Compared to a guess, can we really say it is superior

Introduction

In the previous article, we explained the meaning of swapping the order in the second test, the “Monte Carlo shuffle.” This time, we will discuss the third test, the “random comparison.”

What is random comparison

In this test, we compare the performance of trades made under actual conditions with the performance of entries made without any conditions—chance entries. Even if you decide the timing of buying and selling completely at random, you may occasionally get good results by chance. In periods where the market itself has a bias toward rising or falling, you can obtain “plausible” outcomes with a certain probability even without choosing a direction.

Why compare to a “guesswork”

If you only look at verification results that use conditions, you may want to think, “This performance came from having these conditions.” However, if that performance falls within the range that could be obtained even with guesswork, the conditions themselves cannot be said to have meaning.

In random comparison, we create the distribution of performance when random entries are repeated many times, and we check whether the actual performance under the given conditions falls within the common range of that distribution or is clearly out of position.

What cannot be judged without a comparison

A sense of “goodness” cannot be judged at all without a basis for comparison. Random comparison serves as a kind of yardstick for us to determine whether the conditions themselves truly have meaning.

What you might have felt so far

We have seen many conditions in our previous validations that even fail against guesswork. That is precisely why we cannot omit this comparison. In the next article, we will discuss the fourth test, the “parameter perturbation,” and explain the meaning of lightly tweaking numerical values.

Disclaimer

・The content published in this article is not intended to endorse or provide investment advice for any specific trades. It is for information sharing only.

・Verification results based on historical data do not guarantee the same results in the future. Markets always change.

・Whether you actually trade is entirely your responsibility.

・The indicators and tools introduced in the future are one of the factors for market assessment, and using them does not guarantee profits.

If you are interested, please also take a look at the product page of the Indicator Institute.