PARMAR 3P Trading System

[stickerman]

Ok, so following my recent question about whether or not I’m doing something wrong, here are my stats:

In the last few days I have followed the original plan of .18, .06, .06. I have had a total of 6 sets of trades. Three were all failures, three were all successes. End product: +45 pips.

So for me it’s certainly a system worth persueing. Thanks to Vijay for info so far. Keep up the good work.

[defiance888]

Yes defiance888. i confused the logic and fixed. but overall backtest results since 2013 janauary to current date only 2 trades and wins.

But i wonder only 2 trades we get from 18 months. is it fine?

that too only one inside bar candle we get for 18 months.

vijay is right, there will be much more inside bars in 18 months. each month should have atleast around 3 usually. i think you are not understanding the inside bar concept too well. perhaps then it would be best to stop.

but thanks for giving it a go and everything

[vijaykumar]

In keeping to [I]Vijay’s[/I] original strategy in which there are 3 orders with 3 different profit targets and with a percentage allocation to guarantee that the first order produces a profit even if the 2nd and 3rd are a loss, I would like to suggest the following alteration to the percentage allocation which in my view is mathematically more advantageous.

However, I do need to point out that my formula is based on equal probabilities of all outcomes which I know is not true. In order to be totally accurate I would need to factor in those probabilities. However, even though I am not doing so yet (maybe at a later stage), I feel that the following suggestions would still be more favourable and mathematically more sound than [I]Vijay’s[/I] original allocations of [I]60%, 20%, 20%[/I]; which seem to be arbitrary as far as I can guess.

My suggestion is thus as follows.

Let the values be as such:
[ul]
[li]“a” represents the percentage allocation for the 1st order with R:R=1
[/li][li]“b” represents the percentage allocation for the 2nd order with R:R=2
[/li][li]“c” represents the percentage allocation for the 3rd order with R:R=3
[/li][/ul]
And the conditions to be met in order to improve profitability, should be as follows:
[ol]
[li]“a + b + c = 100%” (So that total loss is our total risk amount, no more, no less).
[/li][li]“a >= ( b + c )” (Therefore we infer from point 1, that “a >= 50%”. This is so that the first order makes a profit, or breaks-even, even if the other two are a loss).
[/li][li]“2 * b >= a” (Therefore we infer from points 1 & 2, that “b >= 25%”. This is so that the 2nd order of R:R=2 can be equal or more profitable than the first of only R:R=1).
[/li][li]“3 * c >= 2 * b” (Therefore we infer from points 1,2 & 3, that “c >= 16,667%”. This is so that the 3rd order of R:R=3 can be equal or more profitable than the 2nd of only R:R=2).
[/li][/ol]
So, in order to meet the above conditions and maximise returns, our ratios of “a:b:c” is “50%:25%:16.667%” or the equivalent of “a:b:c = 6:3:2”. So calculating the values gives us the following valid distribution:

[B]“a = 54,545%”, “b=27.273%” and “c=18,182%”[/B]

This would give the following gains for the following outcomes:
[B][ul]
[li]Loss, Loss, Loss = 100% Loss (of Risk%)
[/li][li]Profit, Loss, Loss = 9.090% Profit (of Risk%)
[/li][li]Profit, Profit, Loss = 90.909% Profit (of Risk%)
[/li][li]Profit, Profit, Profit = 163,637% Profit (of Risk%)
[/li][/ul][/B]
Compare this to the original [I]60%, 20%, 20%[/I] which gives the following:
[I][ul]
[li]Loss, Loss, Loss = 100% Loss (of Risk%)
[/li][li]Profit, Loss, Loss = 20% Profit (of Risk%)
[/li][li]Profit, Profit, Loss = 80% Profit (of Risk%)
[/li][li]Profit, Profit, Profit = 160% Profit (of Risk%)
[/li][/ul][/I]

This is just food for thought, because to really maximise returns it would be best to also factor in the probability of each outcome and adjust values in accordance to those probabilities in order to improve the profit factor. If I have time, I will do just that next time, for EUR/USD and GBP/USD for the year 2013.

Best regards,
Carnino

Thanks for your valuable time and effort in giving such precise calculation.
I appreciate it however the only problem is lot size allocation with your method.

Trader want to risk 0.2 lot size than it would be more difficult to allocate with such accurate percentage base.

Also when 1 trade is with profit and 2 with loss it gives 9% compare to original 20%.

I will backtest for last 2 month to see the difference in result.

Once again thanks

Regards
Vijay kumar PARMAR

[FMIC]

Thanks for your valuable time and effort in giving such precise calculation.
I appreciate it however the only problem is lot size allocation with your method.

Trader want to risk 0.2 lot size than it would be more difficult to allocate with such accurate percentage base.

Also when 1 trade is with profit and 2 with loss it gives 9% compare to original 20%.

I will backtest for last 2 month to see the difference in result.

Once again thanks

Regards
Vijay kumar PARMAR

As I referred to in my post, it was purely a mathematical approach where all outcomes have equal probability - [B]which is not the case[/B]. A proper approach which I intend to still do, is to factor in the probability of each outcome.

However, even with a 0.2 lot, that does not invalidate the percentages given. Obviously they will have to be rounded off, giving (0.11; 0.05; 0.04 lots) equivalent to (55%; 25%; 20%) which is an approximation.

In your case, just because for the P.L.L. outcome you get 20% return which is higher than the 9.09%, that does not mean that yours is more profitable. Also note that in the P.P.L. and P.P.P. outcomes, my gains are higher than yours. One can only verify which is more profitable if we factor in the probability of each outcome.

I have still not finished running all the statistics yet, but I can tell you right now from what I have found so far, is that the best distribution is actually the inverse to what you use. In fact it would be more profitable to use something like 18%, 27%, 55%, allocating more to the 3rd larger R:R order instead of the 1st smaller R:R order.

However, I will post the complete stats here next week or maybe even during the weekend.

[FMIC]

Thanks for your valuable time and effort in giving such precise calculation.
I appreciate it however the only problem is lot size allocation with your method.

Trader want to risk 0.2 lot size than it would be more difficult to allocate with such accurate percentage base.

Also when 1 trade is with profit and 2 with loss it gives 9% compare to original 20%.

I will backtest for last 2 month to see the difference in result.

Once again thanks

Regards
Vijay kumar PARMAR

In a follow-up to my previous post, here is just some stats to substantiate what I stated:

This is from a back-test run on the EUR/USD from 2012-12-31 to 2013-12-27 as follows:

[ul]
[li]ZimeZone: UTC+2 (NY Close)
[/li][li]Trading Days: Tuesday-Friday
[/li][li]Spread: 1.0 pip
[/li][li]Breakout Offset: 1.0 pip on both High and Low
[/li][li]Stop Size: 5.0 pips
[/li][li]Risk: 1% (Initial balance $10000)
[/li][li]Pending Orders Expire 24hrs
[/li][li]Open Orders remain open until SL or TP is hit
[/li][/ul]
Overall Statistics:
[ul]
[li]Trades: 654
[/li][li]Win Rate: 41.74%
[/li][li]Reward/Risk Ratio: 182.05%
[/li][li]Win Rate * Reward/Risk Ratio: 75.99%
[/li][/ul]
For the various allocation percentages, individual statistics are as follows:

[ol]
[li]Allocation: 60%; 20%; 20%
[/li]Profit: $3383.50 (33.8% Gain)
Profit Factor: 1.24
Drawdown: 10,95%
Avg. Profit/Avg. Loss Ratio: 173.07%
Win Rate * Avg. Profit/Avg. Loss Ratio: 72.24%
Kelly Factor: 8.08%

[li]Allocation: 54.545%; 27.273%; 18.182%
[/li]Profit: $3503.00 (35.0% Gain)
Profit Factor: 1.24
Drawdown: 10,78%
Avg. Profit/Avg. Loss Ratio: 173.67%
Win Rate * Avg. Profit/Avg. Loss Ratio: 72.5%
Kelly Factor: 8.2%

[li][B]Allocation: 18%; 27%; 55%
[/li]Profit: $5470.00 (54.7% Gain)
Profit Factor: 1.31
Drawdown: 9,81%
Avg. Profit/Avg. Loss Ratio: 182.75%
Win Rate * Avg. Profit/Avg. Loss Ratio: 76.29%
Kelly Factor: 9.87%[/B]
[/ol]
As you can observe, my percentages give more profit, but what really improves the profitability is the reverse allocation giving 54.7% Profit Gain. This is because of the probabilities of the outcomes which have to be factored in as well.

I will give more details on this at a later date and with a more accurate allocation distribution percentages for EUR/USD and GBP/USD!

Please note that I used M1 data from [I]HistData[/I] for these back-tests.

Regards,
Carnino

EDIT: Added some extra statistics!

[Bobkat]

In a follow-up to my previous post, here is just some stats to substantiate what I stated:

This is from a back-test run on the EUR/USD from 2012-12-31 to 2013-12-27 as follows:

[ul]
[li]ZimeZone: UTC+2 (NY Close)
[/li][li]Trading Days: Tuesday-Friday
[/li][li]Spread: 1.0 pip
[/li][li]Breakout Offset: 1.0 pip on both High and Low
[/li][li]Stop Size: 5.0 pips
[/li][li]Risk: 1% (Initial balance $10000)
[/li][li]Pending Orders Expire 24hrs
[/li][li]Open Orders remain open until SL or TP is hit
[/li][/ul]
For the various allocation percentages, all have an overall Win Rate of 41,74% for 654 Trades as follows:

[ol]
[li]Allocation: 60%; 20%; 20%
[/li]Profit: $3383.50 (33.8% Gain)
Profit Factor: 1.24
Drawdown: 10,95%

[li]Allocation: 54.545%; 27.273%; 18.182%
[/li]Profit: $3503.00 (35.0% Gain)
Profit Factor: 1.24
Drawdown: 10,78%

[li][B]Allocation: 18%; 27%; 55%
[/li]Profit: $5470.00 (54.7% Gain)
Profit Factor: 1.31
Drawdown: 9,81%[/B]
[/ol]

As you can observe, my percentages give more profit, but what really improves the profitability is the reverse allocation giving 54.7%. This is because of the probabilities of the outcomes which have to be factored in as well.

I give more details on this at a later date!

This is terrific research Carnino…I greatly appreciate it.

[vijaykumar]

In a follow-up to my previous post, here is just some stats to substantiate what I stated:

This is from a back-test run on the EUR/USD from 2012-12-31 to 2013-12-27 as follows:

[ul]
[li]ZimeZone: UTC+2 (NY Close)[/li][li]Trading Days: Tuesday-Friday[/li][li]Spread: 1.0 pip[/li][li]Breakout Offset: 1.0 pip on both High and Low[/li][li]Stop Size: 5.0 pips[/li][li]Risk: 1% (Initial balance $10000)[/li][li]Pending Orders Expire 24hrs[/li][li]Open Orders remain open until SL or TP is hit[/li][/ul]
For the various allocation percentages, all have an overall Win Rate of 41,74% for 654 Trades as follows:

[ol]
[li]Allocation: 60%; 20%; 20%[/li]Profit: $3383.50 (33.8% Gain)
Profit Factor: 1.24
Drawdown: 10,95%

[li]Allocation: 54.545%; 27.273%; 18.182%[/li]Profit: $3503.00 (35.0% Gain)
Profit Factor: 1.24
Drawdown: 10,78%

[li][B]Allocation: 18%; 27%; 55%[/li]Profit: $5470.00 (54.7% Gain)
Profit Factor: 1.31
Drawdown: 9,81%[/B]
[/ol]
As you can observe, my percentages give more profit, but what really improves the profitability is the reverse allocation giving 54.7% Profit Gain. This is because of the probabilities of the outcomes which have to be factored in as well.

I will give more details on this at a later date and with a more accurate allocation distribution percentages for EUR/USD and GBP/USD!

Please note that I used M1 data from [I]HistData[/I] for these back-tests.

Regards,
Carnino

Great work really appreciate your effort.
Will look and weekend in detail and re-analyze it and will change necessary things in my account as well.
Regards
Vijaykumar PARMAR

[PipLips]

I agree with bobkat. Great job on the analysis carnino! I often wondered if the allocation was reversed with the system would it make it more profitable. Thanks to you to vijay for your efforts in this forwx endevour.
Cheers,
Piplips

[Bobkat]

Any determination as to which days of the week are most profitable? Which of the 3 TP levels is the most profitable (weighted per lot)?

For my very short exposure to this trading system, TP 15 is the most profitable. YMMV

[FMIC]

It seems you are all eager to now the results. OK! I will spare you the boring statistics for now (I will post those later) and give you the “candy” now!

But are you really ready to see the final results for the best percentage allocation?

[B]Well, here it is: 0%; 0%; 100%!!![/B]
That is right – the best most profitable outcome is if one only places
[B]ONE SINGLE ORDER for the R:R=3[/B] for the full amount of risk.

So for the EUR/USD back-test referenced earlier, but with a Single Order of R:R=3 (SL: 5 pips; TP: 15 pips):

Profit: $7237.50
Profit Factor: 1.35
Drawdown: 13.74%
Win Rate: 31.65% (218 Trades)
Reward/Risk Ratio: 300%
Avg. Profit/Avg. Loss Ratio: 292.28%
Win Rate * Reward/Risk Ratio: 94.95%
Win Rate * Avg. Profit/Avg. Loss Ratio: 92.51%
Kelly Factor: 8.27%

[B]I will let you know about even better results when you let the pending orders expire MUCH later!!![/B]

[B]EDIT: Warning!!! Please read my next post!!![/B]

EDIT: Added some extra statistics!

[FMIC]

[B]EDIT: My apologies but I expressed myself incorrectly in this post. Please read my next one too for more correct summation![/B]

OK! So what is the catch?

Would I trade this way? NO!

The reason the most recent results are not reliable (Single Order at R:R=3) is because the (Win Rate * R:R) is lower than 100%.

The reason that it still makes money is because of the compounding effect of using the 1% Risk (instead of fixed lot sizes) as well as positive slippage!

Also, we have to remember that it is after-all a back-test and we should be careful in allocating resources so that we don’t blow our account.

I guess you will just have to wait for the boring statistics in order to get realistic values that are safer to use.

[B]So, please be warned, don’t use the 0%; 0%; 100% just yet on any live account.[/B]

[FMIC]

OK! So here is a better summation!

In order to make profit and not just break-even, the product of WinRate and R:R must be above 50% and in the case of the Single Order at R:R=3 it is and it just below 100%, so it does make money irrespective of compounding or not.

However, I like to have a “safety net” for when things go wrong. Just as an engineer builds a bridge with a safety factor well above the maximum expected load, I like to have the WinRate * R:R above or equal to 100%.

This is my personal approuch to when considering possible strategies for trading, but all things considered, the 0%, 0%, 100% percentage allocation is still quite good. However, during the week-end I will post a better percentage allocation that will have a WinRate * RR that is above 100%.

Regards,
Carnino

[FMIC]

Please note that I have added some extra statistical values to the previous posts, namely Post #205 and Post #210.

[FMIC]

As promised here is the details for values with a Product of Win Rate and R:R greater than 100%:

EUR/USD (from 2012-12-31 to 2013-12-27) - PARMAR EUR-USD Strategy | Myfxbook
[ul]
[li]Spread: 1.0 pip
[/li][li]ZimeZone: UTC+2 (NY Close)
[/li][li]Trading Days: Tuesday-Friday
[/li][B][li]Breakout Offset: (spread) 1.0 pip only on High (0.0 on Low)
[/li][li]Pending Orders Expire 240hrs (10 Days)
[/li][li]Open Orders Closed at End-of-Day[/B]
[/li][li]Stop Size: 5.0 pips
[/li][li]Risk: 1% (Initial balance $10000)
[/li][B][li]Allocation: Single Order (R:R=5)[/B]
[/li][li]Profit: $24054.93 (240.55% Gain)
[/li][li]Profit Factor: 1.49
[/li][li]Relative Drawdown: 23.52%
[/li][li]Win Rate: 24.06% (320 Trades)
[/li][li]Reward/Risk Ratio: 492.44% (less than 500% due to close at day-end)
[/li][li]Avg. Profit/Avg. Loss Ratio: 470.49%[B][*]Win Rate * Reward/Risk Ratio: 118.49%
[/li][li]Win Rate * Avg. Profit/Avg. Loss Ratio: 113.21%[/B]
[/li][li]Kelly Factor: 7.92%[/ul]
[/li]GBP/USD (from 2012-12-31 to 2013-12-27) - PARMAR GBP-USD Strategy | Myfxbook
[ul]
[li]Spread: 1.5 pip
[/li][li]ZimeZone: UTC+2 (NY Close)
[/li][li]Trading Days: Tuesday-Friday
[/li][B][li]Breakout Offset: (spread) 1.5 pip only on High (0.0 on Low)
[/li][li]Pending Orders Expire 240hrs (10 Days)
[/li][li]Open Orders Closed at End-of-Day[/B]
[/li][li]Stop Size: 5.0 pips
[/li][li]Risk: 1% (Initial balance $10000)
[/li][B][li]Allocation: Single Order (R:R=7)[/B]
[/li][li]Profit: $52705.11 (527.05% Gain)
[/li][li]Profit Factor: 1.72
[/li][li]Relative Drawdown: 20.00%
[/li][li]Win Rate: 21.40% (299 Trades)
[/li][li]Reward/Risk Ratio: 680.91% (less than 700% due to close at day-end)
[/li][li]Avg. Profit/Avg. Loss Ratio: 632.67%
[/li][B][li]Win Rate * Reward/Risk Ratio: 145.75%
[/li][li]Win Rate * Avg. Profit/Avg. Loss Ratio: 135.42%[/B]
[/li][li]Kelly Factor: 8.98%[/ul]
[/li]Obviously, GBP/USD is much better suited for this strategy!

[FMIC]

Any determination as to which days of the week are most profitable? Which of the 3 TP levels is the most profitable (weighted per lot)?

For my very short exposure to this trading system, TP 15 is the most profitable. YMMV

See my latest posts that contain links to MyFXBook for the strategies with single orders of R:R=5 and R:R=7 for EUR/USD and GBP/USD respectively.

The MyFXBook stats show the profitability for the different days of the week and hours of the day, as well as many other stats.

[Bobkat]

See my latest posts that contain links to MyFXBook for the strategies with single orders of R:R=5 and R:R=7 for EUR/USD and GBP/USD respectively.

The MyFXBook stats show the profitability for the different days of the week and hours of the day, as well as many other stats.

Sorry Carnino, but I can’t seem to find the MyFXBook for your postings. Please provide your listing…thanks!

[vijaykumar]

Sorry Carnino, but I can’t seem to find the MyFXBook for your postings. Please provide your listing…thanks!

PARMAR GBP-USD Strategy | Myfxbook

[sheriffonline]

Hi Carnino,

Great Job.

Pending order expire in 240 hrs(10 days) or end of the day? bcoz previous post you tried pending orders closed at the end of the day.

As promised here is the details for values with a Product of Win Rate and R:R greater than 100%:

EUR/USD (from 2012-12-31 to 2013-12-27) - PARMAR EUR-USD Strategy | Myfxbook
[ul]
[li]Spread: 1.0 pip[/li][li]ZimeZone: UTC+2 (NY Close)[/li][li]Trading Days: Tuesday-Friday[/li][B][li]Breakout Offset: (spread) 1.0 pip only on High (0.0 on Low)[/li][li]Pending Orders Expire 240hrs (10 Days)[/li][li]Open Orders Closed at End-of-Day[/B][/li][li]Stop Size: 5.0 pips[/li][li]Risk: 1% (Initial balance $10000)[/li][B][li]Allocation: Single Order (R:R=5)[/B][/li][li]Profit: $24054.93 (240.55% Gain)[/li][li]Profit Factor: 1.49[/li][li]Relative Drawdown: 23.52%[/li][li]Win Rate: 24.06% (320 Trades)[/li][li]Reward/Risk Ratio: 492.44% (less than 500% due to close at day-end)[/li][li]Avg. Profit/Avg. Loss Ratio: 470.49%[B][*]Win Rate * Reward/Risk Ratio: 118.49% [/li][li]Win Rate * Avg. Profit/Avg. Loss Ratio: 113.21%[/B][/li][li]Kelly Factor: 7.92%[/ul][/li]GBP/USD (from 2012-12-31 to 2013-12-27) - PARMAR GBP-USD Strategy | Myfxbook
[ul]
[li]Spread: 1.5 pip[/li][li]ZimeZone: UTC+2 (NY Close)[/li][li]Trading Days: Tuesday-Friday[/li][B][li]Breakout Offset: (spread) 1.5 pip only on High (0.0 on Low)[/li][li]Pending Orders Expire 240hrs (10 Days)[/li][li]Open Orders Closed at End-of-Day[/B][/li][li]Stop Size: 5.0 pips[/li][li]Risk: 1% (Initial balance $10000)[/li][B][li]Allocation: Single Order (R:R=7)[/B][/li][li]Profit: $52705.11 (527.05% Gain)[/li][li]Profit Factor: 1.72[/li][li]Relative Drawdown: 20.00%[/li][li]Win Rate: 21.40% (299 Trades)[/li][li]Reward/Risk Ratio: 680.91% (less than 700% due to close at day-end)[/li][li]Avg. Profit/Avg. Loss Ratio: 632.67%[/li][B][li]Win Rate * Reward/Risk Ratio: 145.75% [/li][li]Win Rate * Avg. Profit/Avg. Loss Ratio: 135.42%[/B][/li][li]Kelly Factor: 8.98%[/ul][/li]Obviously, GBP/USD is much better suited for this strategy!

[defiance888]

Great Job carnio,

but im wondering. shouldn’t modeling quality be at 99% instead of 90%

[FMIC]

Sorry Carnino, but I can’t seem to find the MyFXBook for your postings. Please provide your listing…thanks!

Read Post #214