- Curve fitting can make your backtesting look like it will work but when activated it has been so tailored to the data that it doesn’t preform well in the future.
- 10 years of data as this will give enough ups and downs to know how it will react.
- It’s similar to trading against the trend. Just because it may appear to have operated that way in the past it may not always continue that trend.
1 Like
1.Curve-fitting - the strategy is optimized for events that happened in the past, but won’t be able to produce positive results in the future.
2.Ideally 9 to 11 years of data should be used to make the strategy more reliable.
3. With asymmetric trading signals you have separate criteria for long and short trades which makes the strategy prone to curve-fitting.
1 Like
Creating a perfect historical model which may not neccesarily work for the future (curve fitting)
-
How long should a testing period be if you are serious about building a profitable trading strategy ((
At least a year if not more (article stated 9-11 yrs) -
Why should you avoid asymmetric trading signals?
No guarantee they will continue in the future
1 Like
- Making a strategy that will only fit past data - curve-fitting.
- 9-11 years.
- It increases the strategy’s degrees of freedom and makes it excessively prone to curve-fitted solutions.
1 Like
1 What is so dangerous about over-optimization?
If you tweak your strategy to fit past historical data to much it may not work on new data
2 How long should a testing period be if you are serious about building a profitable trading strategy?
In the blog post he says 9-11 years but with crypto being so new I would try 6 months
3 Why should you avoid asymmetric trading signals?
I believe you can use asymmetric trading signals as a strategy When the market is being affected by unusual political conditions
Usually, fundamental analysis will help you figure out when is a good time to find asymmetrical opportunities. only investing a small amount as a risk reward . The 1% Risk Rule
but asymmetric trading signals should be avoided because its potential for curve-fitting
1 Like
- What is so dangerous about over-optimization?
When you change little things in your strategy based on a specific timeframe, you risk doing curve-fitting. - How long should a testing period be if you are serious about building a profitable trading strategy?
The article says about then years. - Why should you avoid asymmetric trading signals?
There’s simply little chance of working out in the future.
1 Like
- What is so dangerous about over-optimization?
Complexity is the mother of curve-fitting. Whenever you give a strategy enough degrees of freedom an optimization will yield curve fitted results. The less complexity and less parameters available within a given strategy the less probable it is that it will ever be curve fitted as systems that don’t have complex criteria tend to be unable to “fit” to the data if a true inefficiency is not present. It therefore becomes extremely important to code simple “elegant” strategies in order to avoid added complexity which will result in curve fitted solutions.
- How long should a testing period be if you are serious about building a profitable trading strategy?
Optimizations should be carried out for long periods of time, ideally 9-11 years of data should be used for the process in order to ensure that a large amount of market conditions become available. If a simple strategy yields profitable results across a ten year period then the probability of curve fitting is greatly reduced as the system has limited degrees of freedom to artificially “fit” all those different market conditions. However a very common practice in system development is to have a certain amount of historic data “outside” the optimization set – usually one year or two – in order to perform a simulated “forward test” (commonly referred to as out-sample test) to see how the strategy behaved under new market conditions without being able to artificially “fit” into this data.
- Why should you avoid asymmetric trading signals?
By developing simple symmetric strategies with limited degrees of freedom and reliable simulations over long periods of time with one or two years of out-sample testing the possibility to find a curve fitted solution will be extremely unlikely.
1 Like
-
What is so dangerous about over-optimization?
Curve-fitting = Generating trading strategies with absolutely astonishing results that will not be achievable going forward. -
How long should a testing period be if you are serious about building a profitable trading strategy?
9-11 years. -
Why should you avoid asymmetric trading signals?
It makes it excessively prone to curve-fitted solutions. Furthermore, separating criteria for entering and exiting short and long trades can be true for some time, but on the long haul, they are the result of interest rate differentials or such similar macro economic variables that inevitably change through economic cycles.
1 Like
1.- What is so dangerous about over-optimization?
The main danger is the curve fitting. This happens for example when you develop a profitable trading system, and the simulations provide excellent results, but when when putting it to work on a live account the performance is far from expectations and not for the good.
In this case the most probable explanation is that you have curve-fitted your system to past data and therefore your system is unable to behave similarly on a different data set.
The term curve-fitted deviated from the fact that any given data set cab be accounted for by a given mathematical function of arbitrary complexity, so you can always find a mathematical function which can predict with absolute accuracy all the items of a data set.
2.- How long should a testing period be if you are serious about building a profitable strategy?
The longer the period the easier it is to avoid curve fitting as the system has limited degrees of freedom to artificially fit all market conditions.
In this sense a period of 9-11 years is good to ensure that large amount of market conditions becomes available.
3.- Why should you avoid asymmetric trading signals?
Because it increases the strategic level of freedom of the system making it excessively prone to curve-fitted solutions.
1 Like
- Your strategy may work based on past data but may not work on future data as it is unlikely to be identical.
- Nine to eleven years
- The more complexity your system has the more prone it will be to curve fitting
1 Like
- It can make the algoryth curve fitted on that particular asset.
- 1 year to even 10 years.
- It increases the freedom of the algoryth and again it may cause a curve fitted algo.
1 Like
- What is so dangerous about over-optimization?
It is dangerous because over-optimizing one particular model data-set will not give an accurate assessment for an algorithmic solution. - How long should a testing period be if you are serious about building a profitable trading strategy?
9-11 years of data should be used for the process in order to ensure that a large amount of market conditions become available. - Why should you avoid asymmetric trading signals?
Because the risk must be equal to the potential reward. Profit potential can be the same as profit loss.
1 Like
- What is so dangerous about over-optimization?
Complexity is the mother of curve-fitting. The less complexity and less parameters available within a given strategy the less probable it is that it will ever be curve fitted as systems that don’t have complex criteria tend to be unable to “fit” to the data if a true inefficiency is not present - How long should a testing period be if you are serious about building a profitable trading strategy?
Optimizations should be carried out for long periods of time, ideally 9-11 years of data should be used for the process in order to ensure that a large amount of market conditions become available - Why should you avoid asymmetric trading signals?
Asymmetric system; Asymmetric information can lead to adverse selection, incomplete markets and is a type of market failure.
1 Like
- 1 According to the article “over optimization” using historic data produces a strategy that is most likely taylored to that data and not neccessarily effective in future trades.
- 2 9-11 years. These numbers don´t look good together…
- 3 I hope i got this one right… It seems that regardless of the direction in which the price is moving there should be one single ruleset. If i find different rulesets in my strategy i most likely “curve-fitted” it
1 Like
- What is so dangerous about over-optimization?
A. Over-optimization, aka curve fitting, is when one adapts their strategy to a data set to account for previously unhandled market events. It limits the profitability of a strategy to the dataset it has been adapted. - How long should a testing period be if you are serious about building a profitable trading strategy?
A. 9 to 11 years of data - Why should you avoid asymmetric trading signals?
A. Asymmetric signals are generally defined based on current (or test data sets) market behaviors. Those behaviors may change over time leading to a less or non profitable strategy.
1 Like
-
Curve fitting and over optimisation is dangerous as you are assuming that the market will behave in the future almost exactly like it has in the past. As the market is almost never the same, the risk increases significantly and you will most likely lose money if running this strategy.
-
9-11 years is ideal to provide data that spans a wide variety of market conditions.
-
Simple and elegant is the best way to create a strategy. When more complex data and analytics are added, the risk of curve fitting starts to become a reality. Using Asymmetric signals can add to much complexity to an algorithm that may result in over optimisation.
1 Like
- What is so dangerous about over-optimization?
there are many factors that go into building a profitable strategy and following the tips included in this article will save you lots of time by preventing common mistakes. - How long should a testing period be if you are serious about building a profitable trading strategy?
9-11 years of data - Why should you avoid asymmetric trading signals?
it is wise to avoid asymmetric signals to minimize degrees of freedom and curve fitting.
1 Like
- What is so dangerous about over-optimization?
Curve fit a trading system to the past given the fact that the future may have absolutely no relationship with it. - How long should a testing period be if you are serious about building a profitable trading strategy?
9-11 years - Why should you avoid asymmetric trading signals?
Adding separate criteria for longs and shorts automatically increases the strategy’s degrees of freedom and makes it excessively prone to curve-fitted solutions. It may work for the past data but there is no guarantee that it will work for the future as well.
1 Like
1.Over-optimization can lead to curve-fitting, one of the most threats of over-optimization.
2.Ideally 9-11 years of data should be used in order to build profitable trading strategies. For Bitcoin, I would say looking at 6-12 months should be sufficient.
3. Adding seperate data for short and long trades is vital to avoid curve fitting.
1 Like
-
What is so dangerous about over-optimization? You can end up curve fitting, creating unrealistic results for future market movement and will have unsuccessful trades. Avoid 30 mins time frame and very low take profits that are less than 10x the spread.
-
How long should a testing period be if you are serious about building a profitable trading strategy? Ideally 9-11 years of data. A simple yield that is successful over a long period with less freedom is more likely to work.
-
Why should you avoid asymmetric trading signals? You create curve fitting outcomes by being more specific for example using a Moving Average of 20 weeks for longs and 14 weeks for shorts.
1 Like