How to design an Alpha ? Any tipps or guidelines ?

Back

Im struggling with creating alphas a lot recently... Im sticking together components that should work and generate a decent amount of profit but they fail miserable everytime... and i have no clue what im doing wrong...

For example the following case... im mostly trading forex, the big major pairs... using price action and the macd. And the strategy is working fine, i backtested it sucessfully in "TradingView" a trading platform where you can develop and backtest python bots.

Heres what i tried with the alpha streams.

 

Alpha : MACD

Universe : G10 ( forex )

Portfolio : Equal Weighting ( Loss minimization )

Execution : Standard Derivation ( Only execute if trend from MACD is confirmed )

Risk : Trailing Stop Drawdown ( Cut losses once trend ended )

 

Letting the MACD & the Standard Derivation working on daily charts...

The result ? Well... years of losses and one "small" profit of 12% in the end... The winning/losing rate is also pretty bad... its more like gambling ;) 

What do i miss by designing alphas ? Any good tipps and tricks from the community ? What are indicators that tell you what part of the alpha can get improved ?

 

 

Update Backtest








 
0

The material on this website is provided for informational purposes only and does not constitute an offer to sell, a solicitation to buy, or a recommendation or endorsement for any security or strategy, nor does it constitute an offer to provide investment advisory services by QuantConnect. In addition, the material offers no opinion with respect to the suitability of any security or specific investment. QuantConnect makes no guarantees as to the accuracy or completeness of the views expressed in the website. The views are subject to change, and may have become unreliable for various reasons, including changes in market conditions or economic circumstances. All investments involve risk, including loss of principal. You should consult with an investment professional before making any investment decisions.


Hi Lars,

For an Alpha to be consistently profitable, we need to have a statistical edge in the market. It may be hard to achieve Alpha with a single technical indicator for buy/sell signals, so I suggest trying various combinations of indicators, setting aside a period for testing after a profitable strategy is found, so we can test this strategy on the unseen period to see if our algorithm has been curve fitted. Furthermore, a good source of Alpha can sometimes be found in research papers, which can be found from sites like SSRN (search “trading”) and arXiv q-fin, but please note arXiv papers are technically challenging.

Best,
Shile Wen

0

The material on this website is provided for informational purposes only and does not constitute an offer to sell, a solicitation to buy, or a recommendation or endorsement for any security or strategy, nor does it constitute an offer to provide investment advisory services by QuantConnect. In addition, the material offers no opinion with respect to the suitability of any security or specific investment. QuantConnect makes no guarantees as to the accuracy or completeness of the views expressed in the website. The views are subject to change, and may have become unreliable for various reasons, including changes in market conditions or economic circumstances. All investments involve risk, including loss of principal. You should consult with an investment professional before making any investment decisions.


Hi Lars,
hint: most technical analysis methods and technical indicators don't have an edge.
I'd suggest thinking creatively, using your imagination, learn all you can about the financial system, market instruments and markets: equity, debt, forex, derivatives: futures, options; volatility, market structure, market types, market players, market makers, market relationships, order flow, risk management, psychology, bubbles, manias, macro economics, monetary policy, history of finance, hedge funds and so on.
Learn how to trade from pro traders with proven track records, either algo or discretionary/systematic.
Be wary of scam artists.
Learn everything about algorithmic system development.
Think about diversification by instrument and system.
Think about asymmetrical risk versus reward.
Backtest, walk-forward test, forward sim test and live test.
Best,
ES.

1

Update Backtest





0

The material on this website is provided for informational purposes only and does not constitute an offer to sell, a solicitation to buy, or a recommendation or endorsement for any security or strategy, nor does it constitute an offer to provide investment advisory services by QuantConnect. In addition, the material offers no opinion with respect to the suitability of any security or specific investment. QuantConnect makes no guarantees as to the accuracy or completeness of the views expressed in the website. The views are subject to change, and may have become unreliable for various reasons, including changes in market conditions or economic circumstances. All investments involve risk, including loss of principal. You should consult with an investment professional before making any investment decisions.


Loading...

This discussion is closed