SIMULATION

Daily simulations are conducted at my end to reinforce the effectiveness of the trading strategy (as well as mental reinforcement) and to display how the trading results can vary on a daily, weekly and monthly basis. When running thousands of simulations I tend to use an in-house built Monte Carlo Analysis tool, however when simulating trading results for a single month then I tend to use the following website http://www.roll-dice-online.com. How I use this website to conduct trading simulations is as follows.

The key data point needed is firstly the mean win rate for my trading strategy which is 55%. Presently I am putting on trades where my risk is 100 and my reward is around 180 (Please note this simulation conducted in May 2018 was based on 1.8R but since Mid July I have moved towards a 2.75R model). I am also trading on average about 4 times per day. This is based on a bank of 5000 so my risk per trade is 2% of my bank which is within the acceptable norm of up to 3% in financial trading (I try to model financial trading rather than typical sports betting).

Therefore with roll-dice-online.com these are the values I enter.

1. I pick a 9 sided dice because its the closest representation I can obtain of a dice that is able to provide a 55% win rate simulation (meaning that 5/9 * 100 = 55.555%). So when rolling the dice if the numbers 1, 2, 3, 4 or 5 come up then we define that as a trading win and if the numbers 6, 7, 8 or 9 come up we count that as a trading loss.

 2. We roll the dice 4 times as this represents the 4 trading opportunities per day.

3. Number of rolls is 31 as this represents the number of trading days in most months.

4. In the tables further below I divide the days into weeks so days 1-7 is Week 1, days 2-14 is Week 2 and so on.

4. Click here for the raw data results. For day 1 (the first test result) we see 2, 6, 8, 8. This means for day 1 there was 1 winner and 3 losers.

5. Since my winning trades are 180 and my losing trades are -100 this means my profit for day 1 was -120

6. This is the logic we apply for test result 2 all the way down to test result 31 in the raw data file.

7. The only other point to be aware of is if my first 3 trades are losing trades then my daily stop loss is reached (happy to lose the battle) and I stop trading. So the maximum loss I will take in any day is -300. The tabulation and conversion of the raw data file into daily profit and loss is as follows.

8. So based on this simulation for a months worth of trades the total net profit is 8420. (I am aware there can be slippage due to betfair costs, possibly having to take slightly less profit for some trades based on how price is moving etc but the 8420 is a target).

9. Out of 31 trading days, 26 days were profitable days and 5 days were non profitable so percentage of profitable days is 83.87%.

10. The mean of the losing days is -156 and the mean of the winning days is 353.84. Therefore the winning day on average is 2.2 times more than the losing day.

11. If I plot the equity curve (based on daily results) it would look like the below.

12. If I plot the standard deviation and daily equity analysis it would look like this

13. The data clearly shows that the strategy has a strong positive expectancy despite the fact that I am only winning 55% of the time. However it is the power of the risk reward ratio and the ability to trade at least 125+ times per month which helps to ensure the strategy has great potential.

14. If you keep clicking "roll dice" it will keep presenting new sets of data so If I was to go through this exercise again the amount earned per month could be down to 6000 or up to 10,000. This is due to the science of random normal distribution so the idea behind a Monte Carlo Analysis is to simulate this experiment thousands of times and then take the average of all those simulations.

15. When you see my live daily results (click here) you can then compare it to this simulation and when you see me going 4-5 profitable days in a row with a strong equity climb instead of being in disbelief you can go back to this simulation or run your own simulation to mirror the type of results you would expect from a strategy with the parameters I provided. If you see week 3 above you will see a whole week where every day was profitable. I can also see that there is normally no more than 2 losing days per week.

16. Imagine earning on average 8400 per month. This equates to an annual salary of 100,800. Remember this is tax free as earnings from activities considered gambling by the UK government are not taxable. I know people who do IT contracting where earning 700 per day is the top end of the contracting earning scale. If you take 3500 for the 5 day contracting week it equates to an annual salary of 182,000. Deduct the 40% income tax and the net annual income is 109,200. For the sake of making 4 trades per day and not having the stress of having to be in a 9-5 job with a boss sitting over your head I know which job I would prefer.

17. The final and most important point here is this. By increasing my position sizing it naturally increases my earning potential. So if my liability instead of 100 was now 150 meaning my winning trades were now 270, imagine the increase this would make to the annual earnings. This is why I had to develop my own system rather than be limited by system peddlers wanting to convince me that its not possible to be earning more than 2000-3000 per month trading full time.

I realized I had to think completely differently to how 90% of the sports trading industry think and approach trading. It had to be a probabilistic and systematic approach based on science and not human bias.

My reward needs to be higher than my risk, my win rate simply needs to sustain the reward to risk ratio (without having the pressure of trying to be right all the time) and to avoid severe volatility/variance (randomness) for weekly and monthly results I need to be trading at least 30 times per week where possible rather than just taking the odd 1-6 trades per week and crying when variance hits and results in a losing week. Every component is a critical aspect of the whole.

If you enjoyed this simulation insight then make sure you join my biweekly webinars which begin in May. Please note I am not charging or selling anything. Instead I remember promising that if I ever made a success out of trading I would try to give back by sharing the fundamentals (click here) of what helped me because I was in a very dark place with my trading previously and know the frustration, turmoil and desperation it can breed when you are following peoples systems and tips that simply don't work long term.

You are different to every one else in the world. You are unique. So why not build something that resonates with you and works for you.

Look forward to seeing you on my webinars and will keep you updated on twitter nearer the time.

 


The key to success in trading is the ability to accept losing a battle whilst knowing you are winning the war

2015-2018 Scientia Trading | Privacy Policy | DisclaimerEmail Us | Twitter | Youtube