Konstantin Boykachev

CEO Proforexea LLC

Honest Coder

Professional Trader

0

No products in the cart.

Konstantin Boykachev

CEO Proforexea LLC

Honest Coder

Professional Trader

Blog Post

How skipping deals affects the overall effectiveness of the strategy

[ad_1]

effectiveness of a trading strategy

Tell me, have you ever had such a situation: you are distracted from the platform for a few minutes or even hours, come back and see that the right moment to open a deal has been missed? Agree, it’s a shame. But how does such a pass really affect the overall effectiveness of a trading strategy? Maybe skipping a few deals is not so catastrophic? Or, on the contrary, it leads to a sharp drop in the profitability of the system? We will try to answer this question in this review.

The effectiveness of a trading strategy with different skipping deals

Testing a trading system on historical quotes implies fixing a separate time period and opening deals on it. For example, over the past 3 years. Testing can be manual or automatic, and it differs from the human factor.

If you write an Expert Advisor according to a manual strategy, it will open deals according to the specified algorithm, but the trading principle will be different. The robot opens a trade with a clear signal. A person thinks for a while, doubts. Or just skips a signal. As a result, a robot conditionally opens 100 transactions in 3 years, a person manually – only 80.

At first glance, it may seem that the effectiveness of the trading strategy in percentage terms will remain. Indeed, among the 20 missed trades, there are both profitable and unprofitable ones. And with a normal mathematical distribution, the ratio of winning and losing trades will remain. What if a trader misses not 20 deals, but 50? Or 80? By the way, we recommend a useful article: “How many transactions should there be to check the stability of the system”

One of the well-known analytical portals recently hosted a scientific tool that was previously conducted for casinos and the stock market by Edward Thorpe. An advisor was taken as a basis, which was run on the same site, changing the “sensitivity threshold” in the settings.

Three types of settings were used: opening 100% of trades, 85% and 50%. That is, in the third case, the EA opened 2 times less deals. The Expert Advisor was launched on 6 currency pairs in the “last 20 years” interval. Based on the results of the run, a histogram was built, which showed the total profit and profit with a monthly breakdown.

Result:

  • The results are on average the same for all currency pairs. The volatility and type of asset (in particular, the currency pair) does not affect the percentage and the mathematical result.
  • Missing 15% of transactions does not cause significant damage to the account. The spread in performance ranged from 0.2% to 0.8%. If, with the maximum possible 100 signals, the trader misses 10-15 signals, this will not lead to a loss or to a significant change in the percentage level of profitability.
  • Skipping 50% of trades can turn a profitable strategy into a losing one. The deviation for individual pairs reached 20%.

Output:

  • A profitable strategy during the test period turns into a losing one when 50% of deals are skipped… Skipping 15% of transactions is not critical.
  • Testing an Expert Advisor using a manual strategy can give radically different results… It is logical that writing an Expert Advisor and launching it in the MT4 tester is faster than manually opening hundreds of deals. But that’s the point. Option: run testing the EA in conditions as close as possible to manual trading. For example, only at certain hours of time.
  • The same goes for testing a manual strategist.and. If you plan to open trades only during the day, test it on the historical period only in the daytime, so that the number of open trades on the tested and real time intervals approximately coincides.
  • The higher the timeframe, the less likely the signal will be missed during manual testing… The effectiveness of the trading strategy practically does not change on the H4-D1 intervals. On minute timeframes, the probability of missing signals and the discrepancy in the number of transactions between the tested and real periods increases.

Ask questions in the comments!

[ad_2]

Taggs:
Write a comment