## Search found 98 matches

Wed Mar 25, 2015 11:41 am
Forum: Trend Indicators and Signals
Topic: 20% upside?
Replies: 1
Views: 1657

### Re: 20% upside?

This rule sounds very subjective - it could be almost anything, i.e. distance to nearest support/resistance defined by previous lows/highs; distance to a Fibonacci ratio retracement or extension; a point at which some fundamental valuation signals an over valued or "good" value price, etc. It's simp...
Wed Mar 25, 2015 11:34 am
Forum: Trend Indicators and Signals
Topic: Should trading system be stopped?
Replies: 6
Views: 4525

### Re: Should trading system be stopped?

oem7110 wrote:Does anyone have any suggestions on what the formula is for gap-less indicators?
Thanks, to everyone very much for any suggestions :>
Add the "distance" of the gap to all prices prior to the gap that fall within the moving window.
Tue Aug 27, 2013 8:04 am
Forum: Testing and Simulation
Topic: Trading a large number of systems without a large account
Replies: 9
Views: 9987
Combining "experts" in this way is known as Linear Opinion Pools. Googling this term brings up several links to academic papers, such as those linked in my answer to my own question on this forum.
Mon Apr 08, 2013 5:52 pm
Forum: Testing and Simulation
Replies: 4
Views: 3053
I always use brute Force method. Just run through the parameters and look at performance. Sorry, but I disagree with this. It demonstrates two serious biases: data mining bias and hindsight bias. Data mining bias because, given a set of data, you torture it enough to give results that are pleasing ...
Wed Feb 27, 2013 7:43 pm
Forum: Trend Indicators and Signals
Topic: Indicators that are scaled in the range 0 to 1
Replies: 3
Views: 3236

### Indicators that are scaled in the range 0 to 1

I have a simple request; can the board suggest indicators that are scaled in the 0 to 1 range, or can be easily adapted to be so?
Wed Oct 17, 2012 7:45 am
Forum: Testing and Simulation
Topic: Minimum Correlation Algorithm
Replies: 0
Views: 6633

### Minimum Correlation Algorithm

Over time there have been several different threads on correlation on this board, so I thought readers might be interested in a new approach using correlation called the "Minimum Correlation Algorithm." An academic paper, a spreadsheet and more information is available from here.
Tue Oct 09, 2012 6:11 pm
Forum: Testing and Simulation
Topic: Curve Fitting or appropriate market selection
Replies: 15
Views: 8227
if a market has NOT been profitable for 30 straight years, or 20 straight years, or even 10 straight years, then shouldn't I assume some characteristic about that market makes it unlikely to trade profitably in the future? Well, yeah! You have a lot of evidence that shows no matter how you tweak th...
Thu Oct 04, 2012 6:55 am
Forum: Testing and Simulation
Topic: Rolling Correlations
Replies: 3
Views: 2478
You might also find the ideas here useful.
Fri Jul 20, 2012 9:38 am
Forum: Money Management
Topic: How to determine the positions size for each trade?
Replies: 11
Views: 7869
When it comes to the Kelly Ratio I've always liked this series of online tutorials.
Fri Jun 29, 2012 9:55 pm
Forum: Trend Indicators and Signals
Topic: Spurious indicators, or uses thereof
Replies: 3
Views: 3600

### Spurious indicators, or uses thereof

In one of the Market Wizard interviews I remember reading a quote to the effect that "angle techniques are no technique at all," I presumed at the time referring to Gann angles, on the grounds that angles drawn on charts are entirely dependent on the scale to which the y-axis is drawn, which of cour...
Sat Jun 16, 2012 10:04 am
Forum: Testing and Simulation
Topic: What to make of this chart?
Replies: 19
Views: 9776
stopsareforwimps,

If you're looking for more statistically sophisticated tests than a simple rolling window back test, try reading the reference manual for this R package and the papers referenced in it.
Wed Jun 06, 2012 4:18 pm
Forum: Testing and Simulation
Topic: CTA Performance Index
Replies: 28
Views: 10152
The code for my attempt at a simple mechanical benchmark is here. I chose to go with R code and a function written using the Rcpp package. Feedback on this code would be welcome.
Thu May 24, 2012 4:22 pm
Forum: Data Providers and other non testing software
Topic: Free sources of historical data
Replies: 5
Views: 4502
Thu May 10, 2012 8:53 am
Forum: Testing and Simulation
Topic: CTA Performance Index
Replies: 28
Views: 10152
Further to my above post, I have started coding my benchmark suite but have a conceptual problem. At the moment the benchmark suite consists of 5 basic systems, 2 of which are stop and reverse types and the other 3 can sometimes be flat. The problem is, assuming 1 contract per signal per system only...
Sat Apr 28, 2012 1:55 pm
Forum: Testing and Simulation
Topic: CTA Performance Index
Replies: 28
Views: 10152
Coincidently I have recently been toying with the idea of creating a benchmark suite of typical, mechanical trend following systems. If I actually get around to doing this, I'll post code on my blog and link to it from this thread. The code will be almost certainly be a C++ .oct function for Octave ...
Wed Apr 25, 2012 5:54 pm
Forum: Testing Software
Topic: R Tutorials
Replies: 4
Views: 3959

### R Tutorials

A link for those who might be interested in learning to use R for trading/testing purposes. Knowledge of R will then enable one to use some of the R packages listed here.

Edit 1: and one more useful link.
Edit 2: and yet another.
Wed Mar 14, 2012 9:37 am
Forum: Testing and Simulation
Topic: Testing Results Question
Replies: 17
Views: 6961
I assume you are not using (or properly using) stops. You cant calculate R-multiples, expectancy, avg risk, etc without stops. If the exit (stop) is undefined when (before) an entry signal is triggered, then you have no means to define the risk on the trade or size the position based on that risk. ...
Sat Mar 10, 2012 8:26 am
Forum: Testing and Simulation
Topic: Most of us are using insufficient data for testing
Replies: 25
Views: 10272
If you don't have enough data for back testing, why not create it? On my blog there are two posts, here and here (both with code), which show different approaches to creating synthetic data. By adjusting the input parameters you could simulate different market environments and stress test your tradi...
Wed Mar 07, 2012 8:53 am
Forum: Testing and Simulation
Topic: annual profit target
Replies: 16
Views: 6550
Quote: ...It might use "Robust Annual Return" (as described in Way Of The Turtle ) to measure profits... An even more robust measure would be to replace the linear regression in the RAR% calculation with a repeated median regression (explanation here and here). Some time ago I coded the repeated me...