Hi Rick
Have a great holiday mate - good luck with those wolves (I'm picturing a Sean Penn taking a picture of the Snow Leopard in the Himalayas in the Secret life of Walter Mitty). I hope all here have a great Christmas period/break too. Looking forward to hearing about your other strategy Rick, and your WA trader if he drops in.
The Tradeview conference I have to say was excellent. It is expensive and I had a question mark in my mind about its value - I justified signing up by making the investment case that if I can make the conference fee back from what I learn from it, then it would be worthwhile. From what I saw yesterday and the 3 algo's that were presented, I believe I can achieve that (and if so will justify going next year - they are planning to include a HFT speaker next year, which would be interesting).
As I said, there were 3 speakers, and each was quite a different style.
The 1st speaker was Tradeview's head of trading/tech support, and his presentation covered an approach based on market geometry which looks at the repeating angles that can be seen in breakouts and retraces, which alludes to a time and price perspective. I thought it was very interesting and got some insights I hadn't considered before. The EA he presented was very simple.
The 2nd speaker was a quantum physicist and he covered a fascinating pairs strategy which had some complex maths behind it but it basically boils down to a simple concept of how correlated markets move together. The EA he presented was the most complicated of the 3 presenters but having said that it still does not have a lot of code - thanks to the fact that Tradeview have built in 3 custom indicators that the EA needs, plus a watchlist GUI from which you can monitor a large range of markets showing when the EA can trade a given pair or not (whether it trades or not is conditional on the market conditions being "good" for correlated price movement).
The 3rd speaker is an ex world algo trading champion and he covered his methodology for developing trading strategies (I was pleased to note that it isn't too different to the process I am using) and gave 3 separate EA's as examples of different aspects of his strategy development process. Again his EA's were simple.
I really liked the pairs trading strategy the best, mainly because I hadn't thought about pairs trading before, but the EA and indicator/watchlist package is very slick and I am looking forward to testing this one.
Common themes from all 3 presenters:
- Keep it simple, avoid developing EA's that are overly complicated with lots of variables
- Don't over-optimise - they all at most optimised only 1 or 2 variables (mostly only 1)
- Don't spend forever backtesting trying to develop a holy grail EA - there is no such thing
- Money management and risk management can turn a poor/average strategy into a good/excellent strategy
- Rather than spend forever developing a holy grail EA with a perfect backtest equity curve, keep it simple and be satisfied with a good EA that makes a profit (at least 1:1 return:drawdown)
- There is no such thing as a holy grail strategy that works well in all market conditions, so rather than obsess on a single perfect EA , build a portfolio of 10 good EA's - diversification is the closest thing there is to a holy grail and is the best tool to build consistent returns
There's probably other stuff I haven't remembered, there was a lot to take in, but the great thing is we get a video of the conference and the presentations and the EA's each with build instructions, so I am sure there is more to sink in.
Cheers, Sharks
