FOREX QUANT

May 28, 2008

Intraday Volatility Breakout System

Filed under: Uncategorized — by TraderMade @ 4:50 pm
Tags: , , , ,

Following a successful forward test on the GBPUSD conducted by our friend in Bali, we then realized that this particular system must had a good potential for several other markets with high volatility behavior. The first thing that came up in mind was the Hangseng index. Then we asked for a demo account from our close friend at Topgrowth Futures, where we were once teamed up in developing trading systems and trained their traders last year. We really like their “Pro-iTrading” platform. It’s very responsive, with good spread and commission for trading the Hangseng, very user-friendly interface, making it an ideal platform for running an intraday trading strategy.

And as expected, the system has been going well (so far) in this market. We’re going to continue the test and watch how it will perform. Good luck to us.

(*)

May 23, 2008

Beautifully Recurring Pattern

Filed under: Uncategorized — by TraderMade @ 11:25 pm
Tags: , ,

One thing that keeps occurring repeatedly in the fx market is: a beautifully recurring volatility pattern.
Yeap, it’s soooo beautiful.

(*)

May 3, 2008

Credit Hedge Arbitrage

Filed under: Uncategorized — by TraderMade @ 7:29 am
Tags: , , , ,

After a successful acquisition of a multifinance company in the year of 2006, and heavily restructured the company along 2007, Risenberg launched its first credit hedge arbitrage program. The purpose of the acquisition was to provide an efficient vehicle for Risenberg to develop and run its arbitrage system. The acquisition was a deal worth IDR 10 Billion (USD 1.1 Million). Risenberg partnered with another group of investors in the deal and had been approved by the Department of Finance of Indonesia.

By participating in the program, corporate investors are able to gain an opportunity in an alternative asset class that has a very low correlation to the traditional market (ie. stock and bonds). The program is able to deliver a fixed 11% – 14% annual return to participants. The program exploits a continuous arbitrage opportunity that exclusively accessible only by a small number of players in Indonesia’s interbank money market. This is a close ended investment program, will only be available until a IDR 50 Billion (USD 5.5 Million) is reached.

For more info on the program, please contact Risenberg at www.risenberg.com

(*)

April 17, 2008

Intraday Volatility Calendar

Filed under: Uncategorized — by TraderMade @ 5:49 am
Tags: , , , ,


http://www.forexfactory.com/calendar.php

This is an example of an economic calendar that I prefer calling it as “Intraday Volatility Calendar”. The default time stamp is in Eastern Time, so do adjust it to local time to get a better sense of it.

This is where i consider some fundamental analysis- to some extent- is acceptable, that is for volatility anticipation alone, not for directional analysis. And I prefer the word “anticipation”, because I don’t like the word “prediction”, lol.

(*)

April 10, 2008

Interest Rate Arbitrage Trading System

Filed under: Uncategorized — by TraderMade @ 6:34 am
Tags: , ,

Arbitrage Distribution

We have just added more participants in one of our trading system. The latest one is the proprietary credits desk of Bank Mandiri, the largest bank in Indonesia. We’re still doing some tweaking on our data bridge to make a smooth data exchange between our side and theirs. Our IT people is now installing the bridge in our branch offices to smooth the hedging procedure.

The system is based on some arbitrage approach. All i can say now is it generally exploits interest rate difference between different market levels. Some mild optimization practices have been employed to maximize the profit margin per trade, which now is reaching a 10 basis point.

We are currently running this system for Bank Eksekutif, Bank Bukopin, Bank Muamalat, and several high net worth individuals. An average of 10% to 11% annual yield (nett, after tax) have been delivered to our clients. A maximum of 2%-3.5% drawdown is the level of anticipated short term negative cash flow. It’s not really a draw down because it’s actually in the form of an open trades. We’ve also been considering that forming some offshore business unit where tax regulation is better might increase the efficiency of it. Singapore or Cyprus or Hongkong, perhaps.

We’re now preparing to tap Bank BNI’46 into the system. They have a very deep pool of liquidity which will add more robustness.

Good luck to us! :)

(*)

March 18, 2008

M*nkey

Filed under: Uncategorized — by TraderMade @ 7:17 pm
Tags: , , , ,

monyet

“A domino of thoughts”, if you like.
My friend who read my previous blog entry, wrote a deeper thoughts on quals and quants’ approach in investing. He said that both quals and quants share the same key in their investing: consistency and understanding.
And that reminds me of many technical traders who use technical tools like trend lines, mathematical indicators, basic or exotic technical tools, but they never have any consistent method in their trading. They just do a ‘comprehensive analysis’ in each trading decision to be made. And to my knowledge, comprehensive doesn’t necessarily mean applying any method/strategy consistently.
An even more sarcastic coment from a fellow is that he call technical analysts as Mony*t (M*nkey). lol.

(*)

March 17, 2008

A Technician is Not Necessarily a Quant

Filed under: Uncategorized — by TraderMade @ 7:22 pm
Tags: , , , ,

An interesting discussion with one of my CFA teacher over emails after our brief discussion in the class the days before. At the end we both agree that technical analysis is a fail approach in investing.

On Mon, Mar 17, 2008 at 1:57 PM, P.M. wrote:
Untuk yang waktu itu nanya ttg John Simmons, (sorry, lupa namanya), here is the link to John Harris Simmons.
http://en.wikipedia.org/wiki/Renaissance_Technologies#_note-WSJ
http://en.wikipedia.org/wiki/James_Harris_Simons

——-start quote—–
Like many other quantitative funds, their RIE Fund had difficulty with the higher volatility environment that persisted throughout the end of ummer 2007. According to an August 10th article in Bloomberg by Katherine urton, “James Simons’s $29 billion Renaissance Institutional Equities Fund has fallen 8.7 percent so far in August when his computer models used to buy and sell stocks were overwhelmed by securities’ price swings. The two-year-old quantitative, or ‘quant,’ hedge fund now has declined 7.4 percent for the year. Simons said other hedge funds have been forced to sell positions, short-circuiting statistical models based on the relationships among securities.”
———-end quote————–

My comments:

John’s track record and methodologies have not stood the test of time.
Almost $4 billion of his total net worth of $5.5 billion only earned in 2004, 2005 and 2006. Last year, his fund suffered as market has swung wildly. Compare that to Warren Buffett’s track record for over 30 decades, which consistenly outperforms market about twice the rate of Standard and Poor’s performance.

And I dont think investment decisions should be delegated to computers.
After all, investment is very subjective and deals heavily with human psychology, which can’t be quantified.

If things get worse, and if the computers still tasked to decide which action to take, we will see another comparable for Long Term Capital
Management, a bankrupt, quantitative hedge fund that also utilised computer model, and had two Nobel prize winners as the directors: Myron Scholes and Robert C. Merton.
Check: http://en.wikipedia.org/wiki/Long_term_capital_management
Attached is the latest letter from Buffet to the shareholders.
Have a good day,

regards,
PM

On 3/17/08, Darma wrote:

Dear Pak PM, a very interesting comment of you, Pak..

we — the quants’ fans — have been very aware of what happened to LTCM.
We’re more considering LTCM as a sample of a mistake that also happen frequently to many discretionary traders. Barings’ Nick Leeson, and the more recent gigantic loss experienced by Societe Generale caused by a reckless discretionary trader is an example of what a big magnitude of a loss, a discretionary trading practice can cause.

In the case of LTCM, those Nobel prize winners were very confident that the low probability events will not happen, much like betting the market wont swing too far away from the twice of it’s standard deviation. Their biggest mistake were they didn’t expect the unexpected to happen. They were dead wrong, and it cost them big.

But a durable quantitative models isn’t about prediction, which we believe is impossible to be done consistently even using any sophisticated mathematical models.

Some durable market models are even done using some simple common sense to exploit some recurring simple market pattern like the “turtles” or any other mechanical trend followers have been doing.

This particular breed of Quants are even expecting the unexpected to happen, an event that will drive the market to make exploitable movement. Some links on this are:

http://www.turtletrader.com/trend_followers
http://www.tradingblox.com/originalturtles/
http://www.turtletrader.com/why.html

You might also want to check on these names: Richard Dennis, Paul Tudor Jones, John W Henry, Ed Seykota, etc.

Although it’s the most popular, but trend following isn’t the only kind of strategy that’s adopted by Quants. Some other kinds are: the range trading strategies, the opening range breakout strategies, etc.

There are a very diverse universe of mechanical trading strategy, ranging from the very short term to the long term strategies. Trend following is only one of them.

I’m getting more interested in knowing your comment on this, Pak. After all, i’m just a learner, a “Quant’s wannabe”, if you like. hehehe…

And I’m very glad to be knowing you, Pak.


rgrds,

Darma

On Mon, Mar 17, 2008 at 3:59 PM, PM wrote:
sorry lg sibuk

tapi brief comments: value is not the same with price. Value is what you get, price is what you pay. and this can’t be quantified :)

and the names that you mentioned never come to my radar. Dont know whether the names appear on the radar of any non-quant fans. If possible, give me the the web links to their strategies, and most important, the performance.
And we will take it from there.

I doubt that their performance will match that of the fundamentalists. If they already outperformed the market for around 20 years (means that they already experienced the bearish period of 1990s and early 2000), then the names may be worth analysing. But if their performance are only accounted for during the bullish period (where everyone is making money, and those who borrow make even bigger money), then these methods have not passed the test of time.

The problem with the trend following is that the method cant predict the turning points (bull all the time during bull period, or bearish all the time during bearish period. The method cant tell when the bear or bull market will end. This is the single biggest weakness of quant (and its siblings: the technical analysis).

Most of us will live to our 70s or 80s (hopefully……..), so I am interested to know which method will pass the test of time so I can live comfortably when I am old. So far, only 2 investment methods worth mentioning: the fundamentalist (if you know what you are doing) and the index investing (if you dont know what you are doing). Other methods, to my knowledge, have failed.

rgds,
PM

Darma to PM

ya pak. very true.
i think an important note is that a true trend following quantitative models arent trying to predict those turning points. predicting turning points is not what “following” means. and i also consider those technical analysts as a dumb group of people, using those lines and mathematical indicators to predict market, trying to outsmart the market. but as long as they still trying to predict, they’ll never be able to perform better than the market.

thanks greatly for the discussion, Pak. I’m learning a lot from it, especially on the test of time thing :):):)

rgrds,
Darma

PM to me 4:54 PM (18 minutes ago)

sama sama
btw katanya quant fans. Koq sekarang komennya lain? Jadi no fans? yg bener yang mana neh?

rgds,
PM

Darma to PM show details 5:12 PM (2 minutes ago)

hehe, saya quants fan, pak. but not a fan of technical analysis.
they two are very different things. technical analysis and quantitative trading are very different. i always mention technical analysts as “tukang ngecap”, no better then a broken clock. even a broken clock is right twice a day. lol.

using trendlines, mathematical indicators, and all those sounds like sophisticated mathematical formulas, doesn’t mean that one trading using a quantitative strategy, Pak.
the common thing I’ve seen is those technicians trade the market using discretionary strategy, they even don’t know what a positive expectancy of a strategy is.
Yes, they use those technical analysis mumbo jumbo, to give a rocket scientist like of explanation to back their BS prediction. hehehe..

rgrds,
Darma

(*)

March 3, 2008

Some Hints from the World’s Largest Hedge Fund

Filed under: Uncategorized — by TraderMade @ 12:54 pm
Tags: , ,

I have just read this wikipedia entry on Renaissance Technology (www.rentec.com), an investment management firm founded by John H. Simon. It’s the world’s largest hedge fund, handling assets worth USD 35.4 billion.

Some interesting points:

  • It uses market modelling, mechanically process market (price) data to generate trade decisions.
  • It trades easily traded financial instruments, not exploiting any complex instruments.
  • The models generally perform badly in highly volatile market.
  • They don’t use high leverage.

Some excerpt from the wikipedia entry:

Scientifically based investment strategy
For over two decades, Renaissance has been at the forefront of research in mathematics and economic analysis. Renaissance employs more than 150 scientific specialists, including mathematicians, physicists, astrophysicists and statisticians, half of whom have a PhD, who review market data to find statistical relationships that predict the price movements of commodities, currencies and stocks. These employees come from countries as diverse as Japan and Cuba [3].

Renaissance uses computer-based models to predict price changes in easily-traded financial instruments. These models are based on analyzing as much data as can be gathered, then looking for non-random movements to make predictions. Renaissance represents a validation of the quantitative trading model and trades with such high frequency that it (the Nova fund, specifically) accounts for over 10% of all the trades occurring on NASDAQ some days.

It is worth noting that Nova trades execute purely electronically on direction from a computer model. Medallion fund trades are (in large part) executed through a trading desk, whose goal is to increase the value of the positions the model directs the desk to take by timing market trends and executing in novel fashions (including intra-desk trading).

Renaissance trades at margin levels uncharacteristically low among hedge funds. This allows them to significantly reduce exposure risk, while the efficiency of their computational model allows for consistently high returns.

Like many other quantitative funds, their RIE Fund had difficulty with the higher volatility environment that persisted throughout the end of summer 2007. According to an August 10th article in Bloomberg by Katherine Burton, “James Simons’s $29 billion Renaissance Institutional Equities Fund has fallen 8.7 percent so far in August when his computer models used to buy and sell stocks were overwhelmed by securities’ price swings. The two-year-old quantitative, or ‘quant,’ hedge fund now has declined 7.4 percent for the year. Simons said other hedge funds have been forced to sell positions, short-circuiting statistical models based on the relationships among securities.”

It has just received approval from the Securities and Exchange Board of India (Sebi) to operate in the nation’s stock markets as a foreign institutional investor (FII), according to a report in the Business Standard. (CNBC).

(*)

January 11, 2008

Trend Following the Intraday FX

Filed under: Uncategorized — by TraderMade @ 9:41 am
Tags: , , , ,

I am going to base the following view with a certain definition of trend following: entering trades when the market start to go in a certain way, maintaining the open trades as long as the market still move in the same direction, and finally exit the trades when the expected trend discontinued. Purely following. The purest form of this strategy can be seen in Curtis’ white paper that was able to be downloaded freely in his website several years ago. Another form of a pure trend following system can be found in Michael Covel’s first book.

So..
In my experience, trading intraday in the FX market by purely following the trend like that, wouldn’t work as good as when we do a pure trend following system in the daily timeframe. The volatility behavior of the intraday price movement would kill the system by producing too many whipsaws. Even if we used some advanced entry technique to exploit a running trend by using multiple entry like what the turtles do, the wild volatility in intraday FX market is a beast can not be tamed with that prescription.
Using a pure trend following system to trade the intraday FX market would be like asking Mr. Bean cleaning your office desk. It would be such a terrible mess. lol.

Intraday market has its own characteristic, thus need to be handled with different approach.
Would elaborate more into that in another posts next time.

(*)

January 9, 2008

ATR-derived Stop Loss Level

Filed under: Uncategorized — by TraderMade @ 10:22 am
Tags: , , , ,

.

After reading my online buddy’s blog entry about MAE/MFE analysis, i thought it’d be worth posting this note about how we calculate stop loss level based on ATR (Average True Range).

Using a fix n pips distance for SL and/or TP is the simplest way a trader would do. But sometimes, the market would perform differently from time to time. Sometimes the market moves wildly with longer candle body, sometimes it moves very calmly with a very low volatility. A fix n pips SL/TP wouldn’t fit into the always changing market’s volatility behavior.

Moreover, we surely want to test a strategy in several different currency pairs. And because each pair would have different volatility behavior, using an ATR-derived SL/TP level would provide us with a system that can literally self-fit into any kind of volatility behavior.

These are some sample lines from our strategy. It calculates ATR first:

atr_SL = iATR(Symbol(), 0, prm_normalization_atr_period, 1) * prm_SL2atr_ratio;

and then this atr_SL value is used to calculate SL distances as follows:

stoploss = NormalizeDouble(Ask + atr_SL, Digits);

so the value of the SL distance would be based on ATR, not a fix n pips.

(*)

January 8, 2008

Anticipating Price Feed Difference between Brokers

Filed under: Uncategorized — by TraderMade @ 2:34 pm
Tags: , , , ,

We have noticed a significant difference in the performance of our systems in several brokers. We have been wondering what may be the cause of it. When we examined the situation more closely, we found that an exact same algorithm has produced fewer trades in a more-filtered price feed environment.

How can this happen? Well, generally, the OTC FX market is a bilateral market, between the market-maker type broker with the traders. Thus the price quoted by each brokers would not be the exact same price. Some brokers apply some filtering algorithm for their feed, creating a smoother chart that may effect the performance of certain trading systems.

So, we have just added some adjustment into one of our algorithm, by adding/subtracting a certain pips off the trigger level as can be seen in the picture below:

(*)

December 19, 2007

Time Dimension of the Market

Filed under: Uncategorized — by TraderMade @ 1:49 pm
Tags: , , , , ,

.

Several short term trading systems we have been developing for the FX market relays heavily on the distinctively unique and consistent nature of volatility behavior between market’s sessions.

We even haven’t found any ways to substitute our time-constraint filters with something purely derived from price-only data. It’s like we actually have a different kind of market for each session that is divided by a rigid time mark, not by some hints those can be found in price action.

This has been bringing some serious thought for us, because we thought that we have to stick to our philosophy to relay our systems purely on price action. But then we realize that the market has the dimension of time along with the dimension of price.

And to our surprise, I found this interesting observation by Mr. Steenbarger, Ph.D. In one of his post today in his blog, he stated a similar conclusion as ours:

What we can see is that these seem to be different markets. Indeed, the daily correlations among the three range from .12 to .18, suggesting that what the market does during one time period is only very weakly related to what it will do in the next one.

Although his research is about the stock market, particularly the S&P, this has put us into believing that time dimension is very truly exists, as we also been observing the same phenomenon in the FX market.

This confirmation from Mr. Steenbarger’s research has put a new item to our research agenda, that we will need to check market’s average range for each session. This would probably be useful for developing a better opening range breakout model, by having a better measurement of the opening range width. And for our range trading model, we might get a better basis for the entry and exit strategy. The keyword would be determining the historical and average range of sessions.

(*)

November 14, 2007

Postmo Darvas Method (?)

Filed under: Uncategorized — by TraderMade @ 11:13 pm
Tags: , , ,

.

Still inspired by the same article as in my previous post, the author suggested a form of a “Modern Day Darvas Method”,

Here are a few ways a modern-day Darvas might prevent micromanaging trades:
* Use Your Platform Tools. Set it and forget it. Whether you have sophisticated tools like bracket orders or just the basics, use the tools your broker has provided to help you structure your trades in such a way that they have the ability to play out without needing your constant supervision. This will help you to trade without emotion and allow you to fully implement your strategy for the trade from start to finish.
* Hide The Number. I’ve written about hiding the number before, and I think it can be quite beneficial, whether it’s your account balance or even just your P&L on trades. A trader friend of mine claims that hiding his P&L has made the biggest positive impact in his trading – more than any tools or indicators. It’s an interesting concept, particularly if you find yourself overreacting to every little pullback or advance. By giving his broker trade instructions ahead of time, Darvas didn’t watch his P&L or spook himself out of trades.
* Turn off the news flow. Darvas found that his absence from the Street allowed him to stay in trades longer so that he could let his original game plan play out. That means he was not surrounding himself with rumors or opinions or the news flow. He didn’t make bets based on gut feel – he stuck to the charts, trading from the price action when opportunies arose. If for you that means turning off CNBC, avoiding comparing opinions with your trading buddies by staying off of instant messenger or Skype, then so be it. Whatever helps you make your money is what you need to focus on. Remember, Silence is Golden!

Yes, they are indeed good tips. But if the method above are called “Modern”, what then he’ll call this method below:

  • Define the strategy into the fullest details,
  • Write them into a working algorithm,
  • Backtest,
  • Evaluate,
  • Modify & improve,
  • Re-test,
  • Re-evaluate,
  • Fine tune,
  • Another re-test,
  • Validation test,
  • Let the strategy be automated through any growing number of supporting brokers.

Will these sequence of strategy automation produce a new method called as “Post-Modern Darvas” then? LOL.

(*)
Postmodern architecture

November 6, 2007

Chart Analysis Excitement

Filed under: Uncategorized — by TraderMade @ 5:27 am
Tags: ,

From my observation so far, traders who really love to show the screen shot of their wonderful charts with all those exotic indicators and lines and curves and all, are traders who would not feel comfortable for others to see their actual trading result. Some kind of a negative correlation there between the two things. lol.

Successful traders don’t spend times drawing exotic pictures off their chart. They have a system/method, and always is a simple method. Often they even get very bored to execute their own system by themselves, then they start to delegate the execution of their simple method. At that stage, paying programmers to automate it would be much cheaper than paying traders to execute the (boring) system.

Traders may have their own method, including all the analysis mumbo-jumbo. Yeah, they said that whatever works should be okay. Each traders have their own way. But from my experience, that is not the truth.

Look at the chart posted above. I can be 99% sure that the trader doesn’t want his trading result to be seen by others. But he is so proud of his painting. Yeah, he better buy some paints, brushes, and canvasses and maybe start to set up an art gallery/museum. lol. I mean, look at the chart. It’s so colorfull, so live (lol), even when I zoomed in the chart I felt a litle bit difficult to find out the candles. lol. I was so amazed. lol.

So, another hint here: if a trader still feel excited doing his daily analysis on his charts, using his amazing imagination on each of his analysis, or discussing the fundamental stuffs with his fellow traders in trading forums/chat rooms, perhaps he should take a look at his trading statement to bring him back down to earth and start to realize that it’s not the way to be profitable.

(*)

October 25, 2007

Why you need a complete trading system

Filed under: Uncategorized — by TraderMade @ 11:31 pm
Tags:

.

Jokes on market analysts. It goes to both the technical and fundamental type of analysts:

A market analyst is an expert who will know tomorrow why the things he predicted yesterday didn’t happen today!
===
Q: Why did God create market analysts ?
A: In order to make weather forecasters look good.
===
The best time to buy anything is last year.
===
The market is weird. Every time one guy sells, another one buys, and they both think they’re smart.
===

I prefer a complete trading system rather than a complete analysis before each of my trades, especially that come from market analysts/gurus/experts.

(*)

October 15, 2007

The idea on grid

Filed under: Uncategorized — by TraderMade @ 10:31 am
Tags: ,

.

This week, we’ll explore several ideas on system development, particularly on the multiple entry and exit ideas that might be a powerful tool to be implemented into the currently used trading system. Not really a new one, actually. The technique has been implemented in several development project by several other fellow quantitative traders, but has never been used in a proper way, imho. Some robots with ‘grid’ approach are usually ignoring the most important thing in system development, which is the risk-management and risk-limiting function. I have seen many ‘grid’ system that perform well in all trades but then suddenly wipe out the account in a single occasion. It’s like having the lion in a well-built cage but the lock is unlocked.

And now, Aditya K, a fellow quants, he is a very young boy (but not as young as my fellow in the picture, lol), has shown me his trading performance using this technique in a counter-trend approach, which generally is ‘gridding’ a counter-trend trades off (a suspected to be) an exhaustion condition of a move on a short to mid term time frame. But it seems like he has never been exploring the use of some volatility-measurement tools and other essential filtering aspects which in my experience are very significant in improving the performance of systems.

Meanwhile, the currently used system, called the Ri-VER (ver vy.6c), has been performing well this month. Its been trading well on the accounts and made a 3% gain on them. I hope the market will provide us with another great months ahead. Yeah, good luck to us. :-)

(*)

September 28, 2007

Algorithmic trading. There’s no other way better.

Filed under: Uncategorized — by TraderMade @ 7:33 am
Tags: ,

.

Kebetulan ada mas deddy yg tanya agak panjang, jd banyak yg mesti dijawab. dan kayaknya tak posting sebagai blog aja deh, instead of hanya private message, jd supaya bisa dibaca yg lainnya juga oleh temen2 di indotraders.

========================================

good morning, mas deddy.
saya senang ada yg ngajakin diskusi beginian.
soal bikin EA, saya nggak coding sendiri kok mas. saya kerjakan bersama seorang programmer, teman kuliah satu kampus dulu di bandung, dia lulusan informatik, sedangkan saya farmasi. saya punya ide, dia coding. lalu utk modifikasi2, nambah2 filter, ngerubah2 perhitungan sedikit-sedikit, saya bisa lakukan sendiri. tapi setelah terjadi perubahan beberapa versi, saya kembalikan lagi ke dia utk diverifikasi apakah codingnya masih bener, gitu mas.. jadi kalo soal belajar coding yg baik, saya sendiri belum lakukan itu, krn saya sendiri belum ketemu metoda yg tepat utk belajar itu.

soal penggunaan mechanical system dalam bentuk EA di MT4, dibandingkan dgn trading manual dgn menggunakan segala kemampuan berfikir manusia beserta unsur emosi dan psikologinya, saya sendiri punya pegangan dalil2 sebagai berikut, mas:

dalil #1:
utk trading, kita hrus pakai metoda atau strategi yg kita yakini bahwa kalau dijalankan, akan profit. metoda atau strategi ini bisa macem2, bisa mekanis ataupun murni feeling, atau tebak2an.. tapi prinsipnya adalah kita akan menjalankan aktifitas trading menggunakan suatu cara -apapun itu bentuknya- yang kita yakini memberi peluang besar utk profit. Lha kalo kita gak punya strategi yg kita rasa akan profit, kita tentu gak akan bisa memulai trading kan ya?

dalil #2:
untuk bisa memilih strategi yang kita yakini akan profit, tentu kita harus melihat dulu, kira-kira apa bener suatu ide tentang strategi itu bakalan profit. Misalnya kalau ada ide strategi trading berupa: “buy kalau market retrace di fibo, lalu exit saat overbought”, tentu akan wajar sekali kalau kita cek ke historical, lihat2 secara visual di historical chart, apakah kalau kita lakukan strategi itu memang benar ada peluang akan profit. Lalu kalau ternyata kita harus perhatikan kondisi tertentu seperti big news time dll, itu juga kita lihat di historical. Hal ini tentu merupakan hal yg biasa dalam mencari-cari ide trading, kan? Setiap trader yg berusaha mencari strategi pilihannya, pasti melakukan hal ini, yaitu pemeriksaan historical, atau istilahnya backtest. backtest ini bisa manual/visual, atau mekanis utk lebih akurat.

dalil #3:
memeriksa suatu ide strategi tertentu dengan teknik pemeriksaan visual secara manual, memiliki banyak sekali kelemahan dan keterbatasan.
Kelemahan2 itu antara lain:

- seringkali kita memeriksa dengan bias/keberpihakan. seorang trader yg terlatih mungkin bisa mengurangi faktor ini. keberpihakan ini maksudnya begini mas: kita lakukan pengamatan visual untuk mencari pembenaran dari ide strategi kita. Jadi tiap kali kita lihat kejadian kalau kriteria entry-exit yg kita ingin periksa ternyata menghasilkan profit, kita hanya perhatikan itu saja. Kita lupa perhatikan bahwa ada banyak kondisi yang ternya berakibat kerugian, dan itu kita abaikan krn hasilnya tidak seperti yg kita inginkan. We see what we want to see.

- kelemahan parah lainnya adalah adanya faktor “oversight”, yaitu misalnya kita buy kalau marketnya breakout ke atas suatu trend line, kita pikir kita bisa jalankan strategi itu, padahal trend line nya terbentuk dari titik-titik fraktal sebelum dan sesudah kejadian itu. ya ini contoh ekstrim, tapi kesalahan “oversight” ini sering kali dialami oleh orang2 yg baru mulai mencoba mengembangkan strategi trading.

- kelemahan lain: bila di historical chart kita melihat ada move naik yg ‘bagus’, dan kita ingin bisa menangkap gerakan itu dan mengkapitalisasi profit dari gerakan itu, lalu kita cari2 kondisi yg mendahului gerakan ini, lalu kita anggap kita menemukan suatu strategi. begitu terus kita lakukan tiap kali kita lihat ada gerakan ‘bagus’ di market yg ingin bisa kita tangkap, kita coba-coba cari kondisi di market yg mendahuluinya. padahal kondisi itu berbeda2, dan itu hanya kelihatan kalau kita sudah lihat gerakan ‘bagus’ yg kita inginkan.

SEMUA kelemahan dan keterbatasan ini akan hilang kalau kita lakukan pemeriksaan menggunakan teknik non-manual, misalnya pakai program statistik, di excel, atau pakai backtesternya metatrader.

dalil #4
Kemudian kita sudah menemukan suatu strategi yg ternyata memang profit, misalnya tiap kali market bikin triangle di jam tertentu, lalu break ke atas saat MACD di long term mengarah naik — dan ini semua sudah ditest dengan cara yg akurat dan menghasilkan kesimpulan-kesimpulan sbb:
– “ooohhh… strategi ini kalau dijalankan memang profit”
– “rata2 dalam sebulan, kena SL hanya 5 kali, sedangkan selebihnya rata2 ada 20 yang kena TP”
Kalau kita sudah punya strategi yg setelah diperiksa memberi kesimpulan2 bagus seperti ini, tentu kita ingin menjalankannya secara disiplin kan?
Nah di sini lah faktor psikologi seringkali sangat mengganggu. Seringkali strategi yg seharusnya profit, tapi malah menjadi merugikan karena tradernya terganggu dari segi psikologi. Macem2 masalah psikologi yg bisa terjadi mas. Misalnya utk contoh trader yg pakai strategi triangle breakoutnya tadi:
– Sang trader enggan menuruti rules nya sendiri krn tiba2 ngerasa “kok kayaknya ngeri ya, buy di sini, walaupun triangle udah break ke atas, tapi kan market udah naik tinggi beberapa hari belakangan ini”..
– Sang trader tiba-tiba mengambil posisi trade padahal tidak ada kondisi yg sesuai dengan rules nya, krn tiba2 berfikir “ini mestinya bakalan naik nih, walaupun gak ada triangle, tapi ini kan marketnya udah bolak-balik di bawah resistance ini, jadi sekarang udah break, kita mesti buruan masuk nih supaya nggak ketinggalan”.
– dan masalah ketidak-disiplinan lainnya, terutama masalah exit, menggeser SL, position sizing yg seenak perutnya, dsb dsb dsb..

nah kalau kita udah test/periksa ide strategi kita dengan menggunakan teknik yg akurat, misalnya menggunakan algoritma di metatrader, mestinya strategi itulah yg akan kita jalankan dengan disiplin kan.. disiplin artinya disiplin, nggak diubah2. gak dibikin pengecualian2 yg bebeda-beda tiap saat. kalau memang ada kondisi yg akan dikecualikan, misalnya gak akan trading kalau market New York libur, ya itu mesti diperiksa juga di historical kalau dilakukan begitu hasilnya gimana, apa memang lebih baik atau tidak. Kalau memang lbih baik, ya itu lah yg hrus dijalankan dengan disiplin, dst.

Trading dengan disiplin ini semua akan jauh lebih mudah bila kita jalankan strategi kita menggunakan suatu computer-programmed algorithm, misalnya pakai tools alert di MT4, atau bahkan pakai tools automatic trading di MT4 menggunakan EA itu.

dalil #5
Kalau kita tidak percaya dengan suatu strategi akan profit dijalankan krn kita pikir market akan berubah sehingga misalnya di contoh di atas, strategi triangle dan MACD itu kayaknya cuman profit di dulu2 aja, tapi besok2 belum tentu, ya lalu kita tidak punya alasan yg kuat utk menjalankannya, bukan? Di sini, menurut saya, kembali ke faktor psikologi dan rasional.

Kalau strategi tarik2 garis dgn supp/res atau fibo itu kita pikir gak akan profit lagi di masa depan, ya berarti kita gak akan pakai strategi itu juga. Kalau kita gak percaya suatu strategi punya harapan akan profit, lalu dengan apa kita akan trading dong. Sampai saat ini, sepenjang pengetahuan saya, utk menjalankan suatu strategi utk ditradingkan live, setidaknya kita hrus punya harapan bhwa strategi itu akan profit kalau dijalankan. Lebih baik lagi tidak hanya harapan, tapi juga keyakinan, yg kita dapatkan dari hasil pemeriksaan yg menunjukkan hasil yg baik, dan pemeriksaannya akurat.
Dan kalau kita yakin strategi trsebut mestinya profit, tentu kita ingin lakukan dengan disiplin se-disiplin-disiplinnya, kan.

Semua dalil2 dalam pengembangan dan eksekusi strategi trading di atas tadi menunjukkan argumentasi yg sangat kuat utk menggunakan automated system. Gak ada pilihan lain, menurut saya.
Kalaupun artificial intelligence kita percaya bisa menjadi tool yg lbih baik, dan kita tidak percaya thd strategi yg hanya linear matematis biasa, ya lakukan itu. Do whatever you want to do.
Kalau setelah kita periksa suatu strategi secara akurat menggunakan EA di MT4, lalu forward testing menunjukkan hasil bagus juga, tapi begitu mau live kita malah tidak percaya bahwa market akan berlaku begitu lagi, ya kita punya hak penuh utk tidak menggunakan strategi itu utk live. Don’t do anything we don’t wanna do.
Tapi kalau dalam keadaan itu lalu kita drop out strategi itu hanya krn tidak berani, ya lalu mau trading pakai strategi apa? There’s no any other way better.

(*)

September 25, 2007

Where’s my price???!!

Filed under: Uncategorized — by TraderMade @ 12:36 pm
Tags: , ,

There was the thing that will happen when using a broker that close earlier than others. There was no price on IBFX‘s feed where my TP at USDJPY 115.52 should be hit.

Look at these pictures below.
The first and second picture are the screen shot of the 15min charts of USDJPY on FXLQ and MIG. There was a tall candle there from the both brokers.

But with IBFX, the candle at 20:45 where the price should hit my TP wasnt there because they close earlier on friday. So there was no tall candle there.

This is FXLQ’s price:

This is MIG’s price:

And look at what IBFX has done to the price:


(*)

September 10, 2007

Ri-VER v.y.6c

Filed under: Uncategorized — by TraderMade @ 8:38 pm
Tags: ,

Need to add more touch on the money-management part.
Now is still using a proportional increment according to the equity.
Cllick on the thumbnail to view a screen shot of the golden goose. =P

Thanks to Mojo who in the previous chat session, has reminded me of some basic concept in system development. The algorithm is so much better now.

Conclusion from a valuable chat session today with him:
Mojo: 1) use faster momentum (osma)
Mojo: 2) segmental optimization
Mojo: 3) don’t depend on timefilter

August 29, 2007

I feel sorry for them, but let them be

Filed under: Uncategorized — by TraderMade @ 2:41 am
Tags:

An interesting article about algorithmic trading I’ve found here.
An excerpts from it:

…………..Recent research from the Tabb Group indicated that in the USA, perhaps 60% of Investment Managers use or experiment with algorithmic trading……………

The majority of many friends and prospective clients i’ve talked with still believe that algorithmic trading is just a new, un-tested, un-reliable, too-risky approach to trade the market. The majority of retail investors still hold a strong believe that trading must be done using technical analysis that needs imagination to draw the trendline, need to have a sensitive-trained eye to spot the support/resistence levels, must use feelings in some extent, that the skill of an experienced trader is too complex to be programmed into a computer program, etc etc etc.. All these have resulted in a popular believe that algorithmic trading is an impossible approach for profitable trading, and it’s only for daydreamers.

Well, I feel sorry for them, feel sad for them, but since it is still –and will still be– the nature of the statistic is 95% of the players are losers, so let them be. They want to do this discretionary, well equipped with all those technical analysis mumbo-jumbo, and with all those fundamental analysis bullshit, so let they get what they deserve to get.

Recent research from the Tabb Group indicated that in the USA, perhaps 60% of Investment Managers use or experiment with algorithmic trading. I believe that they have a better chance to be belong to the winners group than all the rest of the crowd that I feel sorry about.

(*)

August 28, 2007

Diskusi ttg trading dgn Mr. Mojo

Filed under: Uncategorized — by TraderMade @ 11:16 am
Tags: , ,

Pagi ini berkesempatan chat-chat curhat dgn seorang teman yang selalu menginspirasi. A Guru, if you like to call it. Dia salah satu founder dan moderator di Metatrader Expert and Advisor Groups. He has been always inspiring. Terutama di bidang Algorithmic Trading. Kalau di bidang-bidang lain, ya beliau ini pun cukup menghibur, hahahahaha

Chat ini saya posting di sini, karena saya pikir, obrolan ini mungkin bermanfaat dan menginspirasi bagi teman-teman sesama pengembang trading algorithm, terutama untuk teman-teman Indotraders.

Selamat menyimak.

Darma: jo, apakabar. gw pengen tanya. kalau development market modelling seperti yg kita lakukan dgn bikin trading algorithm begini, bidang studi apa yg melakukannya di universities ya? thx
Mojo: financial engineering
Mojo: ato quantitative finance
Darma: ic. kalau seperti gw yg ga ada latar belakang economic/finance, hanya background science aja, mungkin gak bisa ambil postgrad dgn scholarship, terutama di germany.
Mojo: asal lu bisa lewat tes matematikanya biasanya gak masalah
Darma: ok thanks jo for the leads.
Mojo: k
Darma: apakabar jo. masih trading?
Mojo: baek … masih
Darma: my robot, jo. intraday yg main reversal di jam2 ranging, sampai sekarang equitynya bolak-balik mulu nih
Darma: http://img207.imageshack.us/img207/2660/detailedstatementeb9.gif
Mojo: bolak balik naek? keren dong
Darma: di timeframe ini, spread dan cost menjadi terlalu significant
Mojo: masih yg terakhir?
Darma: kagak naek, disitu2 aje
Darma: udah 3 bulan live tuh jo.
Mojo: lha yang laen emangnya gak bisa cover?
Mojo: itu trade selama 3 bulan total?
Darma: yup. di interbankfx. mungkin klo dipindah ke broker yg spread di cable lebih tipis, udah positif tuh. initial test, $300 doang tuh jo
Mojo: ya berarti efisien dong ea-nya
Darma: huahahaaa. efisien apaan
Mojo: sayangnya pengennya yg gak efisien ya
Mojo: iya efisien, kan jadi delta neutral
Darma: kalo jadi IB, makan rebate, enak tuh. transaksinya banyak
Mojo: kalo market efisien ya jadinya gitu
Darma: ic ic
Darma: ini ngandelin premis bahwa pada jam2 sepi antara closing US dan sebelum open EU, market ranging
Darma: jadi sell di top bollinger, buy di bottom bollinger.
Mojo: yg presentasi terakhir kan?
Darma: gitu doangan
Darma: iya jo, yg itu
Mojo: ya tinggal difilter ama direction aja kan?
Darma: cuman udah dimodif exit nya, poisition sizingnya, dll
Mojo: buy bottom on bull, sell top on bear
Mojo: emang ada cara laen buat trading?
Darma: haha, iya sih jo. gw udah coba filter gitu. tapi tampaknya di timeframe itu, marketnya independen, gak dipengaruhi oleh bull/bear timeframe yg lbih besar.
Darma: malah jadi ada profitable trades yg difilter
Mojo: ooo
Darma: kalo strategi ini dimainkan di tmpat yg lbih tipis cost/trade nya, kyknya udah profit
Mojo: bikin bridge ke ECN ajah
Darma: mstinya sih gitu ya. cuman gw sebel banget, kok profitable trading itu ujung2nya sangat dipengaruhi cost per trade ya..
Darma: main long term juga, gw hitung2, profitnya habis dimakan interest.
Darma: mungkin approach strategi gw terlalu konvensional
Mojo: wah konvensional itu gimana ya
Mojo: toh market bukan brownian kan
Darma: jadi udah diantisipasi oleh market maker. di tf kecil, diantisipasi dgn spread. di tf besar, diantisipasi oleh interest.
Mojo: jelas2 bukan 50-50
Mojo: kalo gak efisien ya bisa dieksploit
Mojo: interest mah kecil
Mojo: (kecuali maen carry )
Darma: hmm.. maksud gw, approach gw terlalu seperti ‘trading yg seharusnya’.. gak kayak yg akal2in market itu lho, seperti yg grid, martingale, dll..
Mojo: kayaknya sih gak perlu diakalin ya
Darma: ic ic
Darma: iya, gw jg pikir gitu.. emang mestinya wajar2 aja sih.
Darma: yg risk nya masuk akal
Darma: dan strategi yg wajar gini
Darma: cuman akhirnya di kasus gw ini, cost receh2 menjadi sangat berpengaruh
Darma: akhirnya, sampe sekarang pun, belum kelihatan hasil nih.
Mojo: lha bukannya backtest/forward dulu katanya spektakulieur?
Darma: but let’s see, bulan depan mungkin market menjadi lebih ‘teratur’, yaitu ranging di jam2 itu
Mojo: tapi minimal kan gak kemana2
Darma: yup, backtest oke banget.
Mojo: terus keliatan gak ada ‘musim’ begini di backtest?
Mojo: kudunya kan kasih clue yg jelas
Darma: tapi memang selalu ada masa2 3-4 bulan di tiap tahun yg marketnya lari pas jam2 itu.. tiap tahun ada aja gitu.. mudah2an yg tahun ini, yg equity gw gak ke mana2 ini, adalah kejadian yg 3-4 bulan seperti di tahun2 sebelumnya
Darma: yup, ada musim begini
Mojo: ya asik dong
Mojo: bentar lagi gua ditraktir makan2 di bali
Mojo: mozaic ubud ya
Darma: huahahahah elu bisa aja
Darma: kekhawatiran gw, musimnya kok agak panjang ya kali ini
Mojo: lha, kan abis gelap terbitlah terang
Darma: ya, jadinya tetep, menunggu dulu deh
Mojo: gak papa, gua juga bisa nunggu makan2 … tenaaang
Mojo:
Darma: di tahun2 backtest, misalnya profit 2000pips per tahun, itu dari 600 transaksi.. jadi kalo per transaksi ada cost yg lebih mahal 1-2 pips aja di live trading, maka profil ekuity nya jadi beda.
Mojo: itu MBT kan buka API ma
Mojo: kan ente ada programmer di bali
Mojo: cobain ajah kesono eksekusi signalnya
Mojo: keliatan beda gak
Darma: yup
Darma: jadi kesimpulan gw sementara ini: menunggu. kalo masih gak bagus juga, cari eksekusi yg lebih murah per trade. gitu perkembangan gw jo. lu gimana
Mojo: ah gua mah gak kemana2
Mojo: masih stuck
Darma: lha kalo masih stuck di tanjakan equity sih asik jo
Mojo: amin
Darma: tuh kan
Mojo: pengennya sih stuck kenapa gak bisa >100% seminggu
Mojo: pengennyaaaaa
Darma: anjritt
Darma: jo, thanks for the chat.
Darma: inspiring, as always.
Mojo: sip
Mojo: ah inspiring apanye
Mojo: perspiring iya, panas
Darma: yeyyy

Ya, begitulah. Mojo, selalu inspiring.
Semoga bermanfaat.
Good luck to us!

(*)

The Toni Theme. Create a free website or blog at WordPress.com.

Follow

Get every new post delivered to your Inbox.