1. June 2012: Buy low sell high (most primitive)
Buy when I "feel" it is too low, and sell when I "feel" price is too high. So let's say EURUSD during the past week has never get to the high level of 1.24, now it reaches 1.24, I will short it. I used Fibonacci and set stop loss take profit level. Unleash, I read something on The Economist about a major economics/politics news that drive the EURUSD significantly, I kept following that strategy. I ended up having like 30% profit. Not bad for a starter. I felt that I can do this.
2. July: Chartist
As a mathematic student at heart, that primitive strategy didn't feel really right to me. Me went on and read moving average, then buy EURUSD when MA(12) cut MA(24) from below, and vice versa. I was nodding to myself: I'm quite smart applying maths to trading. But then at the same times I knew that other people must have known this. Probably I need more sophisticated indicators.
3. October: Mean-variance
When I first learned the Markowitz model, I was amazed. The first beauty I found in trading strategies. I was like this makes a lot of sense. It could works. It is so nice that the guy linked portfolio management with mathematics and statistics. I applied it to FTSE100, and got positive returns after a couple days. I thought that was because of the sound model. Turned out it was just pure luck. I attended the London Quant Group Seminar and learned that mean-variance doesn't perform much better than assigning equal weight to each stock. This is also the time of applying GARCH model to predict volatility, of applying Exponential moving average to predict return.
4. November: Accounting approach
Then there were these dudes Thanh and Nam told me about reading firm balance sheet to invest on stocks. I read about P/E ratio on investopia. Once again I felt like this makes totally good sense. Later on I found a dbook, together with a website that already analyse these information and produced an efficient portfolio. http://magicformulainvesting.com. But I know I don't have comparative advantage at this approach.
5. December: Elliot Wave (Most exotic)
This was when a PhD student I stayed with told me to try out this method. Yet another time I felt "ha this is the right method". Then there was me watching the documentary "Trader" of Paul Tudor Jones, a rare authentic trader documentary. It was about him making huge profit in the 1987 crash. He found a pattern that was seen in the 1920s. I never got enough time and motivation to learn this exotic method.
6. January: Financial Time, Bloomberg (Most layman)
This is when I came back to the FTSE100 trading challenge. One night I red the news of Tesco selling horse meat. I shorted Tesco stocks, and made profit the next day. But this idea was short-lived in my head again, as I didn't feel right scanning the news at night just to find a news related to some stupid company.
7. February: Real Time News (Most exiting)
I was engaged in the LSE XEMarket FX Competition. It was an amazing time. It started with me and my dorm mate going to a trading seminar at Amplified Trading. The trader there analysed real-time news such as unemployment claim, FED comments to trade. The feeling was like "Eureka". I came back, found out RanSquawk that provide data the moment it was released. Better yet there was a guy reading on the radio, I felt like I was living on a trading floor. I was counting seconds to the release of German PMI, and then shorted the EURUSD when the number was lower than expected. It went on amazing. Made a 68% profit by trading Gold and bunch of forex pair.
8. March: Automated Trading (Most promising)
This idea came up when I found out I could download 1-minute data of EURUSD to my computer. I was reading an article about different Forex trading method, one of them is MACD. I tested the method on Matlab. To my surprises, 70% of the time the indicator predict correctly the direction of EURUSD the next minute. I limited the testing window to the recent 5 years, 2 years, and they all gave around 70% of correct prediction. I remember back to last summer when I was talking to Hoang Anh, if I could find a trading strategy that works 60-70% of the time, I'd be on top the the world. That was one of the moment that I felt I was standing on top of the world. (of course as it turned out my past self was naive yet again). But I thought to myself I am gonna find a great indicator. I am gonna backtest and implement a model myself.
9. April: Statistical Arbitrage (Most professional)
A bunch of different factors introduced me to this new method. I think I have found the method that mysterious hedge funds out there are using. The idea is simple, for example Gold spot price and Gold miners stock often move in line. If Gold spot price gets too high, we short the gold and buy the gold miners stock. In general, if I can test (Dick-Fuller) that a portfolio of stocks (that cost 0$ to finance) is stationary, I can create a trading strategy that short the portfolio when it reaches some high limit, and buy when it reaches some low limit. This is the idea that was used by Long Term Capital Management. This is the idea that use advance time series model. I think I have finally found "the magical formula" for trading. But of course my naive self could be wrong again.
So what's next for my future me: master time series, find a market-neutral portfolio, backtest the strategy, and write an automated trading program. Sounds so damn cool. May be in the future incorporate fundamental news as well.
Hopefully I'll be like the dogs in the race toward a successful trading strategy
No comments:
Post a Comment