
\u0130\u015flem ba\u015f\u0131na net k\u00e2r std. sapmas\u0131<\/td>\n$5.83<\/td>\n$7.12<\/td>\n<\/tr>\n
Ort. i\u015flem s\u00fcresi<\/td>\n1,4 sn<\/td>\n3,7 sn<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n
\n

Sezgiye ayk\u0131r\u0131 k\u0131s\u0131m<\/h3>\n

Optimize edilmi\u015f hesab\u0131n kazanma oran\u0131 daha d\u00fc\u015f\u00fckt\u00fc<\/strong> (%79,4’e kar\u015f\u0131 %81,2) ve ortalama kazanc\u0131 daha k\u00fc\u00e7\u00fckt\u00fc<\/strong> (7,76 $’a kar\u015f\u0131 5,61 $). Basit bir puan tablosunda daha k\u00f6t\u00fc g\u00f6r\u00fcn\u00fcr. Yine de daha az riskle daha fazla para kazand\u0131. Neden? \u00c7\u00fcnk\u00fc zararl\u0131 i\u015flemleri \u00fc\u00e7 kat daha ucuzdu. Gecikme arbitraj\u0131nda i\u015flem ba\u015f\u0131na avantaj\u0131n\u0131z \u00e7ok k\u00fc\u00e7\u00fckt\u00fcr ve d\u00fc\u015fman\u0131n\u0131z k\u00f6t\u00fc dolumdur. Zarar kuyru\u011funu daralt\u0131n, sonraki her \u015fey iyile\u015fir.<\/p>\n<\/div>\n


\n
\n<\/p>\n

Gecikme arbitraj\u0131nda neden her \u015fey zarar kuyru\u011funa ba\u011fl\u0131d\u0131r<\/h2>\n

Bir gecikme arbitraj\u0131 sinyali milisaniyeler boyunca ya\u015far. Bir platformun \u00e7oktan ula\u015ft\u0131\u011f\u0131, di\u011ferinin ise hen\u00fcz yeti\u015fmedi\u011fi bir fiyata g\u00f6re hareket edersiniz. Dolum temiz oldu\u011funda birka\u00e7 pip kazan\u0131rs\u0131n\u0131z. Dolum ge\u00e7 kald\u0131\u011f\u0131nda istedi\u011finiz fiyat kaybolmu\u015ftur ve art\u0131k zaten ger\u00e7ekle\u015fmi\u015f bir hareketin yanl\u0131\u015f taraf\u0131nda pozisyon tutuyorsunuzdur. Bu kaymal\u0131 bir giri\u015ftir ve para kaymal\u0131 giri\u015flerden s\u0131zar.<\/p>\n

Varsay\u0131lan hesab\u0131n ortalama zararl\u0131 i\u015flemine bak\u0131n: \u22126,2 pip<\/strong>. Bu, stratejinin y\u00f6n konusunda yan\u0131ld\u0131\u011f\u0131 anlam\u0131na gelmez. Kazanma oran\u0131 y\u00f6n okumas\u0131n\u0131n do\u011fru oldu\u011funu kan\u0131tl\u0131yor; i\u015flemlerin %81’i kazand\u0131. \u22126,2 pip emir ger\u00e7ekle\u015ftirme maliyetidir: sinyalin tetiklendi\u011fi fiyat ile hesab\u0131n ger\u00e7ekte ald\u0131\u011f\u0131 fiyat aras\u0131ndaki fark, art\u0131 emir defterini ge\u00e7erken \u00f6denen de\u011fi\u015fken spread. Optimize edilmi\u015f hesap ayn\u0131 t\u00fcrden zararl\u0131 i\u015flem i\u00e7in yaln\u0131zca \u22122,0 pip<\/strong> \u00f6dedi. Ayn\u0131 avantaj, s\u0131z\u0131nt\u0131n\u0131n \u00fc\u00e7te biri.<\/p>\n

\n

Burada “optimizasyon” ne anlama geliyordu<\/h3>\n
    \n
  • Daha s\u0131k\u0131 giri\u015f gecikmesi b\u00fct\u00e7esi.<\/strong> Optimize edilmi\u015f set yaln\u0131zca modellenen gidi\u015f-d\u00f6n\u00fc\u015f emir ger\u00e7ekle\u015ftirme s\u00fcresi arbitraj\u0131n h\u00e2l\u00e2 ge\u00e7erli oldu\u011fu pencereye s\u0131\u011fd\u0131\u011f\u0131nda i\u015flem a\u00e7ar. Bayat sinyaller ge\u00e7 al\u0131nmak yerine atlan\u0131r.<\/li>\n
  • Spread duyarl\u0131 tetikleme.<\/strong> Giri\u015fler sabit bir varsay\u0131ma de\u011fil, canl\u0131 de\u011fi\u015fken spread’e g\u00f6re filtrelenir; b\u00f6ylece hesap ge\u00e7ici olarak geni\u015flemi\u015f bir emir defterini ge\u00e7mek i\u00e7in \u00f6deme yapmay\u0131 b\u0131rak\u0131r.<\/li>\n
  • Daha h\u0131zl\u0131 \u00e7\u0131k\u0131\u015f y\u00f6netimi.<\/strong> Ortalama i\u015flem s\u00fcresi 3,7 saniyeden 1,4 saniyeye d\u00fc\u015ft\u00fc. Pozisyonlar bir an fazla tutulmak yerine fiyat sapmas\u0131 h\u00e2l\u00e2 lehinizeyken kapat\u0131l\u0131r.<\/li>\n
  • Daha fazla agresiflik de\u011fil, daha y\u00fcksek se\u00e7icilik.<\/strong> Optimize edilmi\u015f hesap genel olarak daha fazla i\u015flem yapt\u0131, ancak her biri daha kat\u0131 bir emir ger\u00e7ekle\u015ftirme kalitesi e\u015fi\u011fini ge\u00e7ti. Pahal\u0131 sinyalleri almay\u0131 b\u0131rakt\u0131\u011f\u0131n\u0131zda ucuz ve temiz sinyaller bol oldu\u011fu i\u00e7in i\u015flem say\u0131s\u0131 artt\u0131.<\/li>\n<\/ul>\n<\/div>\n


    \n
    \n<\/p>\n

    Test nas\u0131l y\u00fcr\u00fct\u00fcld\u00fc<\/h2>\n
    \n

    Tasar\u0131m:<\/strong> ayn\u0131 piyasa seanslar\u0131nda paralel \u00e7al\u0131\u015ft\u0131r\u0131lan iki canl\u0131 hesap; b\u00f6ylece her ikisi de birebir ayn\u0131 fiyat hareketini ve ayn\u0131 likidite ko\u015fullar\u0131n\u0131 g\u00f6rd\u00fc.<\/p>\n

    Kontrol edilen de\u011fi\u015fkenler:<\/strong> ayn\u0131 enstr\u00fcman (XAUUSD), ayn\u0131 sabit lot b\u00fcy\u00fckl\u00fc\u011f\u00fc (0,1), ayn\u0131 ba\u015flang\u0131\u00e7 bakiyesi (5.000 $), ayn\u0131 broker ko\u015fullar\u0131, ayn\u0131 gecikme arbitraj\u0131 \u00e7ekirdek mant\u0131\u011f\u0131.<\/p>\n

    De\u011fi\u015ftirilen tek de\u011fi\u015fken:<\/strong> parametre seti. A hesab\u0131, SharpTrader Optimizer’daki bir parametre \u0131zgaras\u0131ndan se\u00e7ilen de\u011ferleri kulland\u0131. B hesab\u0131 standart varsay\u0131lanlar\u0131 kulland\u0131.<\/p>\n

    Zaman aral\u0131\u011f\u0131:<\/strong> 19\u201322 Haziran 2026. Optimize edilmi\u015f hesapta 3.588, varsay\u0131lan hesapta 2.149 kapat\u0131lm\u0131\u015f pozisyon.<\/p>\n<\/div>\n

    Kar\u015f\u0131la\u015ft\u0131rmay\u0131 adil k\u0131lan, her iki hesab\u0131n ayn\u0131 anda \u00e7al\u0131\u015ft\u0131r\u0131lmas\u0131d\u0131r. Pazartesi optimize edip sal\u0131 varsay\u0131lanlarla \u00e7al\u0131\u015ft\u0131r\u0131rsan\u0131z, herhangi bir fark yaln\u0131zca pazartesi ile sal\u0131 aras\u0131ndaki fark olabilir. Yan yana \u00e7al\u0131\u015ft\u0131r\u0131ld\u0131\u011f\u0131nda piyasa sabittir ve fark\u0131 a\u00e7\u0131klayabilecek tek \u015fey parametrelerdir.<\/p>\n


    \n
    \n<\/p>\n

    Optimizer bu parametreleri nas\u0131l buldu<\/h2>\n

    Parametreler tahminden ya da g\u00fczel bir sermaye e\u011frisine a\u015f\u0131r\u0131 uydurmadan (curve-fitting) gelmedi. Emir ger\u00e7ekle\u015ftirmeye duyarl\u0131 stratejileri canl\u0131 piyasan\u0131n onlara ger\u00e7ekte davrand\u0131\u011f\u0131 \u015fekilde puanlamak i\u00e7in tasarlanm\u0131\u015f SharpTrader Optimizer’daki bir \u0131zgara aramas\u0131ndan geldi. D\u00f6rt unsur bu aramay\u0131 bir gecikme stratejisi i\u00e7in g\u00fcvenilir k\u0131lar:<\/p>\n

    \n
    \n
    01<\/div>\n

    Ger\u00e7ek tick verileri<\/h3>\n

    Arama, bar yakla\u015f\u0131mlar\u0131 yerine ger\u00e7ek ge\u00e7mi\u015f tick ak\u0131\u015flar\u0131 \u00fczerinde \u00e7al\u0131\u015f\u0131r. Tek bir tick’lik fiyat sapmas\u0131 i\u00e7inde ya\u015fayan bir strateji i\u00e7in bar verileri, yakalamaya \u00e7al\u0131\u015ft\u0131\u011f\u0131n\u0131z \u015feyi tam olarak gizler.<\/p>\n<\/div>\n

    \n
    02<\/div>\n

    Emir ger\u00e7ekle\u015ftirme s\u00fcresi modellemesi<\/h3>\n

    Milisaniye cinsinden ger\u00e7ek\u00e7i bir emir ger\u00e7ekle\u015ftirme gecikmesi belirlersiniz ve optimizer bunun sonu\u00e7lar\u0131 nas\u0131l de\u011fi\u015ftirdi\u011fini g\u00f6sterir. Standart geriye d\u00f6n\u00fck test ara\u00e7lar\u0131 s\u0131f\u0131r gecikmeli dolum varsayar; bu da gecikme stratejilerinin k\u00e2\u011f\u0131t \u00fczerinde canl\u0131 ortamdakinden daha iyi g\u00f6r\u00fcnmesine yol a\u00e7an varsay\u0131m\u0131n ta kendisidir.<\/p>\n<\/div>\n

    \n
    03<\/div>\n

    Tick ba\u015f\u0131na de\u011fi\u015fken spread<\/h3>\n

    Ge\u00e7mi\u015f spread, sabit bir broker varsay\u0131lan\u0131 olarak de\u011fil, tick tick uygulan\u0131r. Giri\u015f ve \u00e7\u0131k\u0131\u015f\u0131n her biri o anda ger\u00e7ekten var olan spread ile \u00fccretlendirilir; b\u00f6ylece \u22126,2’ye kar\u015f\u0131 \u22122,0 pip fark\u0131, canl\u0131da size mal olmadan \u00f6nce testte ortaya \u00e7\u0131kar.<\/p>\n<\/div>\n

    \n
    04<\/div>\n

    Her iki bacakta kayma<\/h3>\n

    A\u00e7\u0131l\u0131\u015f ve kapan\u0131\u015f kayma a\u00e7\u0131s\u0131ndan ba\u011f\u0131ms\u0131z olarak modellenir. Bir gecikme stratejisi giri\u015fte de \u00e7\u0131k\u0131\u015fta da kaymaya u\u011frayabilir ve optimizer \u00e7\u0131k\u0131\u015f\u0131n bedava oldu\u011funu varsaymak yerine her ikisini de fiyatlar.<\/p>\n<\/div>\n<\/div>\n

    Buna ek olarak, \u0131zgara t\u00fcm CPU \u00e7ekirdeklerinde y\u00fcksek kullan\u0131mla \u00e7al\u0131\u015f\u0131r; b\u00f6ylece 100.000’den fazla parametre kombinasyonu \u00fczerinde yap\u0131lan bir arama g\u00fcnler yerine saatler i\u00e7inde tamamlan\u0131r ve 24 saatlik performans \u0131s\u0131 haritas\u0131 alt\u0131nda avantaj\u0131n ger\u00e7ekte hangi i\u015flem saatlerinde oldu\u011funu g\u00f6sterir. Bu testteki parametreler, emir ger\u00e7ekle\u015ftirme s\u00fcresi, de\u011fi\u015fken spread ve iki bacakl\u0131 kayma d\u00fcr\u00fcst\u00e7e hesaba kat\u0131ld\u0131ktan sonra en y\u00fcksek puan\u0131 alan settir.<\/p>\n


    \n
    \n<\/p>\n

    Gecikme arbitraj\u0131 yap\u0131yorsan\u0131z bundan \u00e7\u0131kar\u0131lacak dersler<\/h2>\n
      \n
    • Kazanma oran\u0131 i\u00e7in optimize etmeyi b\u0131rak\u0131n.<\/strong> Kazanma oran\u0131nda daha k\u00f6t\u00fc g\u00f6r\u00fcnen hesap daha fazla para kazand\u0131. Bunun yerine k\u00e2r fakt\u00f6r\u00fc ve ortalama zarar b\u00fcy\u00fckl\u00fc\u011f\u00fc i\u00e7in optimize edin.<\/li>\n
    • Zararlar\u0131n\u0131z bir y\u00f6n metri\u011fi de\u011fil, bir emir ger\u00e7ekle\u015ftirme metri\u011fidir.<\/strong> \u0130\u015flemlerin %80’i kazand\u0131\u011f\u0131nda, \u22126 pip’lik ortalama zarar size sinyalinizi de\u011fil, dolumlar\u0131 ve spread’i anlat\u0131r. Sorunu emir ger\u00e7ekle\u015ftirme katman\u0131nda \u00e7\u00f6z\u00fcn.<\/li>\n
    • Varsay\u0131lanlar bir ba\u015flang\u0131\u00e7 noktas\u0131d\u0131r, asla biti\u015f \u00e7izgisi de\u011fil.<\/strong> Varsay\u0131lan set k\u00e2rl\u0131yd\u0131. Ama ayn\u0131 zamanda net k\u00e2r\u0131n yakla\u015f\u0131k \u00fc\u00e7te birini masada b\u0131rak\u0131yor ve bunun i\u00e7in daha fazla d\u00fc\u015f\u00fc\u015f ta\u015f\u0131yordu.<\/li>\n
    • Gecikme ve de\u011fi\u015fken spread a\u00e7\u0131kken test edin.<\/strong> S\u0131f\u0131r gecikmeli dolum ve sabit spread varsayan her geriye d\u00f6n\u00fck test, bir gecikme stratejisini oldu\u011fundan iyi g\u00f6sterir ve en \u00f6nemli parametreyi tam olarak gizler.<\/li>\n<\/ul>\n


      \n
      \n<\/p>\n

      S\u0131k\u00e7a Sorulan Sorular<\/h2>\n
      \n
      Daha y\u00fcksek k\u00e2r fakt\u00f6r\u00fc yaln\u0131zca daha az i\u015flem anlam\u0131na m\u0131 gelir?<\/div>\n
      \n

      Hay\u0131r. Bu testte optimize edilmi\u015f hesap daha fazla<\/strong> i\u015flem yapt\u0131 (2.149’a kar\u015f\u0131 3.588) ve yine de daha y\u00fcksek k\u00e2r fakt\u00f6r\u00fcne ula\u015ft\u0131 (5,41’e kar\u015f\u0131 11,3). \u0130yile\u015fme daha az i\u015flem yapmaktan de\u011fil, daha ucuz zararl\u0131 i\u015flemlerden geldi.<\/p>\n<\/div>\n

      Optimize edilmi\u015f hesab\u0131n kazanma oran\u0131 neden daha d\u00fc\u015f\u00fck ama k\u00e2r\u0131 daha fazlayd\u0131?<\/div>\n
      \n

      \u00c7\u00fcnk\u00fc k\u00e2r, kazanma oran\u0131 \u00e7arp\u0131 ortalama kazan\u00e7 eksi kaybetme oran\u0131 \u00e7arp\u0131 ortalama zarard\u0131r. Ortalama zarar\u0131 \u22126,20 $’dan \u22122,04 $’a d\u00fc\u015f\u00fcrmek, biraz daha d\u00fc\u015f\u00fck kazanma oran\u0131n\u0131 ve daha k\u00fc\u00e7\u00fck ortalama kazanc\u0131 fazlas\u0131yla telafi etti. Matematik, y\u00fcksek isabet oran\u0131ndan \u00e7ok zarar kontrol\u00fcn\u00fc \u00f6d\u00fcllendirir.<\/p>\n<\/div>\n

      \u00dc\u00e7 g\u00fcnl\u00fck bir test herhangi bir \u015feyi kan\u0131tlamaya yeter mi?<\/div>\n
      \n

      Mutlak getirileri de\u011ferlendirmek i\u00e7in \u00fc\u00e7 g\u00fcn k\u0131sad\u0131r; bu y\u00fczden dolar toplam\u0131yla de\u011fil, yap\u0131sal metriklerle ba\u015fl\u0131yoruz. Yakla\u015f\u0131k 5.700 i\u015flem boyunca ortalama zarar, k\u00e2r fakt\u00f6r\u00fc ve zararl\u0131 i\u015flem ba\u015f\u0131na pip maliyetindeki farklar, man\u015fetteki net rakamdan \u00e7ok daha istikrarl\u0131d\u0131r. Dolar fark\u0131 k\u00fc\u00e7\u00fclebilir. Zarar kuyru\u011fu fark\u0131 ise parametrelerin bir \u00f6zelli\u011fidir.<\/p>\n<\/div>\n

      Ayn\u0131 yakla\u015f\u0131m alt\u0131n d\u0131\u015f\u0131ndaki paritelerde de i\u015fe yarar m\u0131?<\/div>\n
      \n

      Y\u00f6ntem aktar\u0131labilir, ancak parametreler aktar\u0131lamaz. Her enstr\u00fcman\u0131n kendine \u00f6zg\u00fc spread davran\u0131\u015f\u0131, volatilitesi ve emir ger\u00e7ekle\u015ftirme profili vard\u0131r; bu nedenle \u0131zgara aramas\u0131 her enstr\u00fcman i\u00e7in yeniden \u00e7al\u0131\u015ft\u0131r\u0131lmal\u0131d\u0131r. Burada alt\u0131n se\u00e7ildi \u00e7\u00fcnk\u00fc spread ve tick davran\u0131\u015f\u0131 emir ger\u00e7ekle\u015ftirme maliyetini g\u00f6rmeyi kolayla\u015ft\u0131r\u0131r.<\/p>\n<\/div>\n

      Tam olarak ne optimize edildi?<\/div>\n
      \n

      Giri\u015fleri ve \u00e7\u0131k\u0131\u015flar\u0131 filtreleyen gecikme ve spread e\u015fikleri; ger\u00e7ek tick verileri, modellenmi\u015f emir ger\u00e7ekle\u015ftirme s\u00fcresi, tick ba\u015f\u0131na de\u011fi\u015fken spread ve iki bacakl\u0131 kayma ile puanlanan bir parametre \u0131zgaras\u0131na g\u00f6re ayarland\u0131. \u00c7ekirdek strateji mant\u0131\u011f\u0131 her iki hesapta da ayn\u0131yd\u0131.<\/p>\n<\/div>\n<\/div>\n


      \n
      \n<\/p>\n

      \n

      BJF trading ara\u015ft\u0131rmalar\u0131na abone olun<\/h3>\n

      Yeni emir ger\u00e7ekle\u015ftirme \u00e7al\u0131\u015fmalar\u0131, optimizer anlat\u0131mlar\u0131 ve \u00fcr\u00fcn s\u00fcr\u00fcmleri \u2013 yay\u0131nlad\u0131\u011f\u0131m\u0131z anda size ula\u015ft\u0131r\u0131l\u0131r.<\/p>\n

      <\/div>