That's one of the guys I was talking about in my previous post! So if he is the best then it explains why MC is a such "funny" app.
Backtesting is tricky. A major academic publication with many backtests of momentum strategies considered seminal in fund industry was recently shown by a professor in Finland to suffer from look-ahead bias. http://papers.ssrn.com/sol3/papers.cfm?abstract_id=2743119 If professors can make such terrible backtesting mistakes what can you expect from retail traders? I would go for Amibroker. A little harder to learn but quite reliable when you sort all the details out. NinjaTrader is more suitable for autotrading in my opinion.
I've been very happy with MC as a testing and automation platform. I use the Easylanguage version. I don't have a need for low latency. If that's the case, don't even waste your time with retail platforms.
Exactly. These high frequency traders have an average-time-in-trade of seconds. Many algo traders have the same metric in minutes....and multicharts with a decent data feed should accomodate that with high reliability.
How can anyone trust his/her money on a funny company such as MC where developers and support don't even understand basic math? The one below is one of the funniest things I have ever seen in my whole life. A guy of a company called MC thinking that by simple conversion of one currency to another a loss would become a profit and vice versa and seriously asking "Why do you consider it to be wrong?". If that is not hilarious then I don't know the true definition of hilariocity. If I were a customer I would be terrified. Absolutely shocking. I would even go so far to say that they possibly never traded (much) themselves.
I've never had any calculation issues with MC performance reports. Only thing I noticed was their bar magnifier for backtesting isn't very indicative of real time fills. I expected that and that slippage is already priced into my proprietary model. As far as real-time automated trading goes, MC performance exceeds my expectations. It performs well if coded properly.
I agree, that's pretty bad. They obviously have a bug in their currency conversion methods or functions. However, to me, that is not a show stopper. Now, that being said, if they were not computing the Sharpe or Sortino ratio correctly, that WOULD be a show stopper.
I could care less about MC performance monitoring. I wouldn't trust it from any retail third party platform. I use my own custom spreadsheet for this. If I wanted the closest to real-time from a platform, I would generate the signals from MC through the IB paper-trader API and look at their reports. IMO, there are two things a third party platform needs to have: 1. Capability to code my logic in the simplest language possible 2. Ability to send orders and manage positions (IOW automate) If it can do this, the 2k you drop on it is chump change. Seeking edge refinement from these platforms is a fools game