A mechanical trading strategy (continued...)

To be honest, I have not yet built a mechanical strategy that I think is robust enough. Like the strategy that I built in my last post. It looked like that it had some trading edge but it cannot hold up when put into real trading a few months. A robust strategy should be able to hold up its trading edge for an extensive period of trades historically and in future trades also.

Knowing what a robust strategy is, we can now start to design our strategy. There are a lot of ways to trade the market: fundamental trading, momentum trading, trend trading etc. You can pick your own favorite method and formulate it into a quantitative strategy. If you are an experience trader, it should not be difficult.

The difficult part for most of the people is to test their strategy against historical data if they do not have tools like MetaStock, TradeStation. Some people would suggest to visually back test your strategy tick by tick using historical charts. But I think it is too time consuming and you won’t have the patience to do so. If are really interested in the context of mechanical trading, I recommend you to learn how to do it programmatically. You don’t need to know advanced programming techniques in order to do so. The aim of these software tools is to help you simplify the process of building your own strategy. All you need is to start to look into it seriously (google it now).

Once you know how to test your strategy programmatically, there are more things you need to take care of like money management, strategy optimization etc. before you can come up with a robust strategy.

令人驚喜的同仁堂 (8069.hk)

自從上期推介了內需股, 魏橋紡織(2698.hk)和同仁堂(8069.hk)之後. 首先魏橋紡織繼續大幅上升. 從$14.5 Monkey繼續沽貨. 在$16左右終於沽清,它仍然 但一直升到超過$18. 令次真是望價興嘆. 還好繼魏橋紡織之後, 同仁堂又大發神勇. 雖然Monkey之前也看好這隻股票, 但萬萬想不到升勢如此強勁.一個星期之內, 從$15直上$21.7. 令人側目. 奇怪的是除了有間投資基金Hamon增持之外, 好像不覺得有什麼好特別的好消息刺激.


最近大陸的製藥股都造好. 好像之前提過業績奇差的中國製藥(1093.hk). 新華製藥(0719.HK), 廣州藥業(087.4HK),都大幅上升. 但根據新聞內地近日再調整260種西藥零售價格，其中70%品種價格下調，平均降幅19%，最大降幅62%，其餘30%廉價藥品則維持原價，而對市場供應緊缺或斷貨18種藥品，則適當提高售價。各大製藥股的大幅上升, 令人感到有點莫名奇妙. 不過手中的持股上升了心情還是十分舒暢.

最 近Monkey檢討自己投資的策略, 發現自己其實都有普遍散戶投資所犯的錯誤. 就是當某隻股票大幅下跌時, 就會覺得股價超值, 於是買入. 但往往低處未算低. 在股價大幅上升時, 又擔心股價會回吐, 於是又過早賣出. 損失了一大段的利潤. 好像今次魏橋紡織就是一個過早賣出的例子. 這種投資心理學最基本的錯誤實在應該盡量避免.

這令我想起投資大師Philip Fisher 的 Common Stocks and Uncommon Profits 所提出的原則, 就是. 認真用書中所提的十五個原則去挑選股票. 選好之後, 除非股價高得瘋癲, 或者是公司本身出現結構性的變化, 否則永遠不用考慮沽出股票. 回想起來, 如果一直有紀律地依據此原則去操作. 利潤比現在高得多. Philip Fisher的投資建議, 絶對是金石良言, 非市面上的股評家可以比較.

A Mechanical trading strategy

I am going to share more about mechnical trading and trading strategy here. Okay, let us start with some basic stuff...

When I started to trade HSI future. I found that it requires much of my attentions. It can drop or up like hell in a few minutes. If you are in a wrong position, you can lost your month's salary in a minute. I was a busy guy (actually a lazy guy). I cannot afford to have my only two eyes on this thing every minutes. I will definitely get fired before I can make any fortune from the future market. I finally found this tool call Tradestation and the concept all mechincal trading that safes me from getting fired.

So what is a mechanical trading strategy. A mechanical trading strategy is a quantified decision mechanism that tells the trader or the "trading robot" when to enter or exit a position, when to stop and determines the position size. the key point here is "quantified'; that means there is no trend reading by drawing trend lines (although this can also be quantified), no emotional guessing, no personal guts feeling.

More importantly, if you don't have a "trading robot" (automatic trading program) to execute your trading strategy. You have to straightly follow the signal of your trading strategy when you trade or else your trading strategy means nothing.

A trading strategy can be as simple as:

Entry: when 20 days moving average cross over 50 days moving average, buy.
Exit: when 20 days moving average cross under 50 days moving average, sell.
Stop: 10% cut loss

But his may not a robost strategy. How to buil a robust mechnical strategy? to be continued...

傳媒業的新景象

左輪右輪, 終於輪到一向被視為夕陽工業的傳媒股. 傳媒大亨梅鐸屬下的新聞集團宣佈意向收購Dow道瓊斯. 接著Thomson Finance 又想收購路透社. 一下子, 傳媒股變成寵兒. 連帶香港的某些傳媒股也大幅上升. 比較明顯的是SCMP(0583.hk),從$2.85左右, 幾日內升上$3.3. 足足升了16%. 對於一向沉悶的傳媒股來講, 是令人欣喜的升幅. 細心觀察, 會發現, 香港傳媒股的上升只限於SCMP和經濟日報. 正如Monkey之前所猜想 (http://monkeyforecast.blogspot.com/2007/03/scmp583hk.html), SCMP是海外資金收購的理想對象. 香港是亞洲的商業及金融中心, 而SCMP(南華早報)在香港和附近地區對中產有很大的影響力. 再加上SCMP幾乎是香港獨市的英文報紙. 所以如果新聞集團或其他資金想進軍香港或東亞的話. SCMP決對是不二之選.


為了保持競爭力, 全球性的合併想信,將會是傳媒業的大趨勢. 每天在地鐵站免費派發的都市日報就是一個很成功的例子. 全球的採訪, 然後每個地區共享資源. 既可降低成本, 又可對新聞擁有控制能力, 不用依賴路透社或其他大新聞機構.同時 對於吸引國際的廣告客戶也有好處.

今天SCMP有大幅的回吐, 相信主要原因是因為新聞集團的內慕交易的影響, 下跌了5%. 但Monkey認為只有暫時的 影響. 最令Monkey覺得有興趣的是, 在昨天的交易中, 見到大部份的大手買盤為澳洲投資銀行麥格理. 由於新聞集團和麥格理同以澳洲為基地. 所以會否收購SCMP, 都不是完全没有可能.

上了電的魏橋紡織 (2698.hk)

上年魏橋紡織的表現令人沮喪, 大家都一窩蜂大升時, 它郤逆市下跌, 最低跌穿了$9. Monkey一向此股有頗深入的研究, 所以一直都在增持. 自從上兩個月以來, 魏橋紡織的股票開始上升, 而且好像食了興奮劑, 升勢越升越急. 由於升勢超乎尋常, 所以在$13.8左右放了一批. 誰知, 它的股價略為回一回後, 今天又大幅上升.直上到今天的$15.3. 之前太早買出實在可惜. 還好還有不少的存貨. 未至於望價興歎.
其實在大家炒完金融股之後, 相信下一浪應該是內需股. 大陸有十幾億的農村人口. 經濟好了, 收入增加了, 自然要改善生活. 人生的必需品, 衣食住行中, 食和住對於他們來講完全不是問題. 而他們想改善的是穿比較好的衣著, 和離開村莊到外面走走見識一下. 所以紡織股和旅遊行業可能得益. 而魏橋紡織以紡紗和製造坏布為主, 是所有成衣業的原材料, 自然可以享有行業的增長.

農村人口另一個迫切需求就是醫療. 醫療製藥股將會得益. 買藥最怕買假藥, 所以品牌最重要. 其中Monkey比較看好同仁堂(8069.hk). 同仁堂科技的產品有其獨特之處. 而品牌也有相當號召力. Monkey一直在想, 以同仁堂產品的質素, 如果可以好像2036靈芝苞子那樣推廣. 業績一定會大幅增長. 只是管理層在市場推廣方面, 還是比較保守. 只要找到適當的人才, 有可能好像房署商場交到領匯手上, 完全可以脫胎換骨.

Mechanical trading of Hang Seng Index Future (HSIF) 程式買賣指數期貨

Last month, I finally figured out how to connection my TradeStation(TS) program to the real-time HSI future data and my brokerage to place order directly from the signal generated by TS. Now I can easily change my trading strategy with the flexibility given by TS.

Monkey is doing a lot of prediction in his blog posts. I do not believe in prediction. I tend to be a trend follower more then predictor when I develop my trading strategy. From now on, I will write more on this blog to share my finding of the continues experiment of my ultimate HSIF trading strategy.

My first trading strategy:

The above TS generated strategy performance report looks promising right? Net earning of 1 future contract is around 4000 pts (worth HKD200K) in 7 months. I think I can retire soon…but….share with you my findings next time.