|
PureBytes Links
Trading Reference Links
|
I'd say no...see below...works out to 1.38% per week in real time. Whether
you like Aberration or not, the real time results speak volumes on how Mark
Johnson (janitor) ran the system.
Can't say how Mark did in 2000 and 2001...Mark?
MT
At 03:59 PM 12/6/01 +0100, Noel Falzon wrote:
>Is having a system that consistanly makes 1% a week
>asking too much ?
To: omega-list@xxxxxxxxxx
From: Mark Johnson <janitor@xxxxxxxxxxxx>
Subject: Vendor System, Real Profits for 3 Years
Various list readers have been requesting that I post this
article for quite a while. I wrote it and it was published
in a newsletter called _Club_3000_News_ back at the
beginning of the year. Six months have passed since it was
originally printed and sent out to Club members, and so
I've recently sought and gotten the Editor's permission
to reprint it here. The Club's URL is http://www.club3000.org
and subscriptions to their newsletter are very reasonably
priced; click over to their website and check it out. -mj
____________________________________________________________________
(Reprinted, with permission, from Club 3000 News
issue # 2000.01, dated February 18, 2000)
>
> Vendor system, real profits for three years -Mark Johnson
> -----------------------------------------------------------
>
> I bought a futures trading system from a vendor who
> advertised it here in Club3000 News, and I've been trading
> it in my real-life, real-money brokerage account. For the
> period 12/31/96 to 12/31/99 (i.e. for the three calendar
> years 1997, 1998, 1999), the system has produced net profits
> of $546,762 in my account, including commissions, slippage,
> vacations, rollovers, and human mistakes. Starting from my
> initial capital, that's a compound annual growth rate of
> 71.6 percent per year, for three years. Year by year
> results were as follows:
>
> Net Return Worst
> Year Profits on capital Drawdown
> --------------------------------------------
> 1997 $118,443 87.8 % 42 %
> 1998 $102,619 41.5 % 33 %
> 1999 $325,700 91.5 % 39 %
>
> (I've sent copies of my brokerage account statements to the
> Editor of Club 3000 News, to authenticate and validate these
> results.)
>
>
> I want to encourage other Club3000 members that it truly is
> possible to make money trading 100% mechanical systems
> purchased from vendors. I ITALICS{know} it's possible,
> because I've done it myself, three years in a row. My aim in
> writing this article to isn't to brag about myself, or to
> boost sales of some vendor's product. Instead, I'm reporting
> my profitable results to show that not only ITALICS{can} it
> be done, it ITALICS{has} been done, in real life, by an
> ordinary citizen. A regular guy who works 40+ hours a week
> in a normal job unrelated to investing or futures or
> finance, and who trades commodities on the side, in the
> evening.
>
> Plenty of people will scream at you that this is
> ITALICS{obviously impossible.} It even appears in the
> advertisements right here in the Club3000 newsletter: "Don't
> become another losing systems junkie" says the ad, and the
> WWW site it promotes contains a Trader Education tutorial
> which claims "... systems sold have never actually
> performed in real-time trading ..." Well, I present my own
> real-money, real-time trading results (above) as a
> counterexample.
>
> When I began to investigate futures as an investment/
> speculation vehicle, it took me quite a long time to find a
> trading approach that fit my personality. Eventually I
> discovered what was right for me:
>
> * I need to use a 100% mechanical trading system, to
> eliminate my own (very poor!) subjective judgement
> from trading.
>
> * I need to use the exact same system and the exact same
> system parameter values on ALL markets that I trade.
> Otherwise, I become nervous that I may have "overfitted"
> the system to past history.
>
> * Long-term trendfollowing suits my personality. It produces
> large profits (and large losses!) per trade, making
> commissions and slippage less of a life-or-death issue.
>
> * I need to research the system ITALICS{myself}, in
> computerized historical backtesting. I found that I am
> unable to "pull the trigger" unless I've run a huge battery
> of historical backtests myself, and spent hours pondering
> the results.
>
> * I need to trade the mechanical system's entry and exit
> signals, with a betsize selection algorithm (some books
> call this "money management") that trades more and more
> contracts as my account grows larger and larger.
>
> * I decided that I probably wouldn't be able to emotionally
> survive an equity drawdown bigger than 40 percent; I'd
> probably panic and stop trading. So I chose a betsize
> selection algorithm that produced less than 40% drawdown in
> historical backtests, with the mechanical system.
>
>
> Remember, this is just me. Each trader is different, and
> what's right for me might be exactly wrong for you. But I
> found that when I adjusted my trading style to align with
> ITALICS{my} personality, I started achieving good results.
>
> I purchased the mechanical system from the vendor, and spent
> a year backtesting it on a variety of markets and using a
> variety of betsize selection algorithms. I also telephoned
> the vendor to ask to see brokerage statements from trading
> the system with real money. He faxed them to me that same
> evening. Regrettably, BOLD_ITALICS{very few} system vendors
> give out account statements, which might make you wonder
> whether they're hiding something.
>
> Since that time, the vendor has gotten a Web Site (who
> hasn't?) and has scanned in his brokerage statements so that
> visitors can study them for themselves. The web address is
> http://www.trade-system.com .
> I hope other system vendors will adopt this practice soon.
>
> After doing all this backtesting and studying, I decided to
> trade the system on a portfolio of 25 futures markets,
> using the exact same system and the exact same parameter
> values (namely, "80" and "2") on all markets. I decided to
> trade the currencies, the energies, the softs, and the
> bonds/notes/interest-rates.
>
> I also decided to use a betsize selection algorithm that
> risked a percentage of my total account equity on every
> trade. But this percentage is ITALICS{not fixed} -- it
> varies as the account equity fluctuates. The risk-versus
> -equity curve for my betsize approach is shown in Figure 1,
> and the math formula behind the curve was presented in
> Club3000 News #99.04.
>
> note: The curve of Figure 1 was generated in Microsoft
> Excel, and may be viewed or downloaded from the web, at
> http://www.mjohnson.com/plots/fig1_2k01.xls
>
> Previous articles have detailed my 1997 trading results (see
> Club3000 News #98.02) and my 1998 results (#99.02). In
> calendar 1999, my account equity peaked at $788,750 on
> Sept. 23. The lowest subsequent valley was on Nov. 5, at
> $481,782. Thus my worst case drawdown in 1999 was 38.9
> percent {math: (789-482)/789 = 0.389 } . I started the
> year at $355,903 and ended it at $681,603 so my account
> grew 91.5 percent in calendar 1999.
> {math: (682-356)/356 = 1.915 }. Here is a table of my
> account equity values for the three year period:
>
> ..................................................
> 12/31/96 $149,841 06/30/97 $194,022
> 01/15/97 (withdrew $15,000) 07/31/97 $279,860
> 01/31/97 $181,740 08/31/97 $223,839
> 02/28/97 $237,543 09/30/97 $238,358
> 03/31/97 $319,647 10/31/97 $268,751
> 04/30/97 $293,402 11/30/97 $221,343
> 05/31/97 $241,270 12/31/97 $253,284
> ..................................................
> 12/31/97 $253,284
> 01/30/98 $306,979 07/31/98 $206,376
> 02/27/98 $283,581 08/31/98 $328,982
> 03/31/98 $289,500 09/30/98 $443,173
> 04/30/98 $248,353 10/31/98 $390,351
> 05/29/98 $214,811 11/30/98 $336,843
> 06/30/98 $223,702 12/31/98 $355,903
> ..................................................
> 12/31/98 $355,903
> 01/29/99 $331,508 07/30/99 $646,242
> 02/26/99 $459,914 08/31/99 $717,237
> 03/31/99 $510,545 09/30/99 $735,686
> 04/30/99 $583,033 10/29/99 $537,519
> 05/28/99 $588,308 11/30/99 $634,698
> 06/30/99 $675,430 12/31/99 $681,603
> ..................................................
>
>
> Since the system follows long-term trends, it holds winning
> trades a relatively long time; 6 months is not uncommon.
> So I do a lot of rollover trades in addition to system-entry
> trades and system-exit trades. These increase the total
> commissions and slippage I pay.
>
> For example, the system signalled to go long Crude Oil on
> 11 Mar 99, and I bought 8 contracts of the June99 crude.
> I rolled this position over to the August crude on 11 May,
> using a spread order ("Buy 8 August crude and sell 8 June
> crude, as a spread trade, at market"). The system remained
> long so I rolled the position again on 12 July, moving from
> August to October crude. The system remained long so I
> rolled again on 13 September, moving from October to
> December crude. Finally the system signalled to exit on
> 11 Oct 99, and I sold my 8-contract position. Even though
> this was theoretically a single trade of 8 contracts, in
> real life I executed 4 round-trips of 8 contracts each, so
> I paid 32 commissions at about $25 each. Total slippage
> (actual vs. system theoretical) was $1840 on this 8-lot
> position. In theory the system made $58,560 on its
> 8-contract position. In reality, I made $55,920. And
> $2640 was lost to commissions and slippage.
>
> I'll call the four individual pieces of that one Crude Oil
> trade "legs": there was a leg when trading CL99M, another
> in CL99Q, another in CL99V, and a fourth leg in CL99Z. The
> account book in which I keep my trading results tabulates
> "legs" of trades (since they are individual round-trips).
>
> Looking over my records for the years 1997, 1998, and 1999,
> the mechanical system generated a total of 507 individual
> trade "legs"; round-trips that represent either a full trade
> or a portion of a trade that lasted so long it had to be
> rolled over. I would suggest that 507 legs is a
> statistically significant number, for a system that uses the
> exact same logic and the exact same parameters on all trades
> in all markets. I would also suggest that the system's
> favorable trading results for three years in a row, across
> 507 legs, are not merely due to chance alone.
>
> In summary: I figured out a trading approach that suited my
> personality, purchased a 100% mechanical trading system from
> a vendor, carefully ran it through lots of backtests, and
> traded it for three years. The system and I were profitable
> in all three years, yielding a compound annual growth rate
> of 71.6 percent per year (net profits of $546,762). It's
> not impossible: ITALICS{I did it}. Others can, too.
>
> Editor's note: documentation provided.
>
--
Mark Johnson Silicon Valley, California mark@xxxxxxxxxxxx
"... The world will little note, nor long remember, what we
say here..." -Abraham Lincoln, "The Gettysburg Address"
|