Over optimization occurs when the creator of the algorithm bot tries to over-perfect the outcomes of the simulated trading results in a backtesting scenario. Basically, the programmer tries to fit the algorithmic scheme to perform historically well. Unfortunately markets don’t necessarily repeat the past performance. We have a saying, "past performance doesn’t indicate future results.
Over-optimization generally produces an algorithm that will not perform correctly under a wide range of market scenarios and can result in very poorly executed trades. These trades, because they are automated and performed by the program, can and usually result in large losses and overly risky trades based on inaccuracies and fallacies and assumptions within the program itself.
Generally speaking, profitable trading strategies will stand the test of long periods of time within a particular market. Backtesting a strategy over an extended period is wise because one can measure the effectiveness over multiple cycles in the assets history. 8-10 years is considered the industry standard.
Asymmetrical trading signals can be particularly dangerous as they apply different parameters to execute trades in different types of market structures. A computer program doesn’t know what type of cycle a market is in. it’s hard enough for humans to seewhen the markets are changing, and having different parameters for different types of trades overcomplicates the strategy and leads to poor decisions.
The best trading strategies are consistent, simple, elegant, and functional. Trying to be perfect with a strategy or an algorithm can be very dangerous, leading to lots of loss and inconsistency.
-
What is so dangerous about over-optimization?
Curve-fitting the algorithm to a specific set of data. Such an algorithm is likely to under perform with a new set of data. -
How long should a testing period be if you are serious about building a profitable trading strategy?
According to the article, which I assume was written for stocks, 10 years. For crypto, I’d consider the creation date of the crypto, market cap, volume, and the general market trends to determine the relevant testing period. I’d also keep a sample of the relevant data out of this testing period used in the optimization process to run a verification test at the end of the optimization process. -
Why should you avoid asymmetric trading signals?
They are often driven by past data and are poor indicators for future data, thus they lead to curve-fitting.
- you could be curve-fitting your strategy, and it less likely to perform well in the future
- 5-10 years
- it is a sign of a curve-fitted strategy
- What is so dangerous about over-optimization? Curve fitting. Do not want to generate 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? 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? There are many different small things you can do wrong when performing optimizations, such as using highly correlated procedures, fine grid tests or ignoring the “surroundings” of your intended results. A very important part of avoiding curve-fitting is to avoid these common mistakes to perform coarse and efficient optimizations that do not predispose your system towards curve-fitted solutions
-
What is so dangerous about over-optimization?
You can run into curve-fitting. It’s when your model is over optimized for past data and not future markets. -
How long should a testing period be if you are serious about building a profitable trading strategy?
Testing periods should be at least 9-11 years of data used. -
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.
-
What is so dangerous about over-optimization?
Over-optimization makes so your model only works for past data but it’s completely useless for future data. Meaning your strategy will not be profitable. -
How long should a testing period be if you are serious about building a profitable trading strategy?
Period should be around 10 years. -
Why should you avoid asymmetric trading signals?
This makes it more likely that the model is curve-fitting as it introduces more degrees of freedom.
- What is so dangerous about over-optimization?
Can lead to curve fitting where good results are generated from back testing past data that cannot be replicated in the future.
- 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?
This increases the strategy’s degrees of freedom and makes it excessively prone to curve fitting.
1)It has the potential to ruin a trading careers if not dealt correctly.
2)9 to 11 years of data
3)It tend to lead to curve fitting
-
What is so dangerous about over-optimization?
that the most common mistake when doing optimization is -without a doubt- the length of the testing period used to optimize. Strictly speaking, optimizations are not bound to be meaningful fit they are done within periods of less than 5 years given that smaller periods of time are not statistically relevant according to long term changes in market volatility. So if you want to optimize your system and avoid curve fitting, use a period of at least five years. Using a smaller period will most likely “fit” your strategy to very specific market conditions and will make it unable to perform correctly as the market changes. -
How long should a testing period be if you are serious about building a profitable trading strategy?
5 years but I believe you should always review it every 6 to 12 months for peak performance -
Why should you avoid asymmetric trading signals?
We should avoid asymmetric trading signals because the asymmetric can eventually lead to curve-fitting in the long run.
- It leads to curve-fitting based on past data
- 9 - 11 years of data
- They increase complexity and freedom of the trading strategy and result in curve-fitting
1.over optimization is dangerous as it can lead to curve fitting
2. 9-11 years
3. increased degree of freedom and complexity makes it prone to curve fitting
- What is so dangerous about over-optimization?
Over optimization can lead to “curve-fitting”. This will result in un-fit model that does not function properly and may result in significant losses. - 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?
Adding separate criteria for longs and shorts automatically increases the strategy’s degrees of freedom and makes it excessively prone to curve-fitted solutions.
1. What is so dangerous about over-optimization?
The danger of over-optimization is that you aren’t really making things better at predicting you are simply making it better at fitting the currently known data set.
2. How long should a testing period be if you are serious about building a profitable trading strategy? The recommendation was to have a decade of data to analyze and should not be done with very tight constraints (not looking at minute by minute trends)
3. Why should you avoid asymmetric trading signals?
Increasing the complexity allows for more things to go wrong and increases the probability that you have only best fit your strategy to the data.
-
What is so dangerous about over-optimization?
The biggest danger resulting from over-optimization is Curve-Fitting where data too accurately reflecting the past is inadequate for predicting the future. -
How long should a testing period be if you are serious about building a profitable trading strategy?
Simulations should be run on periods which are as long as possible.
The longer the trading period, the more statistically significant the data set is and the less likely it is to show curve fitting.
Ideally periods of 9 - 11 years are preferable. -
Why should you avoid asymmetric trading signals?
Using asymmetric trading signals automatically increases the strategy’s degrees of freedom and makes it excessively prone to curve-fitted solutions.
-
The danger of over-optimization is that indicators and tools are used in a specific way that brings back the highest yielding results based on the past data. This level of manipulation will make future results much less usable.
-
9 to 10 years in order to capture multiple market conditions over time.
-
Using asymmetric trading signals increases the strategies degrees of freedom and makes it excessively prone to curve-fitted solutions.
- What is so dangerous about over-optimization?
We don’t want to generate trading strategies that gives good results with historical data but yet unusable, non-profitable or impractical in future/on-going market scenarios.
- How long should a testing period be if you are serious about building a profitable trading strategy?
A testing period should be around 9 to 11 years of data, these includes the various market cycles and economic events that might have affected pricing.
- Why should you avoid asymmetric trading signals?
Asymmetric trading signals have a tendency to be derived from curve-fitted solutions and optimizations, which should be avoided altogether.
- What is so dangerous about over-optimization?
High risk for curve fitting, or that the system will get so limited that it will hardly run at all ++
- How long should a testing period be if you are serious about building a profitable trading strategy?
Ideally 9-11 years if possible, 5ish years on crypto or as long as possible.
- Why should you avoid asymmetric trading signals?
Increases the strategy’s degrees of freedom and danger of making it curve-fitted.
- curve fitting = just because you can find a model that fits into past data does not guarantee it will apply to the future.
- 10 years of past data would be great.
- Strategy should be simple and elegant instead of complex.
1. What is so dangerous about over-optimization?
Over optimization will result in curve fitted date which may have been good for past data but won’t be achievable in the future.
2. How long should a testing period be if you are serious about building a profitable trading strategy?
9-11yrs
3. Why should you avoid asymmetric trading signals?
Having asymmetric trading signals will increase the degrees of freedom which makes it prone to curve-fitted solutions.
- What is so dangerous about over-optimization?
It may lead to “curve-fitting” that produce a strategy that is only tailor to specific past data. - 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 may lead to “curve-fitting” because asymmetric signal works for past data may not work for the future.