Why I started to backtest my strategy

Maybe because of the current situation with my portfolio, I started to trust less my trading strategy and the way I judge the entry and exit points of the transactions.

The concept is not new. Backtest refers to the method of testing a strategy or robot, on previous historical data with the assumption that past results can be an indication of further gains (same principle like technical analysis). Basically you re-run your strategy on old data (for example 01.01.2018-05.02.2018) and see what would you get. It is recommended to simulate with more months than one and on multiple currencies.

Many traders do this 1-2 hours before even start looking into transactions for the current week/day/period. It is like practicing for an exam. It is training, in a comfortable environment, with the sole purpose of exercising your strategy (and maybe focus).

Part of the emotions (from the actual trading) is reduced or missing, as you use a paper account, but the benefits from better judging your entry transactions can’t have a price (as you will have further gains in real-life situations). Another great benefit is that you can simulate one month of activity in 20-30 minutes ( even 1 minute) depending on what are your main objectives.

Now, as I tried to explain the fundamentals, we get to the part that raises questions. How exactly we do that?

How??

There are multiple software tools on the open market, some free, some with a demo period, some super expensive (for me) that let you charge a strategy, or let you trade on historical data.

As I am already using TradingView in my analysis, I preferred sticking with it and not complicate my life with an additional tool ( even if there are better available). I recently discovered that TradingView platform has a bar reply feature, that lets you go in the past, set the speed of printing a bar, press play, and let it work.

The bar replay function, with play, pause, skip to end (current date) and bar speed

This lets me simulate the history of a currency from a certain date, to present. I chose to begin my testing from 01.01.2018 (it is also a limitation of the number of history bars the platform can play, so for 1H we have this date, for 4h we can go further in time…or backward?).

08.01.2018 (Time Travel)

This also comes with a price as the bar replay for intra-day periods is not free of charge. The good part is that I get 1 free month trial and then the price for the next month is equal with the value of 1% from one open transaction. If I see the benefits long term I’m decided to go with it(at least until I find a better solution).

As a minus, the platform didn’t integrate a paper account function in this Bar Reply feature, but this won’t stop me and I found a workaround. They do have paper account trading but it works only on current data. In a bar reply situation, if you use the account, the transaction will reflect on the current price.

The solution

The workaround I found is related to the open buy/sell indicators I was using and an excel report I have on the other screen.

As you saw in the picture the open buy/sell feature lets me select my account size, how much I want to risk on trade (the stop loss pips value will reflect the % of the account), the open price, the take profit price, and the stop-loss price. For all the prices we can see the number of pips.

The screenshot from the second monitor shows an excel that contains the number of trades I took, the currency name, the type of trade, the profit made, the value of the account after the trade, the date(I stopped the transaction or simulation) and some judgment (as I saw I fail a lot and want to document what I did bad. Like don’t do this anymore).

How does it work?

I’ve started with a 1K account and want to grow it. I am not interested in leverage, free margin and position sizing as all is set in percents automatically. Each time I do an analysis, I mark my desired trades on the graph, I hit play and let the bars to run. I hit pause each time I want to: make notes of a new trade, end a current trade, document the progress.

Each time a trade is finished (in either way: t.p, s.l., cut short, and I even have a Trailing Stop system in my head) I readjust the account value in the platform. It takes some time, it is not the perfect system, but it works for me.

As I like making lists, let’s do a pro and against bullet list for both the backtesting strategy and testing on the platform.

Pro:

  • I can really practice my entry and exits and see the transaction how it plays (instead of cutting it short because of uncertainty and fear)
  • it helps improving my moral and the rate of winning trades by cutting short wrong entries
  • I can practice money management and see the results, long term
  • keeping a log of the transactions will let me see the ones that produce results and in what conditions ( afterwords, I can apply the Pareto principle to reduce the number of wrong entries)

Against:

  • the amount of work to make the log, the transactions and adjust each time the account size
  • the need to keep an excel file to track of the transactions
  • paid service
  • not being a real account the emotions are faded (or inexistent)
  • to give a real feeling of growing an account, cross currencies, I should simulate each pair on month, then recalculate the account size each time or consider the 1% value at the beginning of the month and use it as reference for all simulations ( readjust it after 1 month of the simulation)

I will try my best to finish the simulation for 10 currencies in 2 weeks and present to you what are the results. I hope they are worth reading.

Thank you for reading this!

Disclaimer: I am not a financial consultant, all the information you find here are my decisions, I taken at that moment, on my own analysis. I am open to any type of discussion about money. If you want to replicate my portfolio take into consideration that it is your money.

2 thoughts on “Why I started to backtest my strategy”

  1. Nice article, but I have to disagree with two points:
    First, testing a strategy for old data is definitely different from technical analysis as technical analysis only refers two price patterns based on the graphs.
    Second, you should backtest for far (!) more than 1 month. If you could, you should test it for several decades that include both bear and bull markets.

    Best regards
    Jan from Manage your Money

    Liked by 1 person

    1. Thank you, Jan,
      I totally agree with you and would love to do that. But for the moment the paid plan I have, only covers the last two years from data history and I would have to buy another software that will give me this advantage. But I think that backtesting with T.A. each time would improve my way of analyzing the graph and identifying the pattern (or at least the trend and reversal). But I would love to know more, why do you think they are different?

      Thank you again for your help!

      Like

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.