- Search Crypto Craft
- 1,424 Results (11 Threads, 1,413 Replies)
- mikkom replied Apr 7, 2011
Have you tested both in-sample and out-sample? If those are in-sample curves then you WILL find all kinds of strategies that "work" after testing a lot of strategies. Unfortunately that is due to survivorship bias - that's why the fitness algo is so ...
Systematic trading
- mikkom replied Apr 7, 2011
One question you should ask yourself is are the profits under the new high really so unimportant that you don't want to look their results?
Systematic trading
- mikkom replied Apr 6, 2011
What does "higher high frequency" mean? Higher amount of trades with even distribution?
Systematic trading
- mikkom replied Apr 5, 2011
IMHO programming is by far the easiest part of the project.
Systematic trading
- mikkom replied Apr 2, 2011
Wow.. lots of questions. I'm not going to answer all of them but I'll answer some.. I have several kinds of events, there are binary ones (on/off on certain time period) and price level (under/over) ones. Events can occur on multiple timeframes or ...
Systematic trading
- mikkom replied Mar 24, 2011
I would also like to recomment this book, it's really one of my favourite basic books and goes through many basic concepts quite well and it's quite well written (the author is the founder of Trading Blox and one of the original turtles).
The Birth of a Quant Trader
- mikkom replied Mar 24, 2011
To be honest, I have no idea how banks do it.
The Birth of a Quant Trader
- mikkom replied Mar 24, 2011
That is not what I'm trying to say at all. What I'm trying to say is that it would be much more useful to be clear on comments on what you are exactly talking about than post comments that rely on authority rather than facts.
The Birth of a Quant Trader
- mikkom replied Mar 24, 2011
For example because there are volatility trading strategies where people estimate volatility better than IV and make money with it? This "quant" stuff is not all about directional simple strategies.
The Birth of a Quant Trader
- mikkom replied Mar 24, 2011
You still didn't answer the question. The wikipedia page contains at least 5 different cases of volatility. LTCM fiasco kind of proved that "traditional" volatility formulas are not exactly correct as they are based on flawed assumptations (they ...
The Birth of a Quant Trader
- mikkom replied Mar 23, 2011
That is one of the most incorrect comments of this thread. Of course there are academic papers with value out there. This thread has turned into quant vs non-quant approach war and I think it's totally unnecessary. Oh well, I go back to my cave
...The Birth of a Quant Trader
- mikkom replied Mar 3, 2011
I have. The problem with that approach is that exponential looking curves give you the best results (assuming you only use downside deviation - I don't personally like using non-directional deviation, there is no point in penalizing good behaviour). ...
Systematic trading
- mikkom replied Mar 2, 2011
TF system? I'm not sure what you mean. Yes, it's a combination of 21 systems (of about ~95000 candidates (which also are all profitable and have certain features why they are included at the pool)). Fitness algo is the core of evolutionary ...
Systematic trading
- mikkom replied Mar 2, 2011
What a difference a good fitness algo does — I have been testing a new fitness algo for a few days and I have to say I'm a bit astounded.. I thought there is a bug or something but after trying different methods (mar etc) and comparing ...
Systematic trading
- mikkom replied Feb 27, 2011
IMHO kelly is far too aggressive (you can always use percentage of kelly if you want to but you will realize that people, including you, tend to want to optimize to something else than aggressive profits) and the usual percentage risk actually is ...
Systematic trading
- mikkom replied Feb 22, 2011
The question is what you are looking for in a position sizing algorithm - if you trade multiple systems/positions you might want to limit risk to certain level for example or you might want to limit directional/correlated risk to some predefined ...
Systematic trading
- mikkom replied Feb 21, 2011
Of course you need out-sample verification. All in-sample data results should be considered as false (IMHO).
Systematic trading
- mikkom replied Feb 17, 2011
All that traders do is curve fit our systems / behaviour to past data. The actual question is: How to find those events that repeat themselves so you curve fit to the correct parts of the curve. One answer is to use so much data that the system that ...
Systematic trading