Live Trading Journal

[nonlinear5]

I've started trading with JBT live today. I'll record my experience
here. I am not going to describe exactly details of the strategies,
but I am willing to say that they are variations of the strategy known
as Defender in the JBT distribution. I will be running three
strategies, each trading one ES contract. All three are rather
conservative, meaning that they settle for small gains and small
losses. All three trade the ES from 10am to 3:30pm EST. From
backtesting results over the last 16 months, the profit factor is
around 3 and the PI is around 6. The number of trades is between 250
and 400 per strategy for that period.

Today's results: ran three strategies, D2, D3, and D4. Ended up with +
$328 for the day. D2 is +$101, D3 is +$126, D4 is +$101. Each strategy
made 5 trades. Slippage was near zero, except for one trade where it
was one tick (position was sold at 1152.75 while the current best bid
was 1153). Out of the 15 trades, 12 were profitable.

The purpose of this journal is not so much to publish the performance,
but to report any technical and logistical issues related to JBT in
live trading. For those of you who are trading live with JBT, please
report your observations here. Thanks.

[John-Crichton McCutcheon]

Are you trading D2 , D3 , D4 Simultaneously on the ES ?

[Gab]

Hi Eugene

Good 1st day, congratulation. Let's hope you'll continue with those gain and never look back with a drawdown. 

Glad to ear you're live trading.

--

Gab

[nonlinear5]

> Are you trading  D2 , D3 , D4 Simultaneously on the  ES ?
>

Yes, each one is trading one ES contract independently and
simultaneously. Effectively, this means that I could be 3 contracts
short, 3 contracts long, or anything in between. In practice, the
holding period is rather short (somewhere around 30 minutes or so), so
most of the time, I am flat on all three.

[new_trader]

CONGRATS to Live Trading!
You finally made it. Keep on rolling...

[John-Crichton McCutcheon]

Thanks for sharing that. Something to think about. Can think of it
as kind of
diversification of Book Signals ? By the way, its always great to see
somebody actually
"put their money where their mouth is" in the investing world.

[Eugene Kononov]

Today's results: ran three strategies, D2, D3, and D4. Ended up with +
$328 for the day.

I looked at my account in TWS today, and it reports my P&L for today as +$340. It turned out that the commissions were $2 per trade, instead of $2.40 per trade as assumed by JBT. Everything else is the same. I'll hit the IB web page to see if this cheaper than I expected commission is a temporary promotion or a permanent deal, and make the corresponding adjustments to CommissionFactory.java.

[nonlinear5]

> I'll hit the
> IB web page to see if this cheaper than I expected commission is a temporary
> promotion or a permanent deal, and make the corresponding adjustments to
> CommissionFactory.java.

I looked it up, and it turned out to be:
0.85 (flat rate) + 1.14 (exchange fee) + 0.01 (regulatory fee) = $2
Reference: http://individuals.interactivebrokers.com/en/p.php?f=commission#futures1
I'll fix it in the next release.

[Keith]

Thanks for sharing, I will take note of the defender strategy and
follow your experiences closely.

[nonlinear5]

October 7 results: +$242. There were 2 trades, both profitable.
D2: 1 trade, +$108. D3: 1 trade, +$134. D4: 0 trades.

[Gab]

On Thu, Oct 7, 2010 at 3:38 PM, nonlinear5 <eugene....@gmail.com> wrote:

October 7 results: +$242. There were 2 trades, both profitable.
D2: 1 trade, +$108. D3: 1 trade, +$134. D4: 0 trades.

Cool. I'm happy for you.

Sorry for being a little off on this question, but how much is required in your account to trade that strategy ?

--

Gab

[nonlinear5]

> Sorry for being a little off on this question, but how much is required in
> your account to trade that strategy ?

The intra-day margin for one ES contract is $2,813. To avoid the
margin call, you should have more.

[Paul]

Do you plan to host your strategies on C2 (for a fee of course)?

On Oct 6, 12:57 pm, nonlinear5 <eugene.kono...@gmail.com> wrote:

[Eugene Kononov]

Do you plan to host your strategies on C2 (for a fee of course)?

I'll see how it goes for a month, and decide on C2 at that time.

[nonlinear5]

October 8: ran D2, D3, D4. No trades. This is quite normal. There were
enough days in the backtesting period without any trades.

[nonlinear5]

The time shifts still occur in recorded data. I thought I fixed them,
but it turned out that they still happen. I am changing my NTP server
to us.pool.ntp.org, too see if this fixes the issues.

[nonlinear5]

> The time shifts still occur in recorded data. I thought I fixed them,
> but it turned out that they still happen. I am changing my NTP server
> to us.pool.ntp.org, too see if this fixes the issues.

I looked at it again and found a bug in JBT. When multiple strategies
are running, JBT creates multiple instances of NTPClock, which may
interfere with one another. I've fixed the bug and set time server
back to ntp2.usno.navy.mil. Everything seems to be normal today. I'll
put this fix in the next release.

[nonlinear5]

Oct 11 results: no trades.

[nonlinear5]

Oct 12 results: no trades.

[Javier]

I assume that you are using 3 different accounts, right? What's the
recommended setup for this? Running each one in a separate machine/
environment?

[nonlinear5]

> I assume that you are using 3 different accounts, right? What's the
> recommended setup for this? Running each one in a separate machine/
> environment?

No, I am using a single TWS account and a single JBT instance, on a
single machine. JBT was designed from the start to support trading
multiple strategies simultaneously.

[Javier]

> No, I am using a single TWS account and a single JBT instance, on a
> single machine. JBT was designed from the start to support trading
> multiple strategies simultaneously.

Then what happens when D1 is long and D2 is short? TWS position will
be 0?

[Florent Guiliani]

They canceled themselves each other...

--
Florent,

[nonlinear5]

There were no trades on Oct 13. On the Oct 14, I am running D2, D3,
and D6, instead of the normal D2, D3, and D4. D6 is parametrized to be
more active than the others.

[nonlinear5]

Oct 14 results: +$84. Breakdown: D6: 1 trade, +$84. D2 and D3: no
trades. Slippage was 0. No NTP clock issues.

[nonlinear5]

Oct 15 results: +$5. Breakdown: D6: 2 trades, +$5. D2 and D3: no
trades. Slippage was 0. Execution time was between 150ms and 450ms. It
was a quiet and stable week overall, from the JBT perspective. Trading
results were in line with the backtested results. The event report
looks clean. I'll probably release JBT 7.10 over the weekend, with
just minor changes, and I expect the 7.10 release to stay unchanged
for a while. Next trading day will be Monday, October 18. I am
planning to stick with D2, D3, and D6. This combination seems to
provide some diversification, although the underlying strategy is the
same.

[nonlinear5]

October 18 results: no trades.

[nonlinear5]

October 19 results: -$545. D2: 2 trades, -$220, D3: 1 trade, -$79, D6:
2 trades, -$246. All 5 trades were losers. Slippage was 2 ticks in my
favor. Executions were between 200ms and 400ms.

[nonlinear5]

Trading D2, D3, and D4 today. Had I stuck to my original plan, my P&L
would have been $400 higher.

[nonlinear5]

October 20 results: 3 trades, +$213. D2: 1 trade, +$58. D3: 1 trade, +
$84. D4: 1 trade, +$71. Slippage was 0. Executions were in between
150ms and 450ms. All 3 trades were profitable. Average time in trade
was about 15 minutes.

[nonlinear5]

October 21 results: no trades.

[nonlinear5]

October 22 results: no trades.

[nonlinear5]

October 25 results: no trades.

[nonlinear5]

October 26 results: 1 trade, -$4. D2: 1 trade, D3 & D4: no trades.
Slippage was 0. Executions were 250ms and 400ms.

[nonlinear5]

October 27 results: no trades.

[nonlinear5]

October 28: I accidentally started JBT in the forward test instead of
the trading mode today. Results: -$12. D2: 2 trades, +$42. D3: 1
trade, -$54. D4: no trades.

[nonlinear5]

I am taking a little bit more risk today and trading Defender3,
Defender4, and Predator3.

[nonlinear5]

October 29: no trades. It was an uneventful week. I am up since I
started live trading, so I am content.

[ShaggsTheStud]

Every day I look forward to reading your reports, it's like I get to pretend I am part of the action. :P

On Fri, Oct 29, 2010 at 4:14 PM, nonlinear5 <eugene....@gmail.com> wrote:

October 29: no trades. It was an uneventful week. I am up since I
started live trading, so I am content.

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

[Kos]

FillPrice Date Commissions Strategy
PnL

1161 10/12/2010 0:21:44 2 JBT7.08 B1
1160 10/12/2010 0:23:54 2
1159.25 10/12/2010 22:50:05 2
1160 10/12/2010 23:11:23 2 12.5 4.5
execution 1100 ms, 1283 ms, 800 ms, 1200 ms

1165.75 10/15/2010 22:34:22 2 JBT 7.09 B1
1172 10/15/2010 23:45:12 2 JBT 7.09 B1
1169 10/16/2010 1:45:06 2 JBT 7.09 B1
1170 10/16/2010 1:52:40 2 JBT 7.09 B1
1168.25 10/16/2010 1:59:11 2 JBT 7.09 B1
1169.25 10/16/2010 2:07:21 2 JBT 7.09 B1
1169.75 10/16/2010 2:13:59 2 JBT 7.09 B1
1171 10/16/2010 2:42:08 2 JBT 7.09 B1 -275 -291
execution 1050 ms, 1284 ms, 1340 ms, 693 ms, 782 ms, 1010 ms, 744 ms,
750 ms


1163.5 10/19/2010 1:58:12 2 JBT7.09 B1
1160.5 10/19/2010 2:13:37 2 JBT7.09 B1
1158 10/19/2010 3:13:37 2 JBT7.09 B1
1156.25 10/19/2010 3:30:00 2 JBT7.09 B1 -237 -245
execution 2100 ms, 681 ms, 694 ms, 639 ms

1179 10/20/2010 2:34:04 2 JBT 7.09 D
1177.75 10/20/2010 2:45:51 2 62.5 58.5
execution 1400 ms, 1226 ms

1181.25 10/26/2010 22:21:56 2 JBT 7.09 D
1181 10/26/2010 22:29:04 2 JBT 7.09 D
1179.25 10/26/2010 2:12:27 2 JBT 7.09 D
1177.75 10/26/2010 2:21:08 2 JBT 7.09 D 88 79
execution 1511 ms, 1222 ms, 1009 ms, 683 ms

[nonlinear5]

Thanks for posting your results, Kos. You may want to attach them as
html, otherwise it's difficult to read. Your executions times are
significantly greater than mince. Where are you located and what kind
of network connection are you using? I also wonder if the longer
execution times mean greater slippage for you. You can determine the
slippage by comparing the "fill price" with the "expected fill price"
in the strategy reports.

[nonlinear5]

November 1 results: 4 trades, -$304. D2: 2 trades, -$33. D3: 1 trade, -
$142, D4: 1 trade, -$129. Slippage was 0. Execution times were between
160ms and 640ms. The event report is very clean: no market data
issues, no NTP clock issues, no connectivity problems.

[Kos]

Defender:
11/01/10 14:36:13.072 EDT -- 1 1182.75 1182.5 2 -- -8.25
11/01/10 14:52:59.757 EDT 1 0 1180 1180 2 -141.5 -141.5
11/01/10 15:15:38.230 EDT -- 1 1174.25 1174.25 2 -- -143.5
11/01/10 15:25:11.711 EDT 2 0 1176.25 1176.25 2 96 -45.5

[Kos]

executions: 2600 ms! 675 ms, 1270 ms, 750 ms

[Kos]

D1 and B1 -no trades, no broken connections, no MarketData issues.
Trades are placed from Singapore using 10 Mb broadband internet
connection of SingTel

[nonlinear5]

> Trades are placed from Singapore using 10 Mb broadband internet

Ok, so this does explains your execution times. Your signal is
traveling at least 10,000 miles to the Chicago Mercantile Exchange.

[Gab]

executions: 2600 ms! 675 ms, 1270 ms, 750 ms

With those execution times, I'd consider renting a server in the US or Canada if you plan to live trade for a long time.

My 2 cents. :-)

[wigginx]

I'd suggest a virtual private server from pair.com (http://
www.pair.com/services/vps/). I used to have a dedicated server with
them and would get very good response times from IB servers. I
recently switched over to the virtual private server instead for the
significant cost savings, but I am still evaluating the performance
difference. It has not been noticeable so far.

If this is something that more people are interested in, I can put
together a description of the configuration tweaks I made in order to
be able to run TWS and JBT in their FreeBSD environment.

-Adam

[Astor]

Adam, please post those tweaks. That will be very helpful.

---

From: wigginx <wig...@gmail.com>
To: JBookTrader <jbook...@googlegroups.com>
Sent: Tue, November 2, 2010 12:14:43 PM
Subject: [JBookTrader] Re: Live Trading Journal

--
You received this message because you are subscribed to the Google Groups "JBookTrader" group.
To post to this group, send email to jbook...@googlegroups.com.

To unsubscribe from this group, send email to jbooktrader+unsub...@googlegroups.com.

[nonlinear5]

November 2 results: no trades.

[nonlinear5]

November 3 results: 2 trades, +$392. This was my best day since I
started the journal.
D2: 1 trade, +$184
D3: 1 trade, +$208
D4: no trades

All the defenders were remarkably subdued, considering the fact that
there was plenty of market action today. I expected a lot more trades.
I'll take a look at the performance chart later to see what happened.
Slippage was 1 tick against me. Executions were between 160ms and
600ms. Average time in trade was about 6 minutes.

[nonlinear5]

I am taking a little bit more risk today. Trading Predator1,
Defender2, and Defender3.

[nonlinear5]

November 4 results: +$484. Made a new all time high in my equity
curve. Predator1 entered long at 10:08, held it until the end of the
trading interval, and ended up with +$484. D2 and D3: no trades.

[Kos]

3/11

B: no trade, D1: 104$, D2: 123$
execution 1500ms, 2400 ms

4/11

B, D1, D2: no trade,

[nonlinear5]

November 5 results: no trades for P1, D2, D3. Very good week overall.

[Gab]

On Fri, Nov 5, 2010 at 3:36 PM, nonlinear5 <eugene....@gmail.com> wrote:

November 5 results: no trades for P1, D2, D3. Very good week overall.

Indeed. I'm glad for you. It gives you a little buffer for when thing won't run as smooth as now (if it ever happens)

Congrats on the good work Eugene. 

--

Gab

[Kos]

6/11:
B: 58.5, slippage 0 and 0.25 on open and close, respectively.
execution 1200 ms, 895 ms
D,D1,D2 no trade

[nonlinear5]

Kos, was that for Friday, Nov 5?

[nonlinear5]

November 8: P1, D2, D3: no trades.

[nonlinear5]

November 9: no trades for P1, D2, D3.

[Kos]

I usually close the next day in Singapore. Here open is at 5th Nov
9-30 pm, close 6th Nov 4 am local time.

[Kos]

November 10: B 108, D1-3 no trades.
Slippage 0, execution 1200 ms , 867 ms

[Eugene Kononov]

On Tue, Nov 9, 2010 at 9:13 PM, Kos <p...@trosy.com> wrote:

I usually close the next day in Singapore. Here open is at 5th Nov
9-30 pm, close 6th Nov 4 am local time.

That's going to be confusing if we decide to compare the results for a given day. I suggest using the same time zone.

[nonlinear5]

November 10: no trades for P1, D2, D3.

[nonlinear5]

Nov 11: no trades for D2, D3. Did not trade P1, as I realized that my
new volume-based strategy is superior. However, it's not ready for
live trading yet.

[nonlinear5]

Nov 12: trading my new volume-based strategy today for the first time,
in addition to D2 and D3. The new strategy is called TensionSeeker1.
In backtesting, it produced about 2.5 times more profit than either D2
and D3, and more trades. However, it's also about 1.5 times riskier,
based on PI and PF.

[nonlinear5]

Nov 12: interesting day today. I ended up with -$58 on 5 trades.
D2: 2 trades, -$8
D3: 1 trade, +$46
PS1: 2 trades, -$96

Slippage was 1 tick in my favor. Executions were from 130ms to 900ms.

[Kos]

D1: 2 trades, -$33
D2: 1 trade, +$46
B: no trades

Executions as usual for Singapore: worst 1620ms to best 620ms.
Slippage is 0 despite long executions

[Eugene Kononov]

D1: 2 trades, -$33
D2: 1 trade, +$46
B: no trades

Looks like your D2 is the same as my D3.

[nonlinear5]

This thread has served its purpose. I am happy with JBT performance in
live trading. It seems quite stable and is functioning as designed. I
have not experienced any market data issues, or any order execution
issues. The overall slippage is near zero. The NTP timer is reliable.
Simultaneous running of multiple strategies works well. The
orchestration of events appears to be solid, as well: no events seem
to be missed or ignored.

I am still going to maintain this thread, but updates will be provided
on a weekly basis, instead of daily.

[Astor]

Eugene, now that the JBT performance is so good, can you update Jarbitrager or add spread trading capability to JBT?

---

From: nonlinear5 <eugene....@gmail.com>
To: JBookTrader <jbook...@googlegroups.com>
Sent: Mon, November 15, 2010 7:52:03 AM

Subject: [JBookTrader] Re: Live Trading Journal

[nonlinear5]

> Eugene, now that the JBT performance is so good, can you update Jarbitrager or
> add spread trading capability to JBT?

Astor, I know that you've been advocating merging the capabilities of
JA and JBT into a single framework, but I believe they are
conceptually different enough to evolve independently. Over the life
cycle of JBT, I've quite religiously resisted the functionality which
I believe to be "bells and whistles". That includes support for other
brokers, database storage, command line and batch processing,
portfolio management, and various third-party packages for web
console. You can browse these features here:
http://code.google.com/p/jbooktrader/source/browse/#svn/tags/jbooktrader-fg-0.3

I am a big proponent of simplicity. Reducing the JBT code base makes
me happier than adding to it. I'd much prefer to have JBT defined and
developed in a narrow scope, with the small, compact, and well-
understood set of features. In the end, my purpose is not to maximize
the user base, but to maximize the code efficiency in the sense that
would allow me to use JBT to accommodate my trading strategies with
the smallest effort required to maintain and evolve the code base.

Now, I fully understand that everyone's approach to trading is
different. It may be well be the fact that combining spread trading
with the limit book trading is beneficial. However, until proven so, I
am not willing to put my efforts into it. I hope you stick around,
Astor, as you contributions to both JBT and JA have been valuable.

[Astor]

Eugene, it is hard to prove the benefit of combining spread trading with book trading without actually doing it.  I think spread strategy can benefit from some of the indicators available in JBT and vice versa. It does not make sense to keep the two types of approaches separate, because they are complimentary. All JBT indicators zoom in on market sentiment. Spread trading indicators zooms in on relative valuation. When both types of indicators say "buy", the security is both, undervalued and market sentiment is turning in its favor - should be a far more consistent performer than separately.

 

Ability to maintain a market neutral exposure is also very important.

 

In any event, I will stick around.

---

From: nonlinear5 <eugene....@gmail.com>
To: JBookTrader <jbook...@googlegroups.com>

Sent: Mon, November 15, 2010 9:06:49 AM

Subject: [JBookTrader] Re: Live Trading Journal

[ShaggsTheStud]

The best thing for Eugene to do is to work on JBT in the form that he likes it.   If he no longer likes working on it anymore, then his progress will stop.  I don't think he will be upset if anyone uses it as a base for some other variant (although you should give proper credit, as this code that he has given us is quite a gift!)  You just have to be willing to get your hands dirty.

On that note, I am now putting out a formal challenge for someone to fork the project to handle multiple instruments! Who has what it takes to see that through from start to finish?  I know I have the skill, but I personally do not have the time.

To unsubscribe from this group, send email to jbooktrader...@googlegroups.com.

[Eugene Kononov]

On Mon, Nov 15, 2010 at 10:15 PM, ShaggsTheStud <shaggs...@gmail.com> wrote:

I don't think he will be upset if anyone uses it as a base for some other variant (although you should give proper credit, as this code that he has given us is quite a gift!)  You just have to be willing to get your hands dirty.

Indeed. JBT source code is free to you all to use as you please, be it commercial, educational, or another spin-off open source project. My preference is, of course, to keep it all in one place, and to improve it with everyone's feedback and participation. The bottom line, however, is that it's my trading system which I happen to share with you. I get free QA from many different people running JBT, and that worked spectacularly well for JBT and me as the user. You get the codebase which would have probably taken you a couple of years to develop from scratch, or a couple of hundred thousand dollars to develop if you hired someone. That's how I view our contract.

 

[Kos]

I just wonder why all this discussion goes into Live Trading Journal
thread? I think it's valuable to check live performance of JBT,
especially in the turmoil market, which I am will duly come.
16/11
B -16
D1 -66
D2 +46
I am doing these and all other strategies (incl my) in simulated. The
results are not the same. In simulated B=-60, D1=-66, D2 =0
execution from 600 to 1200 ms, slippage 0-1 ticks.

[nonlinear5]

> D1 -66
> D2 +46

I had these results:
D2: -$16
D3: +$154
D4: +$34

When you optimized your Defenders, which selection criteria did you
use? I almost always use PI.

[nonlinear5]

Week of Nov 15 to Nov 19: 4 trades, +$172. Defenders continued to be
very selective.
D2: 1 trade, -$16
D3: 2 trades, +$154
D4: 1 trade, +$34

[nonlinear5]

There are shortened sessions for the ES on Nov 25 and Nov 26, so I am
not going to trade on those days. That means I am done for the week of
Nov 22 to Nov 26 with these results: 5 trades, +$368. All 5 trades
were profitable.
D2: 3 trades, +$176
D3: 2 trades, +$192
D4: no trades

The incidents included one day when I lost power for half a day and
the other day when slippage was unusually large (0.75 of a point on
one trade). Aside from these issues, it looks good so far.

[new_trader]

> Nov 22 to Nov 26 with these results: 5 trades, +$368. All 5 trades
> were profitable.
> D2: 3 trades, +$176
> D3: 2 trades, +$192
> D4: no trades

this is a really great performance!
so it seems that your strategies are able to make about 400$ per week,
~$1.600$ a month.

have you thought about increasing position size per trade?
with these reliable results that we have seen the last weeks your
profit can scale directly with position size.

[Eugene Kononov]

have you thought about increasing position size per trade?
with these reliable results that we have seen the last weeks your
profit can scale directly with position size.

Since D4 is very inactive, next week I'll trade 2 contracts with D2 and 1 contract with D3.

[nonlinear5]

Results for the week of Nov 29 to Dec 3: 1 trade, -$66.

[nonlinear5]

Results for the week of Dec 6 to Dec 10: I was out of town most of the week, so I only ran JBT for two days that week. There were no trades.

[Alexana]

Any results for January?

[nonlinear5]

I was running JBT live every day, but there were no trades since December 3 last year. Last weekend, I decided to re-optimize for the recent market period, and this week I've been running two of these re-optimized strategies. There have been some trades. I'll post this week's results on Friday after the market close.

[nonlinear5]

Results for the week of January 24 to January 28: 10 trades, 70% profitable trades, +$311. I was running two new strategies, optimized for the period of the last four months, as explained in my previous post. Here and in the future, I'll refer to these strategies as ActiveDefender2 and ActiveDefender3.

[nonlinear5]

Results for the week of January 31 to February 4: outstanding week, 8 trades, 100% profitable trades, +$1018.

DefenderActive1: 2 trades, +$267
DefenderActive2: 4 trades, +$434
DefenderActive3: 2 trades, +$317

[Vlad Palnik]

Eugene,

Are the trades live or paper account??

Vlad Palnik

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

--
Vlad Palnik

vlad...@gmail.com

(360) 624-5422

[nonlinear5]

Are the trades live or paper account??

All the trades and the results posted in this thread are live.

[Victor Martin]

Eugene, what is your trading schedule?

I'm running in a real account a non-customized Defender strategy from 9am - 5pm EST, optimized for last few months, but Net Profit for this week was slightly negative.

Did you make any modification to Defender strategy for achieving this profits?

Keep up the good work!

On Sat, Feb 5, 2011 at 1:12 AM, nonlinear5 <eugene....@gmail.com> wrote:

Are the trades live or paper account??

All the trades and the results posted in this thread are live.

--

[Eugene Kononov]

My trading schedule is the same as before, 10-15:30 EST. All three Defenders are exactly the same strategies as before. Optimization period was October 6, 2010 to January 2011. I selected three "sweet" spots from the optimization maps, which made my three Defenders. I'd definitely recommend excluding the early and the late trading session periods from your schedule. What selection criteria  do you use when optimizing?

[Eugene Kononov]

For reference purposes, see the attached backtest results for the Defenders optimized from Oct 6 2010 to Feb 4 2011. When optimizing, you should probably seek these (or better) results.

[Eugene Kononov]

Minor correction in regards to the backtest results in my post above: the optimization period was from Oct 6 2010 to Jan 21 2011. The backtest period is from Oct 6 2010 to Feb 4 2011, so it includes 2 weeks out of sample.

[Eugene Kononov]

My trading schedule is the same as before, 10-15:30 EST. All three Defenders are exactly the same strategies as before. Optimization period was October 6, 2010 to January 2011. I selected three "sweet" spots from the optimization maps, which made my three Defenders. I'd definitely recommend excluding the early and the late trading session periods from your schedule. What selection criteria  do you use when optimizing?

I've got to make another correction. The exit code did in fact change in my new Defenders. To avoid any confusion, I am attaching the strategy source code. I did leave out optimization parameters, so it's up to you to find the set of parameters that fits your risk tolerance.

[Victor Martin]

Thank you, I'll try this strategy how it performs with a reduced trading schedule

El 05/02/2011, a las 17:58, Eugene Kononov <eugene....@gmail.com> escribió:

My trading schedule is the same as before, 10-15:30 EST. All three Defenders are exactly the same strategies as before. Optimization period was October 6, 2010 to January 2011. I selected three "sweet" spots from the optimization maps, which made my three Defenders. I'd definitely recommend excluding the early and the late trading session periods from your schedule. What selection criteria  do you use when optimizing?

I've got to make another correction. The exit code did in fact change in my new Defenders. To avoid any confusion, I am attaching the strategy source code. I did leave out optimization parameters, so it's up to you to find the set of parameters that fits your risk tolerance.

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

<DefenderActive.java>

[Klaus]

I tried to integrate it, but it gives me a syntax error.
getcurrentPosition unknown.

There is, however, a method getPosition. Is this the right one?
(I am wondering also how this mismatch of names can be?
Are you not using the current version?)

>  DefenderActive.java
> 3KViewDownload

[Eugene Kononov]

On Sun, Feb 6, 2011 at 2:43 PM, Klaus <klaus.s...@googlemail.com> wrote:

I tried to integrate it, but it gives me a syntax error.
getcurrentPosition unknown.

There is, however, a method getPosition. Is this the right one?
(I am wondering also how this mismatch of names can be?
Are you not using the current version?)

Yes, I made some changes to framework and forgot about it. I'll publish a new version sometime next week. In the meanwhile, you can use getPosition() instead of getcurrentPosition() in that code attached in my previous post.

[Paul]

Can someone send me today's data (Feb 10th)? My connection crashed
prior to the open.

On Feb 6, 2:47 pm, Eugene Kononov <eugene.kono...@gmail.com> wrote:

[nonlinear5]

What crashed, TWS, JBT, or your network? If it's JBT-related, I'd like to know. I'll post the Feb 10th  data after the market close today.