JBookTrader 8.07 release

[nonlinear]

JBookTrader 8.07 has been released.

Latest release download: http://jbooktrader.googlecode.com/files/JBookTrader-8.07.7z
Release Notes: http://code.google.com/p/jbooktrader/wiki/ReleaseNotes
User Manual: http://docs.google.com/View?id=dfzgvqp4_10gb63b8hg
Historical market data sample: http://jbooktrader.googlecode.com/files/ES-sample.7z

Changes in this release:

--  Improved the methodology of indicator initialization

--  Fixed date scroll issues in the charts (thank you, Judson)

--  Optimization results are now sortable by table column (thank you, Judson)

--  Fixed the problem associated with the 2104 error code (thank you, RV)

--  Staring from this release, I no longer package my best-performing strategies with JBT. However, these strategies can be purchased:
    https://docs.google.com/document/d/1cYjooxpxqUftqkvUWkb3CBlTFyQSx2jmi16KDVXeykk/view

[Javier]

> --  Staring from this release, I no longer package my best-performing
> strategies with JBT. However, these strategies can be purchased:

Nonlinear, for how long did you traded the 11 strategies live?

[nonlinear]

Nonlinear, for how long did you traded the 11 strategies live?

Every day for the last 6 months.

[Allen Whitt]

I should know the answers to this, but I don't remember.

• The strategies' results are given for backtest, but you also ran them live. What were the per-strategy results of running them live?
• Does the 'trades' column represent contracts traded or entry/exit points?
• If the latter, what was the max position achieved during the period?

On Mon, Jan 16, 2012 at 3:14 PM, nonlinear <eugene....@gmail.com> wrote:

Nonlinear, for how long did you traded the 11 strategies live?

Every day for the last 6 months.

--
You received this message because you are subscribed to the Google Groups "JBookTrader" group.

To view this discussion on the web visit https://groups.google.com/d/msg/jbooktrader/-/CvQvFV7u3q8J.

To post to this group, send email to jbook...@googlegroups.com.
To unsubscribe from this group, send email to jbooktrader...@googlegroups.com.
For more options, visit this group at http://groups.google.com/group/jbooktrader?hl=en.

[Eugene Kononov]

• The strategies' results are given for backtest, but you also ran them live. What were the per-strategy results of running them live?
• Does the 'trades' column represent contracts traded or entry/exit points?
• If the latter, what was the max position achieved during the period?
  

The "trades" column represents the number of trades made by the strategy during the specified time period. The "trade" is counted as the round trip completes. For example, enter long and subsequent exit count as one trade.

Each strategy always trades one contract. So, if it's a long-only strategy, it can be either flat or long one contract at any point in time. If it's a short-only strategy, it can be either flat or short one contract at any point in time. All strategies that I trade are flat most of the time, as illustrated by the frequency of trades and trade duration, which, on average, happens to be around 20 minutes (that is, the average time between the entry and the exit is about 20 minutes). The time separation between the trades is about 2 days, on average. This effectively means that the typical strategy is in the market about 2% of the time. The remaining 98% of the trading hours are spent waiting with a flat position. To put it in more simple terms, the typical strategy is very patient to enter the market, and very quick to exit, which I believe to be a very good quality.

In regards to live performance, I'd note that I continuously and periodically re-optimize the strategies with the most recent data, so I don't keep the "out of sample" stats which would tell me what would happen if I traded a fixed strategy for extended periods of time. However, by definition, live trading is "out of sample", and what I found is that the results are satisfactory, although not as good as the back-tested results.

One thing that I am working now is to address the very uneven distribution of trades. When I look at the performance charts over the period from October 2010 to January 2012, almost all trades occur during the period of time between August 2011 and November 2011. Sometimes weeks and weeks pass without a single trade, although market makes significant price moves. It would be good to be able to capture these moves, which I am trying to model using volatility, but I have not been able to do it successfully without sacrificing the performance in the other time periods. I've attached my typical equity curve for illustration.

 

[Javier]

>> Nonlinear, for how long did you traded the 11 strategies live?
> Every day for the last 6 months.

What's the delta between the optimization results and the satisfactory
real performance that you got during the last 6mo?

> One thing that I am working now is to address the very uneven distribution
> of trades

I've concluded the same, the strategies work well in range/downtrend
markets with medium/high volatility but don't take trades in uptrend
markets with low volatility (e.g. Sep'10 to Mar'11). I'm thinking in
adding some volatility/volume filter to achieve an even distribution,
any other ideas on this?

[Judson Wilson]

I suspect that you simply won't make a lot of trades with these systems in low volitility, because they are count-trend, if I remember correctly.

I would suspect that trades can start going against you in a big way in low volatility if the volatility picks up big - something that the volatility compensation would need to adopt for.

--
You received this message because you are subscribed to the Google Groups "JBookTrader" group.

[Eugene Kononov]

On Sun, Jan 22, 2012 at 3:37 PM, Judson Wilson <wilson...@gmail.com> wrote:

I suspect that you simply won't make a lot of trades with these systems in low volitility, because they are count-trend, if I remember correctly.

I would suspect that trades can start going against you in a big way in low volatility if the volatility picks up big - something that the volatility compensation would need to adopt for.

My previous set of strategies performed well in the low volatility environment. In the high volatility environment, that set of strategies was starting to lose money. My new set of strategies was optimized on the longer period which includes both low- and high-volatility market, and it performs well. The drawback is that in the low volatility, these new strategies don't take trades at all, even though the market makes significant price moves. I only had a few trades during the last two months. So, the challenge is to make a strategy "volatility-aware", so that it can be traded in diverse market conditions.