Номинальная скорость обработки ордеров - МТ4 vs МТ5. Тестируем вместе. - страница 6
Вы упускаете торговые возможности:
- Бесплатные приложения для трейдинга
- 8 000+ сигналов для копирования
- Экономические новости для анализа финансовых рынков
Регистрация
Вход
Вы принимаете политику сайта и условия использования
Если у вас нет учетной записи, зарегистрируйтесь
Корректно. Будем детально разбирать после 1 мая.
Немного по-теме:
Немного по-теме:
Прослеживается непонятная тенденция "чем больше спред, тем дольше время исполнения"
Не понятно, почему так???, чем больше спред тем проще же брокеру найти контрагента или я ошибаюсь?
Не наблюдал такого.
Все же нужна статистически значимая выборка.
Раз есть гипотеза, проведите ее простое конструктивное исследование и выложите результат (любой) здесь.
Here are my network test results from Limassol to Amsterdam, ping to MetaQuotes-Demo servers is 130 ms.
Tests on MetaTrader 5 build 642:
MetaTrader 4 build 419 tests:
It is clear that 10 transactions on MetaTrader 5 in asynchronous mode flew away in 31 ms, on average 3 ms per order. This is the rate of fire of the asynchronous mode "fire and forget" without any confirmation from the server. The first responses (in a batch) from the server will come somewhere in 145 ms, as in the usual synchronous mode. But just as 10 requests were fired in 31 ms, the reception of their results should also come in a batch in 145 ms.
Synchronous mode shows 149 ms execution of each order, which is a good result with a basic ping to the trading server of 130 ms. Small network packet size + efficient network protocol allow very fast data delivery.
MetaTrader 4 has only synchronous mode and the average execution time of one order turned out to be 207 ms, which is longer than the execution time of MetaTrader 5. One explanation for this difference is the cost of using a separate network context for trading operations . which can lead to some delays. In addition, in MetaTrader 5 we have applied more efficient mechanisms for order queues and execution.
It is important to remember that any measurements of network speed characteristics directly depend on the network itself. In my case, the base ping to MetaQuotes-Demo servers was at the level of 130 ms. This means that it is impossible to compare the results obtained between different measurements without taking into account and adjusting for the base ping.
I have attached two scripts so that anyone can run their own tests. Use them on the same MT4/MT5 servers, located in the same physical location, with the same access points and the same ping values. This will allow you to compare the results obtained in MT4 vs MT5 mode. The addresses of our demo trading servers MetaTrader 4 (demo.metaquotes.net:443) and MetaTrader 5 (access.metatrader5.com:443).
Version for MT5 got an error.
Is there anyone can help to correct it ? Thanks in advance