$15,000 to Over $750,000 in 7 yrs Impossible? Think again!

[barney]

Cathers.

An important question which I feel (I could be wrong) many would like answered in the way of guidence in their pursuit of developing a trading methodology.
I'm thinking of putting together a thread in answer to your question basically walking you through the steps of systems/method developement.
Not actually designing one although we could do that but more pointing in the right direction.
I expect there would also be some good questions.

If people would like a seperate thread let me know and I'll get started.

Hi lads, Enjoying my "education" on this forum. Just thought I'd mention an old thread which I found last night (actually it was very very early this morning!!.........Musos hours unfortunately :guitar: )
I have reposted it in the General Chat section under Trading Plans....
Tech, maybe you could take the thread on from where it is up to..........It is a great start for "learners" like me.......and you could get into a bit heavier stuff for you experienced lads............I still like reading the "heavier" technical stuff even though a lot of it goes over my head.......(maybe I should stop ducking all the time :hide: Cheers.

 

[Staybaker]

Staybaker,
Tech is using Leverage.

Yes, I know. My question is about the part before he introduces leverage.

Cheers, Staybaker.

 

[tech/a]

Sorry Stay just saw your post.

You are looking at a linier return.

In actual practice a tested method will return an average over the period tested. This will include losses,and outlier profits some of 100's of %.

Trades can last days or years so the reinvestment is anything but linier.

I can and will post the yearly and Monthly returns if this makes it easier.

If not then I will go deeper into explaination.

 

[Sir Burr]

Re: $15,000 to Over $750,000 in 7 yrs Impossible? Think again!

Tech,

Sorry if this has been repeated, haven't read all the posts but 3 points on the back-testing:

• No consideration of tax. Software doesn't have any options to add this. No sure how to consider tax besides wiping off a percentage (suggestions?).

• Extraordinary market conditions over the testing period. Try testing with data in other market conditions.

• Survivorship bias. Instead of using the current BT list, try turnover to pick stocks.

I think that the above 3 points may considerably alter the outcome?

SB

 

[tech/a]

Re: $15,000 to Over $750,000 in 7 yrs Impossible? Think again!

Tech,

Sorry if this has been repeated, haven't read all the posts but 3 points on the back-testing:

• No consideration of tax. Software doesn't have any options to add this. No sure how to consider tax besides wiping off a percentage (suggestions?).

• Extraordinary market conditions over the testing period. Try testing with data in other market conditions.

• Survivorship bias. Instead of using the current BT list, try turnover to pick stocks.

I think that the above 3 points may considerably alter the outcome?

SB

(1) Tax is a consideration regardless of how you trade. It would be rediculous to suggest that returns be quoted Nett Free of tax on a forum such as this.
But as I stated in a post further up once you use reasonable sums( $250,000) the returns become such that you'd happily pay your tax and hop on the next plane to the bahamas,infact youd probably live in Menorca and not have to worry about tax issues.

(2) The testing period is over the last 8 yrs and is a long system,its designed to return best profit during bull markets. The arguement that a method should be designed to profit equally well in ALL market conditions again is rediculous if trading a portfolio of stocks. However if trading Indexes or Futures which are singularities,then best results would be found from trading both bullish and bearish market conditions.

(3) Survivorship is always an issue when testing. However if a sample is big enough I find that the results are telling us that regardless of what stocks you choose,if you trade in the manner formulated in the tested methodology you are going to get a result within the parameters of the Highest and Lowest deviation of the average.This has certainly been the case with the systems I trade and of the results I am privvy to of traders trading systems,and with knowledge of their systems blueprints.

Im happy to trial based upon Turnover if you would define for me how you define "Turnover" and how to determine it. I have a liquidity filter within the system shown based upon minimum turnover over a month of trading ($500,000).

The exercise is meant to get people thinking that high return isnt necessarily linked to high returns from your trading---more from consistant returns, Compounding and sensible use of leverage.

 

[Sir Burr]

The exercise is meant to get people thinking that high return isnt necessarily linked to high returns from your trading---more from consistant returns, Compounding and sensible use of leverage.

OK definately but:

1) Yes it is ridiculous, $15,000 to Over $750,000 in 7 yrs Impossible? Think again and lower your expectation to 30-50% of that.

2) "The arguement that a method should be designed to profit equally well in ALL market conditions again is rediculous if trading a portfolio of stocks." Yes exactly my point, expect a return of over $750,000 in 7 yrs IF we have similar market conditions.

3) Firstly include de-listed shares in your database then create a ticker with the total market turnover in it. Add a moving average to it and use a % of the turnover to find the approx ASX300. Using 500000 will exclude many stocks in the past compared with today (bias).

SB

 

[Sir Burr]

Tech,

I just noticed there doesn't seem to be any Monte Carlo profit (or drawdown) stats for the $15,000 to Over $750,000 in 7 yrs?

Profit Stats
Maximum Profit: $????????
Average Profit: $750,000 (750.00%)
Minimum Profit: $????????
Standard Deviation: $??????
Probability of Profit: ???%
Probability of Loss: ???%

What's the Min/Max profit and deviation?

Also, in regards to your comments about tax and the title of the thread. Yes is impossible to accumulate the amount in question as there will not be the same compounding effect with tax being paid out of your account.

SB

 

[tech/a]

God why bother.

I have purposely chosen an example where the system return is 15%.
If I alter even that variable to 40 or 50% which is possible in a longer term trading methodology the 15-750K scenario looks more like 15-3 million which in turn less everything you have suggested Im sure will nett the $750K I cant be bothered running it. Some trades would only be subject to 25% tax as they would be held over 365 days infact pretty well all the big winners would fit here.

Actually to prove a point I altered 2 variables SLIGHTLY and it doubles return.To $1,350,000 from the original $15,000
I altered position sizing from 12.5% of capital to 13.5% and I altered Leverage from 2.5:1 to 3:1 That should take care of your tax and other issues-- So yes it can be done.

Table 1

But if it makes you happy I see your point.

Hopefully those who may benifit have understood and digested mine.
Nothing is impossible if you think outside of the square.
Little is possible if you dont. Most get caught in the semantics and lose sight of the message.

Table 2 is the Montecarlo you requested of the ORIGINAL 15-750K over 20000 portfolios.

Hmmm seems that there is a computer problem I cannot write piks to the forum.
I think its a problem my end.
I can email the files to you if you wish just private mail me your email address and i will post off tommorow.

Sorry about that.

 

[MichaelD]

Tech,

There have been some valid objections to your projected returns here which really need to be pointed out without being steamrollered over.

Problem 1. Tax is a major issue which you have simply waved away. The effect of accounting for tax is a major reduction in the power of compounding - taking 30% of the compounding amount out per annum to pay the necessary tax will halve the projected return.

Problem 2. This is a MUCH bigger problem and no one has mentioned it so far in this thread. Inclusion bias. (Not survivor bias which has a negligible effect.) If you examine the universe of returns I'll guarantee you that you'll have several trades such as Paladin (PDN) in there from 0.02 to 5.00 (or the rough equivalent). Such a trade would NEVER be taken in real life because PDN would not have been in the trading universe to begin with. Thus, the extreme outlier trades which make the numbers look so fantastic should not be counted in the outcome. In my opinion, this is probably THE biggest gotcha of backtesting, and almost nobody seems to consider it.


I'm NOT saying it isn't profitable. I'm saying it's not as profitable as it first would appear.

 

[tech/a]

Michael.

The tax issue I will apologise for not heading the thread $750,000 before tax.
However As pointed out above with small changes same method returns $1,350,000 before tax,so I'm sure it would return $750,000 after tax.

Inclusion basis.
I certaintly see your point,however in this case this is not the case.
The method only returns 15% net and relies as much on win rate as it does on Profit/trade.
The largest winners in the trade log are HSP<TOL<CTX 344/306/279% return respectively all held around a year and a half.

I'm happy for you to veiw the trade log. Ill email it if you wish.

Ive often pondered the very point you make. In conclusion this will occur but the whole premise of a good method is that with ANY universe of stocks the results will be within the parameter of the Blueprint of the methodology tested.Today the BT list will not be as it is in 2010 some will be added and some deleted.
As is the case in testing regardless of the combination of trades in the portfolio the results remain in the Blueprint. This is the case when I test universes from other bourses which have similar characteristics to a list like BT.

So to make everyone happy I will submit the second group of settings as the example so that the results are more likely.

Detailed Report
( ASX100 XX1)

Simulation Summary
Simulation Date: 12/09/2006
Simulation Time: 9:31:10 PM
Simulation Duration: 0.35 seconds

Trade Summary
Earliest Entry Date in the Trade Database: 8/01/1999
Latest Entry Date in the Trade Database: 25/08/2006
Earliest Exit Date in the Trade Database: 15/01/1999
Latest Exit Date in the Trade Database: 1/09/2006

Start Trade Entry Date: 8/01/1999
Stop Trade Entry Date: 25/08/2006
First Entry Date: 8/01/1999
Last Entry Date: 25/08/2006
First Exit Date: 15/01/1999
Last Exit Date: 1/09/2006

Total Trading duration: 2793 days

Profit Summary
Profit Status: PROFITABLE
Starting Capital: $15,000.00
Finishing Capital: $1,372,880.02
Maximum Equity/(Date): $1,357,880.02 (1/09/2006)
Minimum Equity/(Date): -$2,731.38 (24/09/1999)
Gross Trade Profit: $1,697,759.39 (11318.40%)
Gross Trade Loss: -$339,879.37 (-2265.86%)
Total Net Profit: $1,357,880.02 (9052.53%)
Average Profit per Trade: $11,807.65
Profit Factor: 4.9952
Profit Index: 79.98%
Total Transaction Cost: $13,800.00
Total Slippage: $0.00
Daily Compound Interest Rate: 0.1618%
Annualized Compound Interest Rate: 80.4437%

Trade Statistics
Trades Processed: 356
Trades Taken: 115
Partial Trades Taken: 0
Trades Rejected: 78
Winning Trades: 56 (48.70%)
Losing Trades: 59 (51.30%)
Breakeven Trades: 0 (0.00%)

Normal Exit Trades: 109 (94.78%)
Delayed Normal Exit Trades: 0 (0.00%)
Open Trades: 6 (5.22%)
Protective Stop Exit Trades: 0 (0.00%)
Time Stop Exit Trades: 0 (0.00%)
Profit Stop Exit Trades: 0 (0.00%)

Largest Winning Trade/(Date): $252,480.00 (1/09/2006)
Largest Losing Trade/(Date): -$40,394.08 (3/02/2006)
Average Winning Trade: $30,317.13
Average Losing Trade: -$5,760.67
Average Win/Average Loss: 5.2628

Trade Duration Statistics
(All Trades)
Maximum Trade Duration: 749 (days)
Minimum Trade Duration: 6 (days)
Average Trade Duration: 134 (days)
(Winning Trades)
Maximum Trade Duration: 749 (days)
Minimum Trade Duration: 6 (days)
Average Trade Duration: 231 (days)
(Losing Trades)
Maximum Trade Duration: 161 (days)
Minimum Trade Duration: 6 (days)
Average Trade Duration: 42 (days)

Consecutive Trade Statistics
Maximum consecutive winning trades: 8
Maximum consecutive losing trades: 8
Average consecutive winning trades: 2.55
Average consecutive losing trades: 2.81

Relative Drawdown
Maximum Dollar Drawdown/(Date): $61,087.98 (28/10/2005)
Maximum Percentage Drawdown/(Date): 31.3900% (16/03/2001)

Absolute (Peak-to-Valley) Dollar Drawdown
Maximum Dollar Drawdown: $67,020.99 (7.1620%)
Capital Peak/(Date): $935,835.97 (16/06/2006)
Capital Valley/(Date): $868,814.99 (1/09/2006)

Absolute (Peak-to-Valley) Percent Drawdown
Maximum Percentage Drawdown: 39.7300% ($14,753.85)
Capital Peak/(Date): $37,135.38 (24/03/2000)
Capital Valley/(Date): $22,381.52 (16/03/2001)


Semantics aside,I hope the example of the use of Compounding/Leverage and how the smallest changes in variables NOT WITHIN a methods structure can alter the outcome considerably---has been helpful.

 

[nizar]

tech

I find it odd that average losing trade time is (substantially) longer than average winning trade time. if a trade goes against u, wouldnt u pretty much get stopped out straight away?

(while when it goes up, u would just ride the trend?)

 

[Sir Burr]

Hi Tech,

Personally I bothered because I just had some points to make. I believe I put them across with no malice intended and in a non-derogatory manner.

The above figures you posted are from the Detailed Report not the Monte Carlo report. Monte Carlo indicates the range of outcomes of all statistics rather then a possible best case put forward in a single simulation run.

OK say we exclude tax, compounding, curve fitting, outliers and biases etc and accept that the profit is $750000 over the past 7 years. My main point of reply was that the period of testing is majority a bull market and not to expect the same returns going forward. Just one out of the 7 years testing period was down, 2002.

I understand your points about money management and they are excellent but it would be interesting to see results in a bear market on margin. If you were using a method which keeps you out of a bear market, then again these high profits would not be seen as I believe the system tested is long only.

Try reversing the system making it a “short” and see what the returns would have been over the same 7 year period. An exception with this is the maximum profit on a trade would be 100% only.

About the tax, correction: 25-50% tax (+ Medicare).

SB

Edit: Tech/A, being traders it is our job to try and rip apart any theories to try and find problems with trading systems (isn't it?). :

 

[tech/a]

Trade Duration Statistics
(All Trades)
Maximum Trade Duration: 749 (days)
Minimum Trade Duration: 6 (days)
Average Trade Duration: 134 (days)
(Winning Trades)
Maximum Trade Duration: 749 (days)
Minimum Trade Duration: 6 (days)
Average Trade Duration: 231 (days)
(Losing Trades)
Maximum Trade Duration: 161 (days)
Minimum Trade Duration: 6 (days)
Average Trade Duration: 42 (days)

Sorry Nizar not with you??

Average winning trade time 231 days
Average losing trade time 42 days

 

[tech/a]

Sir B

I'll answer your post in greater detail later.

I am very aware that the Report posted wasnt as asked.

Here is the Monte Carlo report of the original. 15-750k

Monte Carlo Report
( ASX100 XX1)

Simulation Summary
Simulation Date: 13/09/2006
Simulation Time: 7:57:10 AM
Simulation Duration: 35.47 seconds

Trade Parameters
Initial Capital: $15,000.00
Portfolio Limit: 100.00%
Maximum number of open positions: 100
Position Size Model: Equal Percent Units
Trade Size (% of total cap): 13.00%
Pyramid profits: Yes
Transaction cost (Trade Entry): $60.00
Transaction cost (Trade Exit): $60.00
Margin Requirement: from Database

Trade Preferences
Trading Instrument: Stocks
Break Even Trades: Process separately
Trade Position Type: Process long trades only
Entry Order Type: Default Order
Exit Order Type: Default Order
Minimum Trade Size: $0.00
Accept Partial Trades: No
Volume Filter: Ignore Volume Information
Pyramid Trades: No
Use Level Zero trades only: Yes

Simulation Stats
Number of trade simulations: 20000
Trades processed per simulation: 356
Maximum Number of Trades Executed: 128
Average Number of Trades Executed: 123
Minimum Number of Trades Executed: 118
Standard Deviation: 1.45

Profit Stats
Maximum Profit: $926,828.39 (6178.86%)
Average Profit: $785,185.41 (5234.57%)
Minimum Profit: $587,810.38 (3918.74%)
Standard Deviation: $63,795.80 (425.31%)
Probability of Profit: 100.00%
Probability of Loss: 0.00%

Percent Winning Trade Stats
Maximum percentage of winning trades: 50.83%
Average percentage of winning trades: 48.27%
Minimum percentage of winning trades: 45.97%
Standard Deviation: 0.81%

Percent Losing Trade Stats
Maximum percentage of losing trades: 54.03%
Average percentage of losing Trades: 51.73%
Minimum percentage of losing trades: 49.17%
Standard Deviation: 0.81%

Average Relative Dollar Drawdown Stats
Maximum of the Average Relative Dollar Drawdown: $4,858.44
Average of the Average Relative Dollar Drawdown: $3,857.02
Minimum of the Average Relative Dollar Drawdown: $3,214.72
Standard Deviation: $340.92

Average Relative Percent Drawdown Stats
Maximum of the Average Relative Percent Drawdown: 2.8634%
Average of the Average Relative Percent Drawdown: 2.5953%
Minimum of the Average Relative Percent Drawdown: 2.3725%
Standard Deviation: 0.0960%

Maximum Peak-to-Valley Dollar Drawdown Stats
Maximum Absolute Dollar Drawdown: $70,168.24
Average Absolute Dollar Drawdown: $41,421.12
Minimum Absolute Dollar Drawdown: $27,015.00
Standard Deviation: $8,939.94

Maximum Peak-to-Valley Percent Drawdown Stats
Maximum Absolute Percent Drawdown: 30.7737%
Average Absolute Percent Drawdown: 30.7217%
Minimum Absolute Percent Drawdown: 30.6682%
Standard Deviation: 0.0518%

 

[dj_420]

tech/a

just wondering what do i have to do to use a system like this in my current trading plan. what software should i buy that will allow me to perform the calculations you have made.

if i have a possibility of making some excellent returns using a tech a system then i want to try it.

what do you think is better tech/a, fundamental long term trades or using a syetm with many short term tech analysis trades

 

[stock_man]

Congrats on a great thread topic Tech.

I have just read over the whole thread, and there are some great ideas and comments.

However, it seems the thread is full of negativity. From my understanding the point of this thread was to open people's minds up to the possibility that this kind of capital growth is viable.

People need to understand that by being so cynical and critical of ideas, only caps their thoughts. Its only when we expand our horizons and thought patterns that the unimaginable becomes a reality.

You must remember that once the earth was thought to be flat - people ridiculed this idea and called them insane. Or maybe when the idea of internation flights was dreamed upon.

Thanks again Tech, your thread has only opened up my thought process to think of other possibilities of creating $1M.

 

[MichaelD]

Inclusion basis.
I certaintly see your point,however in this case this is not the case.
The largest winners in the trade log are HSP<TOL<CTX 344/306/279% return respectively all held around a year and a half.

Fantastic. That's really good to hear and certainly improves my estimation of the robustness of the system.

I'll raise two more objections, though;

1. The time frame traded. I for one would like to see this same system traded from 1-Jan-1988 to 31-Dec-1992 with all other parameters as initially set. This will give a better indication of system survivability during lean times.

2. I'd also like to see the closed equity curves for both time frames to get a feel for how the system trades/how long it is in drawdown in good and bad times.

 

[tech/a]

Cathers.

The analysis itself isnt important.
I use T/A as I can write formulas which can then be tested. The key I believe in introducing Compounding and Sensible leverage is the ability to have the parameters known that testing will provide.

Stockman.

Its a common Aussi trait. Unfortunately it blinds many to the possiblilites of thinking outside the shoe box. Most live in the shoe box.

Michael.

its NOT ABOUT THE SYSTEM PRESENTED,we could dissect it endlessly.
I will run some tests and sure I agree in less than ideal times the results will alter.
When I Started techtrader years ago I had no idea what the future bought,I have no idea what the next 4 yrs will bring either.

What I do know is that whatever I trade I will continue to trade provided it falls within the "Blueprint".
I am at the moment looking into "Switches" as I like many others believe there are times to be fully exposed and times to have no exposure at all.
Not having a position wether it be trading or property is in fact HAVING A POSITION.

Money---Big money comes not from being right constantly but from the WAY YOU USE your money. I have proven this time and again.

 

[jwatkins]

T/A... I completely agree that the system is quite irrelevant, but I'd be terrified in using a 70-75% LVR as you suggest!

Does there even exist a system that would not generate a margin call repeatedly during a bear market (without the benefit of knowing in hindsight that you were in a bear market)? I know you argue that you'd use a different system to T/A during a bear market, but
a.) what sort of ideas would the system use
b.) would it impale you if it turned out you weren't in a bear market afterall
c.) could it survive a major shock?

I'd take 15%... heh I think I'd take 10% p.a. over the next 7 years. Of course, I'm very open to proof that 10% returns are pathetic

 

[Sir Burr]

Does living in a shoe box mean I am closed to ideas? Maybe this is some gang of weirdo's who think that backtesting statistics over a bull market should never be questioned?

Granted, a virgin starts out with Techtrader a few years ago with market conditions as they have been, has pocketed a motza like everyone else. A virgin starts out with Techtrader in '87 and for 7 years what have they achieved, $750,000, $200,000, -$15,000?

My problem is not that $750,000 is impossible (in 7 years), simply that it is dependent on market conditions (using "long" only).

You have made it look slick as and this is not a tricky question but again, how would trading on margin look in a market, say '87 thru '93?