Today you will learn how to verify a trading strategy.
These are the same methods I use in stock trading to see if the system is reliable.
You’ve found a trading strategy, and you want to know whether it is good or bad. However, you do not know how to test it. Well, you’re in luck, because that’s what I’ll teach you in this tutorial.
Verify a trading strategy by backtesting manually or using trading software. Use Amibroker or any software, and be sure to have reliable data of the stocks. Evaluate the results to find out if it’s a viable strategy.
Question for you, did you find a good trading system? How do you know that it is doing what it’s supposed to do?
Well, you can only evaluate it through backtesting. In this post, I’m going to teach you the best practices on how to do it.
Let’s talk about what is backtesting, curve fitting, getting reliable data, and optimizing your strategy.
You need to understand that without the basics, you’ll have unreliable results, and I don’t want that to happen.
If you want to become a professional trader, you have to know the win rate, statistics and possible income or loss of a strategy before using it in trading.
Do you believe that buying books, and courses can teach you the real thing?
By the way, I’m talking about trading here. Of course, you can learn by reading books, researching, and watching online courses. However, you can’t do that in actual trading.
You have to verify everything you read and learn.
Don’t believe easily. You’ve got to protect yourself from the wrong information because there are lots of it.
The only thing you can do is evaluate the trading system yourself.
However, if you don’t like to do the hard work, you can go to our recommended strategies page.
Moreover, you can read further below if you’re interested, and you’ll learn something.
By backtesting, you simulate a trading strategy with the use of historical data of the stock universe. If you can quantify the parameters, you can test it.
For example, you can’t test a system based on good products, manager traits, and market sentiment.
There are many ways to do it and here’s what I do, and I’m warning you that it takes a lot of time.
I do backtest by using sample and out of sample data. Basically, I test the strategy on historical data, and then I paper trade on live data.
I don’t use actual money because I’m still verifying if the system is working.
After six to 24 months, I evaluate the strategy and decide if it is viable.
Moreover, after more testing, it becomes tiring and time-consuming.
Enough about me and let’s go back to the topic!
Your goal verifying a trading system is to find out whether the strategy is viable, and if it proves to be reliable, then congratulation because you found a good one.
However, how do you do that?
- First, get backtesting software.
- Second, have reliable data.
- Next, prepare your trading strategy.
- Third, start backtesting.
- Finally, evaluate.
I use Amibroker, especially when I found a new strategy. Maybe it is just me, but I like to test anything I find.
You can also use Quantopian. You can use it to code trading systems, and you can back test it.
The good thing is that their services are free with data for most countries. However, some stock exchanges are not available.
Moreover, if that is the case, just upload your own data.
At Quantopian, you have to code in Python or hire someone to code for them in case you have some extra money.
Curve fitting is optimizing a strategy too much. Did you know that you can use Amibroker to do it?
The software has a function that can make a trading system adopt to the historical data.
Well, optimize function, if abused, can overfit the strategy to test data, and it is the reason why the strategy will fail.
However, you can avoid curve fitting by using the sample, and out of sample data that we’ve discussed above.
Read it carefully and understand it, because you do not want to use a useless trading strategy.
You want to have a reliable result after working hard with your backtests.
What would you do if the result from historical data is different when you actually implemented the trading strategy?
I know you won’t like it.
According to BuildAlpha, there are three ways to avoid curve fitting:
- As I have mentioned above, use sample and out of sample data.
- The trading strategy must have enough occurrences.
- Validate the system in other markets.
- I always use these tips to avoid curve fitting. For example, for trades less than 2,000, I consider it not enough for 20-year historical data.
In addition, test the trading system in the United States and other countries. If you did your backtests in one stock exchange, then it is not enough.
I recommend using at least five markets. It may seem a lot of work, but it has to be done.
Well, if you do what is uncomfortable, you’ll find a good strategy that is not available anywhere.
When that happens, you’ll feel that it’s worth it.
If you want to verify a trading strategy, you must have reliable data. Having bad data can really give unreliable results.
The free ones are easy to get, including from our site, yet it does not have guarantees. However, even though we are providing the data for free we have been striving to share reliable data.
Nevertheless, we don’t give any warranty or whatsoever. Even so, if you accept the risk, then use our data.
Conversely, you can use Norgate’s Data, but I warn you it is expensive, but it is reliable.
Free data don’t usually have adjusted stock splits when we are talking about securities.
That is why we don’t recommend backtesting while you haven’t corrected every stock price.
If you don’t do that, you may be just wasting your time.
You can also use other vendors to verify whether the results are correct when testing historical data.
Here are the lists of other sellers:
- Quick FS
I will update the list as we find more. Well, I don’t really use other sellers, and I just listed it for your reference.
Can you provide feedback if you have used them?
However, I find Norgate better because it is the one I have used. I can load up the data in Amibroker and set my AFL code. Then, start the verification of the trading strategy to whether it is effective or not.
If you find Norgate too expensive, you can ask your fellow traders to share the costs. I did that one time to save a little more.
You just have to be creative.
Why reliable data is important?
If you want to verify a trading strategy, you must use accurate data, or else your research is useless.
Now can you can do all the things that I have shared, but it takes time and money. The cost can skyrocket. Before you find a good trading strategy through verifying and backtesting.
It can take months or years if you are lucky or not.
However, if you’re willing to work hard you’ll find a good trading strategy that you can use in real trading.
Nevertheless, there is a better way to all of this. Well, if you know something that could fast-track your success will you do it?
Of course, you will, if you could save time and money why not?
I know because I traveled to the same spot as you.
Well, I bought so many books, online courses and even took a lot of seminars.
The only trading strategies I got are garbage because it doesn’t work.
I hope and prayed that someday I’ll find a good strategy that I can use for a long time. Until it stopped working.
Until one day, I found a course that leads me to recommend the strategy. I’m not going to tell you what is it because we have different personalities.
Instead, please go to our recommended page and read what I urge you to do.
All things considered, you can only verify a trading strategy through backtesting.
However, if you want to have a system fast without wasting time and money, go to our recommended strategies above.
It is better not to waste your time doing everything yourself.
Well, I hope you learned something today.