Search results

Definitely get "Smarter Trading: Improving Performance in Changing markets" and "Trading Systems and Methods" by Perry Kaufman. My library isn't very large. Most of my methods are original. I write a lot of code and test ideas constantly. My best advice is to learn to code up and test...

We should write this. Although it could be a weekend project for me, I’m sure others on ET are more qualified to design the authentication, encryption and message protocols. We’ll need to set up a few firewalled (IP secure, few open ports) servers that will host the database, re-distribute...

It depends. Both self-adaptive and manually adaptive systems should not be over optimized by allowing too many DoF, or DoFs that bring too much diversity. Either constrain the parameters to a range (subspace) which is known to be robust based on longer term, multi-market studies, or allow...

I didn’t work on the NNs long enough to get any success. I agree that traditional models are often superior, certainly for the processing speed advantage. Should be very successful. Option pricing models? Or is that already overdone? The SVD is very nice because once you have the...

Think of the advantages of us writing our own application to do this. It would be open source, so everyone could contribute their favorite feature. We could set up an official "track record" database server, trading competitions, or display trading signals and real time PnL in a IM chat...

Here is the matching client source. I don't remember where I found these sources. I tried looking for the link. Couldn't find it. Note these are not my modified versions. My versions are somewhat proprietary in that they are built into my systems.

Most of this can be done using a third party TCP server and client application. The server will… Query the TWS API for current positions, get advised when trades are made. Accept clients to connect and authenticate to a bound TCP port. Broadcast information at regular intervals to...

Diversification is also key, either by splitting size between uncorrelated systems, or by using a system design built on many "additive" non-correlating elements. I do both. I've had systems exceed their historical max drawdown (blow up), stopping out at a loss of 0.01 EUR/USD trading 6E, but...

Yes. You don’t always need to validate over a larger set if the backtesting / optmization is done properly. All systems “discovered” through backtesting will suffer from some amount of hindsight bias. Walk-forward performance may eventually decay to zero, go negative, or oscillate...

One option is to pass your data to a MEX function or save to a data file, then run it through a higher performance toolkit. Don’t they provide tools to export to other formats? Surely someone has figured out the formats. Perhaps instead of using the built-in Matlab toolkits (closed...

I agree. Once the socket interface is working in Matlab it can be rewritten in C. Matlab is much easier to prototype and debug than C. So it’s probably faster to start off with Matlab. It’s probably because most people keep their source code proprietary. I’m an good example. I have...

Implementing a Matlab to TWS connection using native-Matlab sockets code will be slow (much overhead), unless you really know what you are doing, or you only require real time data for a few symbols. I recommend using Matlab + MEX code for prototyping, simulation and backtesting. Use Matlab...

Tweaking the components and testing lots of hardware is great. But what about tweaking your code to run faster in general, or tweaking it to work better on your particular hardware? Assuming you write your own code, you can optimize it to certain processors, system designs or cache sizes...

Wouldn't it be even more usable with hyperthreading and your engine set to low priority? What is wrong with keeping your backtest engine's priority always set to low? It will still get all free CPU cycles.

My experience too. Nitro, do your systems tend to act highly correlated at times? I have this problem with certain systems. Fortunately, most of them are uncorrelated enough so that mixing is not a problem.

The more liquid eminis seem more efficient than the less liquid ones. ES seems to be the most efficient of all, hardest for my systems to crack.

I was wrong. I overlooked the missing "-" in the start menu string. I didn't want to test the strings since TWS was already running.

My destkop icons and Start menu items have the same startup strings. I see this on both Win2000 and Win2003 servers. Maybe that's because I rebooted them after upgrading to TWS 819? Maybe he needs to reboot?

Indeed the startup string changed from: (TWS version 818 and previous) "C:\Program Files\Java\j2re1.4.1_03\bin\javaw.exe" -jar C:\jts\jts.jar C:\jts to (TWS version 819) C:\WINNT\system32\javaw.exe cp C:\Jts\jts.jar;C:\Jts\jcommon-0.9.0.jar;C:\Jts\jfreechart-0.9.15.jar...

Yes, rolling back to 818 fixes it. Starting TWS the usual way (desktop icon or start menu) allows version 819.3 to work properly. The problem is that starting with version 819.3 the TWSStart utility is no longer able to start TWS. This configuration has worked fine for the last 6 months. I...