1. I started this thread to keep track of all postings and discussion related to the paper trading Capitaland using TurtleFarm.

2. Here is the link to oringal thread that discuss turtle trading, which you may find helpful if you need some background information about turtle trading system:
http://z7.invisionfree.com/ChartistsUnited...dpost&p=1820292

3. TurtleFarm is a trading system using Turtle trading rule. Details see:
http://www.technical-analysis.com.sg/prodTurtle.html

Immediately after the launch of TurtleFarm, I play around with the software and data, and made the following post on Dec 19 2006, 08:51 PM
-------------------------------------------------------------------------------

This chart is produced using TurtleFarm, which signals an exit today.

Capitaland (2006-12-19 Closing Price $5.70)



Today it closes the 4 buy orders issued some 3 months ago.

Since I don't use the defualt setting by TurtleFarm, and make a few slight adjustment that I think will be more applicable to SGX stock trading (Long only, increase Notional Account % Sub limit), I think it is only useful if I post my modified settings. I did that on : Dec 20 2006, 12:21 AM
-----------------------------------------------------------------------

...

I have changes the Turtle parameters: see the snapshot below:



The purple line is the total profit and lose of that trade (in this case, 4 pyramid positions), after exiting the trade. It stands UP at $9119.88.

The blue line is the profit and lost calculated based on the closing price; and red line is the profit and lost if exit at the trailing stop loss.

N is used to size the position, and 20 is the parameter (ATR's period). CSK please correct me if I am wrong.

And I added the following on Dec 20 2006, 12:30 AM

-----------------------------------------------------------------------------------

Let me add the summary and the equity curve for those who are interested.

Since I like (more on that later) the result I saw in Capitaland, I decided to do a paper trading on this counter, and posted on Jan 10 2007, 10:36 PM when TurtleFarm did the first entry

----------------------------------------------------------------------------------------

Capitaland (2007-01-10 Closing Price $6.25)

TurtleFarm did an intraday entry today.

And posted here on Jan 10 2007, 10:36 PM when the second entry was added.

---------------------------------------------------------------------------------------------

Capitaland (2007-01-15 Closing Price $6.65)

TurtleFarm added another position today.

And lastly, the third entry when added:

---------------------------------------------------------------------------------------------

Capitaland (2007-01-15 Closing Price $6.65)

TurtleFarm added yet another position today. Total 3 positions so far, and one more pending price move above 6.801.

Current position size mark to market closing price 6.65 is $26,194.35. Trailing stop at 6.306, putting total $1347.71 at risk.

Question for TurtleFarm trader: will you add another one position if 6.801 triggered?

(Please ignore the lot size requirement and the tick size is $0.05 for the time being, I am trying to do a real-time paper trading before doing adaption to SGX context.)





Here are my modified rules for those who are interested:

CSK has responded to my posting on 3rd entry:

------------------------------------------------------------------------------------------

"will you add another one position if 6.801 triggered?"

This is the last pyramid unit. If the trend goes into swing, this last unit can
potentially increase the position profit by 1/3. Of course this will jack the stop
higher by about $0.10 to 6.405 (from 6.306). The risk of the 6.405 stop loss being
hit is higher than the 6.306 stop. Position loss if stopped out is $1295.92, vs
1165.95 if this last unit is not added. Weigh the possible reward (if trend develops)
vs the risk if stopped out.

The Turtle pyramid strategy is to add to winners, never add to losers. If a position
doesn't act right, the stops take care of it. The total position size is divided into 4
equal tranches of 25% each.


Note: I just realize a display bug in the New Orders Alert Window. Just a display
bug that occurs sometimes and not always with no effect on the orders although
the display can be confusing. I just fixed it. Will be in ver 1.0.2 This is the
corrected display:

Followed up by mine:
---------------------------------------------------------------------------------------------


Since CSK is on this topic, I might as well add some of my observations when doing paper trading using TurtleFarm:

1. The allocated size required in this case in 1313 shares, which does not fit SGX's size requirement. Since there are a potential total of 4 entries of 1313 shares, or final total 5252 shares, I am thinking to do 2000 shares each for first and second entry, and subsequently 1000 shares each for third and fourth entry. My rational is trying to keep the average cost of the whole position lower - if all four positions are entered finally. If not 4 are triggered, I get stop out at 1/4 or 1/2 of total intended size, the stop is not that painful (I don't intend to change the trailing stop).

2. The price of 6.801. When doing this round of paper trading, I realised that the calculated entry price of 6.801 is just a little higher then 6.80. My initial reaction is to round it down to 6.80 (or 6.50, 6.60 & 6.70 for the first, second and third entry). But after observing the entry timing, I think we should round up to the next tick, otherwise we may end up in a situation that we are in the trade but TurtleFarm are not if price just kiss the rounded down entry price and retraced.

Any comment?

And Undead's:

------------------------------------------------------------------------------------------

I am a little surprised HC chose CapLand to paper trade.
Because I ran a trial run on CapLand as well (and being lazy, only Capland).

My observation for 3 periods

-Oct 94 to March 97 (Sideway Market)
I did not expect a trend following system to win.
turtle did what was expected, flat result

- March 97 to Oct 98 (One way down and we can only BUY stocks)
I was wondering if the stops will work,
and the stops work.

- April 03 to Now (One way up)
I was hoping to hit jackpot,
unfortunately the result is really not satisfactory.

Perhaps CapLand is not suitable,
and I do not expect turtle to work on everything.

But I am surprised HC chose Capland!

Undead asked: Why Capitaland?

Answer:

1. I am lazy. :) Just pick anyone that come to my eye.

2. Though I did not do a system testing on all SGX stocks, but I choose Capitaland to paper trade for a few reasons:

2a. Since I am talking about trading stocks, and turtle trading is holding the positions for quit some time (at times months), I don't wish to touch those stock that is of poor fundamentals, reducing risk in case that company goes under.

2b. I think I need a certain liquidity, in term of daily trading volume. I have not yet come out with any rule on this, but I think Capitaland is liquid enough.

2c. Equity Curve of Capitaland looks nice to me: quite smooth and upward moving. I will post one on Pacific Andes, which I also did another trial using TurtleFarm, the swing in equity curve is too wide to my liking.

2d. The up-optimised performance report of Capitaland cover quite a time zone, from 1998 till now. So I think we are not just talking about the current bull run since 2004. The system, to me is quite well behaved.

2e. I am still eyeing on a better candidate for paper trading. Care to point me to one?

I welcome all comments.

Here is the performance report generated by TurtleFarm base on 2007-01-16's data:

Equity Curve


Performance Summary


Trades Done


Monthly Results


Quarterly Results


Year Results

In fact I had posted another paper trading on Pacific Andes:

http://z7.invisionfree.com/ChartistsUnited...dpost&p=3896651

But the look of the equity curve make me uncomfortable:

Capitaland (2007-01-17 Closing Price $6.85)

Today TurtleFarm added the fourth position.

---- Deleted -----

Capitaland (2007-01-18 Closing Price $7.00)

This update included the data window, which shows various value as of today's closing.

Notice that the trailing stop is now at 6.406, so to adapt to SGX's tick size, I will do an exit at 6.40. Rounding down is to prevent I prematurely exit the trade while TurtleFarm still in it.

I can't help but also notice the profit marked to market is $1.8k, but to exit at the above trailing stop, I will have to see the whole position going from profitable stage into losing stage, and incurring a loss of $1.3k.

Can TurtleFarm trader sits still and sees that happen?

Capitaland (2007-01-26 Closing Price $6.65)

Chart updated.

TurtleFarm trader: Notice now mark to market profit has all evaporated, while trailing stop still keeping the whole position. Are you tempted to "outsmart" the system buy/sell signal? How about scaling in and out to take advantage of volatility?

I can feel calm when doing this paper trading. But not sure if dealing with real money, especially on the part that sees S$2k + profit just disappear, and I have to take a loss when the system call for an exit.

Capitaland (2007-01-29 Closing Price $6.90)

Trailing stop now kick in at 6.406.

Someone commented to me that using turtle trading system to trade SGX stock now is not appropriate, reason is we are likely at the tail end of the bull run where a lot of volatility is expected. To make thing worse, my setting in TurtleFarm is long only.

That is exactly my intention of doing a paper trading at this time: doing a stress test to the system (and the TurtleFarm trader).

We cannot tell for sure how long and how much more the market will run. If we are impressed with the stimulated result of the trading system and intend to use it to trade real money, we will have to take its signal as and when it is generated. We don't have the privilege of foresight that a long trending bull run will follow the current entry. The trade may or may not work, but the systems supposedly to keep the profit running and loses capped.

It is also under this seemingly unfavorable stage of market, the psychological reaction of TurtleFarm trader will the tested.

(I am now holding on to a comment, waiting for the trailing stop to move up and exit this position at a profit. Stay tuned.)

"Someone commented to me that using turtle trading system to trade
SGX stock now is not appropriate, reason is we are likely at the
tail end of the bull run where a lot of volatility is expected."

When the Turtles were taught the system, they were specifically told
not to use any technical indicators. Explanation were given. This is why
Analyst's TurtleFarm do not have any technical indicators that may
confuse and affect the Turtles' discipline. They were also taught not
to pre-empt the market but to be reactive. So to trade with the
pre-conceived expectation of a tail end may affect the required
discipline to follow the system, not just the Turtle system but any
trading system.

"To make thing worse, my setting in TurtleFarm is long only."

It is not easy to short stocks on SGX, unlike in the US. Can still use
Long Put warrants to take a short stock position. Of course this cannot
be backtested because of the short life-span of structured warrants.

It is unfortunate that the SGX Single Stock Futures did not take off.
Otherwise, they would be a better trading vehicle than the structured
warrants which are initially priced with ridiculously high Implied
Volatility that served to line the issuers' pocket with money in the
first instance - a real fools' market. US stock options will be better.

"... likely at the tail end of the bull run where a lot of volatility is expected. "

The Last Profitable Trade Filter was designed to take care fof this.

"It is also under this seemingly unfavorable stage of market, ..."

The end of trend is not an unfavourable stage. This will not be a problem.
End of trend means trade exit. The problem is a sideway market.

"We cannot tell for sure how long and how much more the market will run."

Yes, agree. Therefore, the trailing exit handles this aspect of the trade
management.


Trends in the futures market start and end. The more the better. More
because more chance to make money. Better because money made.
The problem is a sideway market.

The problem, if there is one, of end of bull trend in the SGX stock market
is an individually created one. The problem does not exist. Singapore
stock traders must start to look at and accept shorting strategies instead
of cursing and swearing at people who even talk of shorting.

If there is a downtrend in SGX stocks, the SiMSCI futures traders will be
just as happy as when there is an uptrend. So too the TiMSCI and Nikkei
futures traders. So there is actually no problem with a bear stock market.

Capitaland (2007-02-04 Closing Price $7.10)

Lion had posted at the following URL a word of caution about the last bar is a doji of long upper shadow. I too had noticed that, but when doing this system trade, such observation has to be ignored, however uncomfortable I felt.

Maybe this why someone commented that those who are new to the market and without and TA background are more likely to be successful in system trading.

http://z7.invisionfree.com/ChartistsUnited...dpost&p=3998648

I guess before we can judge, we need to understand some of the numbers.




From my understanding,

Cumulated Closed Profit and Loss (118,081.66)
This of course tells you whether the system makes or loses money.
Let’s be realistic,
profits is the main motivation for trading.
However let’s look at other statistics,
they will tell us if the system suits us

Total Number of Closed trades (251)
To perform any meaningful statistical analysis,
we will need a large enough sample size.

Percentage of Winning Trades (51%)
It will tell us if our system captures profits regularly.
For a trend following system, a number above 50% is good.

Maximum Drawdown (-7188.88)
This number gives you some indication of the minimum capital
that you will need to trade the market.

For example, if you can only risk $100,000,
a maximum drawdown of $200,000 will put you out of business.

Average Trade (470.44)
This will give you an indication how easy
or difficult it is to make money with this system.
Don’t forget we have not factor in commissions and slippages.

Return on Account (59.04%)
I prefer to look at annualized return.
Else it would be difficult to compare results.

A simple conversion like
<[(1+0.5904)^(1/19)]-1>*100
will be nice.

Total Commissions Accounted and Total Slippage Accounted
Many systems are profitable before transaction costs.
For active systems, commissions will be an important consideration.
And what is slippage?
It is difficult to answer without further information.
For example what is the market volatility,
or can I buy a million shares at $6?

Profit factor (4.27)
A good value will indicate that the equity curve is smoother.

Average Number of Bars in Winning Trades (34)
For a similar system with similar risk/reward,
the lesser time in the market will be better

Others
Of course this is not all.
Personally I will calculate the recovery period,
how long will it take to recover from the drawdown.

For more complicated summary,
I will also have to judge the impact of the new strategy
on my existing strategies.

Capitaland (2007-02-12 Closing Price $7.30)

Today TurtleFarm moves the trailing stop up to $6.599, making the the trailing exit, if called for, to make this trade to be closed in a profitable (slippage and round down ignored) situation. From now on, it is a matter of how much money will the TurtleFarm trader made when exiting this position.

It is of course very comforting to know this fact, after some having a possibility of exiting this trade in a loss for about a month. Having in this type of sure win trades (unless there is gap down, touch wood :) ), TurtleFarm trader should have his mind in peace, and following the system generated trailing exit signal from now on should be easilier.

Capitaland (2007-02-21 Closing Price $7.70)

Chart updated.

Capitaland (2007-02-23 Closing Price $7.85)

Even though this is just paper trading, I can feel that I get "happy" to see now the trail exit profit is $2,085.04, or arround 6% from the initial total cost of $34,936. This "happiness" is coming from the fact this this trade will at least made this amount of money.

Capitaland (2007-02-28 Closing Price $7.00)

Today is a bad day for SGX counters. Capitaland gapped down and open at 7.00, Turtlefarm close all positions at opening price since it is lower than the trailing stop. With the closure, the total profit of these trades added up to $1822.44.

Looking at the price chart, I ponder on the issue of exit at gap down open. SGX don't allow us to key in a order on condition of certain price is met, so as Singapore shares trader, we can only look at the price screen and react. From today's intraday price action, I think if I will to exit this trade, most likely will be $6.95, a few minutes after market open. The slippage will more more than that $0.001 as I set in TurtleFarm. Closing this trade at $6.95, it would cost me another $257.35, reducing the total profit before brokerage to $1565.09.

Capitaland (2007-03-27 Closing Price $7.80)

I am not sure if the buy entry will be triggered soon, so I just post here an updated chart.

Capitaland (2007-03-29 Closing Price $8.00)

TurtleFarm initiates the first entry today.

-Deleted-

Frankly speaking, I don't like this entry now.

It looks to me more like an end season window dressing that triggers the buy signal. But when doing system trading, it is not a consideration.

"... But when doing system trading, it is not a consideration."

Yes, this is important.

There was a very important point taught to the original class
of Turtles. This very important point is NOT in the Original
Turtle Rules that Curtis Faith released as a free project with
the approval of Richard Dennis.

This important point has never been made public. But I will
do it here. This revelation is the first ever to be revealed to
the public.

Once you understand it, it makes you understand the Turtle
Trading System much better. It makes you want to go to the
Performance Report in Analyst's TurtleFarm to look at both
Winning and Losing Trades better. It makes you understand
why some trades are very very important but at trade initiation
you do not know they are that important. It makes you
understand why you will have Losing trades and yet take
them in your stride.

Here it is:

There are two types of errors:

Type 1 error: Losing trades.
All system will have losing trades. This is unavoidable.

Type 2 error: Missing out a good trade.
Trend trading system (the Turtle system is one) rely on that
few trades in the year that bring it in the BIG profit. At the
time of trade initiation, we do not know which are the good
ones. We only know when it is running and running and the
trailing exits are showing more and yet more profit. These big
profit trades are the outliers in a trend trading system. They
cannot be missed. Missing them is Type 2 error and is fatal
to the system.

I am writing my own document of the Turtle Rules and this
very important point will be included. I had thought it would
be easy and quick to write the document but realise it to be
otherwise.

Capitaland (2007-03-30 Closing Price $8.00)

2nd entry is made today.

Capitaland (2007-04-04 Closing Price $8.15)

3rd entry today.

Trader's opinion about the stage and direction of the market does not have a place in system trading. This is where system trading shines, it supposed to take away the judgemental/discretionary part from the trader, and replace it with systematic entry and exit rules.

In the previous trade, someone commented that TurtleFarm's entry was near the tail end of bull run, but the system generated a good profitable trade out from it. Those traders who think that it is at the tail end of bull run and did not trade had missed a profitable opportunity.

Looking at the price action of Capitaland, I think it is likely that my opinion about end season window dressing is wrong this time; it is likely that TurtleFarm has beaten me this time.

Though this trade is yet to be closed.

Capitaland (2007-04-09 Closing Price $8.55)

4th (last) entry is made today.

Capitaland (2007-04-13 Closing Price $8.15)

Chart updated.

Capitaland (2007-04-19 Closing Price $8.05)

TurtleFarm exited all four positions today, incurring a loss of $255.77, before brokerage.

From this (2nd) trade, I noted the importance of keeping moving the trailing stop up as protection. The more difficult part, is how (much) to move.

Capitaland (2007-04-20 Closing Price $8.25)

After knowing DJ closed up 100+ point, 2007-04-19's exit now looks premature. But when exiting on 2007-04-19, we do not know how DJ will perform on 2007-04-19.

And if Capiland's bull has enough power, there is an re-entry $8.70.

Capitaland (2007-04-27 Closing Price $8.60)

TurtleFarm does the 1st entry (of 3rd case study trade) today.

Capitaland (2007-05-07 Closing Price $8.35)

TurtleFarm up the trailing stop to 8.299.