|
PureBytes Links
Trading Reference Links
|
List
Here is some interesting discussion forwarded from the Realtraders list.
Kinda long and sometimes repetitive but instructive.
John Manasco
----- Original Message -----
From: Gary Fritz <fritz@xxxxxxxx>
To: <john@xxxxxxxxxxx>
Sent: Sunday, July 09, 2000 11:25 AM
Subject: Re: [RT] optimal f
> > Recently there was some discussion of the merits and drawbacks of
> > using optimal f for position sizing. If someone has saved those
> > emails could you please forward the pertinent ones to me? I didn't
> > save them and now want to reread them.
>
> Here's what I saved.
> Gary
>
>
>
----------------------------------------------------------------------------
----
From: Kevin243@xxxxxxx
Date: Tue, 29 Feb 2000 21:25:38 EST
Subject: [RT] =?UTF-8?Q?Re:=2020Fixed=20Ratio=20or=20Fixed=20Fractional=3F?=
=?UTF-8?Q?=3F=20=20=E2=80=93=20Pros=20and=20Cons=20of=20=20Each=20...?=
To: <realtraders@xxxxxxxxxxxxxxx>
I would size the first trade as a fraction of the optimal fraction or
"Optimal F". (Recommend you read Ralph Vince's books). Another possibility
is to set up an account hedged with cash, and trading the risk portion of
the
portfolio at the Optimal F. To arbitrarily select a fraction of capital to
risk, is not a good way to trade. If that fraction is beyond the optimal f
for the market/trading system, then you will go broke.
Too far the other way, and you are missing the potential reward for the risk
you are taking.
I personally use fixed ratio, but I'm not about to say that it is superior
to
optimal fixed fraction (optimal f). Mathematically, it is not and Ryan
states this. Your statement that the fraction ought to be the same % for
each trade is correct when the trade outcomes are random ie. not correlated.
If they are correlated, then varying trade size (fraction of account equity)
could be the best way to go. That would depend on whether you can devise a
trading system to effectively benefit from the correlated results.
I have personally settled on fixed ratio for more practical reasons.
Optimal
F calc's are tedious where fixed ratio calc's are simple. I have set up
simulations (fixed ratio vs. optimal f) in Excel based on my actual trading
statistics ie. (ave % gain, ave% loss,max % loss, % drawdown, # of wins, #of
losses, etc.). Optimal F money management easily beats fixed ratio. But
not
necessarily in the beginning. Like Ryan states, there may be an advantage
to
using fixed ratio trading initially and then progressing to optimal f. My
simulations have shown this to be likely, but not necessarily so. (ie. more
often than not).
However, Delta in fixed ratio trading can be set to extremely aggressive
trading.
I'm trading one portfolio very aggressively and another portfolio not so
aggressively. Surprise, surprise, the aggressive portfolio has more
variability and more return than the less aggressive portfolio. I haven't
run any Sharpe ratio stats to see if the added return is worth the risk I'm
taking. I suspect that it is.
There are other practical matters to consider, like can you sleep at night.
The money management strategy you use has to consider your ability to deal
with the math. If optimal f is too complicated for you, then it is not an
option.
More importantly, I would guess that most traders have no concept of money
management (ie. a positive expecatation and how to leverage it), and would
benefit greatly from having that knowledge.
Kevin Campbell
In a message dated 2/29/00 1:34:24 PM Central Standard Time,
andrew@xxxxxxxxx
writes:
> (I apologize as this is somewhat long ...)
>
> I would like to start a discussion on the merits and weaknesses of the
> Fixed Ratio position sizing method popularized by Ryan Jones.
> Specifically how does it stack up against the other predominant position
> sizing method known as Fixed Fractional.
>
> The majority of the research I have conducted was based upon the fixed
> fractional method, or sizing your position based upon the risk the
> position represents and the percent of your total equity you wish to
> risk.
>
> For Example:
>
> You have a $100,000 account balance and wish to risk 10% of your account
> on each trade, or $10,000.
>
> You have an S&P Trading System which buys on a breakout of the highest
> high of the last 5 bars and sells on a breakout of the lowest low of the
> last 5 bars. You would reverse your position with each occurrence. (I
> am not advocating this system, nor have I tested this, I am only using
> it as an example).
>
> Your first trade has a risk of $2,000 so you would trade 5 contracts
> (10,000 / 2,000 = 5).
>
> Your first trade is a $750 winner so your account balance is now
> $103,750.
>
> Your second trade has a risk of $4,250 (the market has a greater 5 day
> range) and you would trade 2 contracts (103,750 / 4,250 = 2.44 ==> 2).
>
> You would continue this as long as you traded.
>
> The above makes sense logically to me. With the Fixed Ratio Method, I
> see some Problems:
>
> 1. How many contracts do you trade on your first trade? The method
> explains how to change your size once you have begun trading, but I am
> unclear on how to size your very first trade.
>
> 2. All trades are sized the same. Why would you want to risk varying
> amounts of capital? In the above example if you were trading a fixed 4
> contracts per trade, you would assume risk of $8,000 on the first trade
> and a risk of $17,000 on the second. I do not understand why different
> trades should represent differing amounts of risk for your account if
> the outcome of each trade is unknown and as the same probability of
> success or failure.
>
> I hope that this is a good start for an educational discussion amongst
> ourselves regarding what many consider the most crucial aspect of
> successful trading. I am looking forward to following the ensuing
> dialogue.
>
> Best regards,
>
> Andrew Peskin
>
-- End --
From: "Glen Wallace" <gcwallace@xxxxxxxx>
To: <realtraders@xxxxxxxxxxxxxxx>
Subject: [RT]
=?Windows-1252?Q?Re:_=5BRT=5D_Fixed_Ratio_or_Fixed_Fractional=3F=3F?=
Date: Wed, 1 Mar 2000 13:05:50 -0800
Andrew:
I read Ryan Jones' book, but I found the mathematical support for his system
weak. I prefer and use Ralph Vince's optimal f, instead.
I think it was Gwenn who mentioned that risking 10% of your capital on a
trade is high. I tend to agree, but I wouldn't conclude that risking 10%
necessarily translates into a near-certain risk of ruin. Risk of ruin is
affected by the characteristics of your system, the markets you trade and
the largest loss you will suffer. One thing it cannot anticipate is the 3
Standard Deviation and 4 Standard Deviation losses (unless you have captured
one in your system testing) that traders are exposed to at some point in
their lifetime. Risk of ruin is certainly something you should calculate
and be aware of, particularly if your money management system is telling you
to risk big money.
You asked about the position size for your first trade. My recall may be
off, but I think this is related to choosing the right "delta", which was
Jones' inherent weakness. If you make a mistake in position sizing, this is
where it will be.
Before implementing a fixed ratio money management system, I would suggest
you read Ralph Vince's Portfolio Management Formulas and his Mathematics of
Money Management for a balanced viewpoint, not to mention a wake-up call
regarding risk.
Regards.
----- Original Message -----
From: "Andrew Peskin" <andrew@xxxxxxxxx>
Sent: Tuesday, February 29, 2000 11:32 AM
Subject: [RT] Fixed Ratio or Fixed Fractional?? - Pros and Cons of Each ...
(I apologize as this is somewhat long ...)
I would like to start a discussion on the merits and weaknesses of the
Fixed Ratio position sizing method popularized by Ryan Jones.
Specifically how does it stack up against the other predominant position
sizing method known as Fixed Fractional.
The majority of the research I have conducted was based upon the fixed
fractional method, or sizing your position based upon the risk the
position represents and the percent of your total equity you wish to
risk.
For Example:
You have a $100,000 account balance and wish to risk 10% of your account
on each trade, or $10,000.
You have an S&P Trading System which buys on a breakout of the highest
high of the last 5 bars and sells on a breakout of the lowest low of the
last 5 bars. You would reverse your position with each occurrence. (I
am not advocating this system, nor have I tested this, I am only using
it as an example).
Your first trade has a risk of $2,000 so you would trade 5 contracts
(10,000 / 2,000 = 5).
Your first trade is a $750 winner so your account balance is now
$103,750.
Your second trade has a risk of $4,250 (the market has a greater 5 day
range) and you would trade 2 contracts (103,750 / 4,250 = 2.44 ==> 2).
You would continue this as long as you traded.
The above makes sense logically to me. With the Fixed Ratio Method, I
see some Problems:
1. How many contracts do you trade on your first trade? The method
explains how to change your size once you have begun trading, but I am
unclear on how to size your very first trade.
2. All trades are sized the same. Why would you want to risk varying
amounts of capital? In the above example if you were trading a fixed 4
contracts per trade, you would assume risk of $8,000 on the first trade
and a risk of $17,000 on the second. I do not understand why different
trades should represent differing amounts of risk for your account if
the outcome of each trade is unknown and as the same probability of
success or failure.
I hope that this is a good start for an educational discussion amongst
ourselves regarding what many consider the most crucial aspect of
successful trading. I am looking forward to following the ensuing
dialogue.
Best regards,
Andrew Peskin
-- End --
From: "Gary Fritz" <fritz@xxxxxxxx>
To: <realtraders@xxxxxxxxxxxxxxx>
Date: Wed, 1 Mar 2000 15:45:24 -0700
Subject: [RT] Re: Fixed Ratio or Fixed Fractional?? Pros and Cons of Each
...
Priority: normal
In-reply-to: <003601bf83c2$172a3780$5f1c4118@xxxxxxxxxxxxxxxxxxxxxx>
> Before implementing a fixed ratio money management system, I would
> suggest you read Ralph Vince's Portfolio Management Formulas and
> his Mathematics of Money Management for a balanced viewpoint, not
> to mention a wake-up call regarding risk.
After reading Vince's book, I did a bit of experimentation with a
spreadsheet. The thing I didn't like about opt-f was that it
requires knowing the largest loss during the period -- which, of
course, you don't know ahead of time. It works great to decide what
you SHOULD have traded, after the fact, but it doesn't seem to me to
work so well for deciding what you should put on for your NEXT trade.
In reality, I think you'd have to use the "worst loss SO FAR," and
see what worked out best. Naturally, if at some point you take a
huge loss that wipes you out, that value of f will turn out to be
"too large" and won't come out as the optimal f value if you do an
opt-f analysis afterward. Which is nice in hindsight but wouldn't be
too comforting if you'd been trading near the previous opt f value
and got wiped out.
A friend of mine uses what he calls a "conservative f" which he
computes as AverageTrade / AverageWin. I forget his derivation of
this but he feels it's a safer approach than optimal f.
I decided to modify Vince's approach to use the AVERAGE loss SO FAR,
and use that to compute the f$ and betsize using that approach. The
formulas I used were:
AvgLoss = (average loss so far in the system's history)
f$ = -AvgLoss / f
BetSize = AccountValue / f$
Same as Vince, except I use average loss instead of largest loss.
Since the average loss is a bit more stable than the max loss, this
seemed to stabilize the results some. I usually ended up with an
"Average Loss f" value that was about 1/3 as large as the optimal f,
but the f$ value is only slightly smaller.
Interestingly enough, the TWR (Terminal Wealth Relative, the account
value at the end of the test) with my "Average Loss" f was MUCH
larger than the TWR for Vince's optimal f, which he claims is the
"largest possible TWR." (Although I used max loss SO FAR, and he may
be referring to the max loss during the whole test -- but that's not
too helpful for realtime trading.) E.g. in the test case I ran, my
"AvgLoss f" TWR was **25 times** larger than Vince's TWR.
Is this "AvgLoss f" any safer than optimal f? I think so. Certainly
trading *at* optimal f is lunacy, and you wouldn't want to do that.
I wouldn't want to trade at my "AvgLoss f" either. But you could
trade at 80% of the "AvgLoss f" (.142 instead of .178) and make as
much as you'd make trading at Vince's opt f, and I'd bet the non-
optimal "AvgLoss f" value is a LOT safer than trading at Vince's opt
f. So for less risk you get equivalent return. It seems to be a bit
more resistant to large losses, but I haven't studied that very much.
Is this really better than Vince's opt f? I have no idea. :-) I
don't have the math background nor the inclination to pursue it to
prove if it's a safer approach. But it sure *seems* to provide
better return for equivalent risk.
If anybody would like to pursue it, I can send you my spreadsheet.
It's about 420kb compressed, so it's too big to post to RT. It's not
really documented, but if you would like to take the above writeup
and a few additional pointers and figure out the spreadsheet on your
own, let me know and you're welcome to a copy.
Gary
-- End --
by coinet.com with SMTP; 1 Mar 2000 22:50:25 -0000
Date: Wed, 01 Mar 2000 14:50:21 -0800
From: Dennis Holverstott <dennis@xxxxxxxxxx>
To: <realtraders@xxxxxxxxxxxxxxx>
CC: realtraders@xxxxxxxxxxxxxxx
Subject: [RT] Re: [RT] Fixed Ratio or Fixed
> I tend to agree, but I wouldn't conclude that risking 10%
> necessarily translates into a near-certain risk of ruin. Risk of ruin is
> affected by the characteristics of your system, the markets you trade and
> the largest loss you will suffer.
I agree except I would make that the biggest STRING of losses you will
suffer. I haven't read Vince's books so correct me if I'm wrong but that
seems to be the big weakness of optimal f. It bases it's calcs on things
like the biggest loser, the average loser and the percentage of losers
but it doesn't consider what would happen if all the losers happened in
a row. So people try to give themselves a safety factor by trading a
some fraction of optimal f but Vince doesn't give you any guidance for
choosing that fraction. So, you might as well just hold your finger up
in the wind and guess. <joke> I've heard optimal-F nicknamed how F-ast
can you go broke. </joke>
> One thing it cannot anticipate is the 3
> Standard Deviation and 4 Standard Deviation losses (unless you have
captured
> one in your system testing) that traders are exposed to at some point in
> their lifetime.
Yup. Niederhoffer, LTCM, and the list goes on. IMHO, a Monte Carlo
simulation is the best way to do that. Take your historical trades (100
or more, trading one contract), shuffle them into a random order, and
calculate the max drawdown. Repeat 10,000 or so times and you can get
some interesting stats like 50% chance of $ww drawdown, 10% chance of
$xx drawdown, 1% chance of $yy drawdown and 0.1% chance of $zz drawdown.
Then decide your risk comfort level (aggressive or chicken) and trade
one contract for each $jj in your account.
For example, with my own system, cutting my size in half cuts my risk of
ruin from 10% to 0.1%. Being basically a chicken, that's a trade-off I'm
more than willing to make.
Actually, that's the simple version. To do it right, you need to adjust
your historical trades so each one had an equal dollar risk. Then run
the Monte Carlo and you can come up with a safe percentage of your
account to risk on each trade. Don't know your risk before you enter
each trade? You should. Once you have those numbers, it's a piece of
cake to implement.
> Risk of ruin is certainly something you should calculate
> and be aware of, particularly if your money management system is telling
you
> to risk big money.
Risk of ruin is the MOST important thing. That's where it all starts.
Without that, none of the rest of it matters.
--
Dennis
-- End --
(Sun Internet Mail Server sims.3.5.1999.09.16.21.57.p8)
with SMTP id <0FQR00GNJMUGZ1@xxxxxxxxxxxxxxxxxxxxxx> for
realtraders@xxxxxxxxxxxxxxx; Wed, 1 Mar 2000 17:06:17 -0600 (CST)
Date: Wed, 01 Mar 2000 17:06:15 -0600
From: Clyde Lee <clydelee@xxxxxxx>
Subject: [RT] Re: Fixed Ratio or Fixed Fractional?? Pros and Cons of Each
...
To: <realtraders@xxxxxxxxxxxxxxx>
Organization: SYTECH Corporation
Just for kicks, why would not WORST CASE DRAWDOWN
to date be something to base the computation on rather
than average loss or total loss in a string.
Remember, I'm asking since I really don't know much about
this -- I'm running a test right now on AverageLoss and
then will test these others on a very simple (Aberration like)
system.
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
Clyde Lee Chairman/CEO (Home of SwingMachine)
SYTECH Corporation email: <clydelee@xxxxxxx>
7910 Westglen, Suite 105 Work: (713) 783-9540
Houston, TX 77063 Fax: (713) 783-1092
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
----- Original Message -----
From: "Gary Fritz" <fritz@xxxxxxxx>
To: <realtraders@xxxxxxxxxxxxxxx>
Sent: Wednesday, March 01, 2000 16:45
Subject: [RT] Re: Fixed Ratio or Fixed Fractional?? Pros and Cons of
Each ...
> > Before implementing a fixed ratio money management system, I would
> > suggest you read Ralph Vince's Portfolio Management Formulas and
> > his Mathematics of Money Management for a balanced viewpoint, not
> > to mention a wake-up call regarding risk.
>
> After reading Vince's book, I did a bit of experimentation with a
> spreadsheet. The thing I didn't like about opt-f was that it
-- End --
From: gary@xxxxxxxxxxxx (Gary Funck)
Date: Wed, 1 Mar 2000 15:31:48 -0800
In-Reply-To: "Gary Fritz" <fritz@xxxxxxxx>
"[RT] Re: Fixed Ratio or Fixed Fractional?? Pros and Cons of Each
..." (Mar 1, 3:45pm)
To: fritz@xxxxxxxx, <realtraders@xxxxxxxxxxxxxxx>
Subject: Re: [RT] Fixed Ratio or Fixed Fractional?
Mime-Version: 1.0
As an alternative, you might look at the Kelly Ratio. Vince discusses
this, and McMillan discusses it in "McMIllan on Options" (as I recall).
[the following is excerpted from McMillan's Option Strategist
newsletter, circa 1994.]
How Much to Invest?
One should not put all of his speculative funds into one position: if there
is only a 40% to 50% chance of any one recommendation making money, then
he should alot his money over several different positions. In fact one
of the things that is hardest for any speculator to decide is "how much
of my speculative funds should I put into this one position?
It is actually possible to derive a mathematical answer to this question,
if one adheres to the "Kelly System". Kelly was a programmer at Bell
Telephone Laboratories, who derived a formula to answer the above question.
His formula is actually used by most gamblers who have a fixed result
(100% or -100% less "commission", on any one outcome). Therefore, we
have to adapt a little to use the Kelly formula. Here is the original
formula in its simplest form:
Amount of bet = (W + L) * p - 1
where W = amount you win, L = amount you could lose, and p = probability
of winning. For example, in a situation where you risk one "unit" and
pay a 10% commission, the Kelly formula would read: 2.1*p - 1.1*p.
Thus, if you can predict winners under the criteria at a 60% rate,
the Kelly system would tell you to bet 2.1*0.60 - 1.1*0.6 = 0.16 or
16% of your total bankroll on one trade. In this case, if p is less
than about 52%, the Kelly system would tell you not to bet all
(2.1*0.52 - 1.1*0.52 = -0.01).
Since the stock market is more complicated, we not only have to guage
the probability of having a winning trade, but also have to take into
account how big the wins and losses are. In this case, the Kelly formula
becomes:
where p = probability of winning and r = average win/average loss
(assuming equal investment in each trade).
In the data on page 2, we had 12 wins and the average winning trade
made $1444; we had 11 losses, and the average losing trade lost $501.
Therefore p = 0.52 (12 winning trades divided by 23 total trades), and
r = 2.88 (1444 divided by 501). This yields 35% as the amount to risk.
That is one should put about one third of his speculative funds into
each of our speculative recommendations, if he believes that our
recommendations will continue to profit at the same rate shown in
the table on page 2.
As we stated earlier, these returns are higher than normal, if you wre
to go back over all 83 speculative recommendations we have made since
our inception in December of 1990, you would find 35 winners and
48 losers (42%) with the average win netting $1113 and the average
loss costing $620. Using these figures, the Kelly System will tell
you to risk 9.7% (about 1/10) of your speculative bankroll on each
recommendation. That is much more conservative, and seems to make
more logical sense in the long run.
The Kelly criteria is a dynamic formula as the percentage of winners
and losers changes over time, but overall it is a reliable way to
be sure that you are risking enough, but not too much on your trades.
-- End --
From: "Gary Fritz" <fritz@xxxxxxxx>
To: <realtraders@xxxxxxxxxxxxxxx>
Date: Wed, 1 Mar 2000 16:36:05 -0700
Subject: [RT] Re: Fixed Ratio or Fixed Fractional?? Pros and Cons of
Each...
Priority: normal
In-reply-to: <38BDA4EA.92343B7B@xxxxxxxxxx>
> Just for kicks, why would not WORST CASE DRAWDOWN
> to date be something to base the computation on rather
> than average loss or total loss in a string.
Might work. Let us know what you find!
> That's what the Monte Carlo does. It assumes there is no
> correlation between the results of one trade and the next and that
> your realtime historical trade order was just a random chance
> (might not be true for all systems but it's true for mine).
That's my biggest gripe about MC testing. For my system at least,
and I suspect for many/most, the trade order is NOT random chance.
E.g. my system has been behaving VERY differently for the last 3
months than it did in the 2 years of testing before that. (And I
**love** it!! :-) The market is acting fundamentally differently
than it did before -- more volatility, etc -- and that affects the
trades that happen during this period.
I think many systems have "memory" like this. E.g. take a simple
breakout system. It's going to tend to have a string of losses
(fakeouts) followed by a big win when the market finally trends.
That's not a random series. MC testing of that will produce
questionable results that do not necessarily reflect reality.
Gary
-- End --
From: "Glen Wallace" <gcwallace@xxxxxxxx>
To: <realtraders@xxxxxxxxxxxxxxx>
Subject: [RT]
=?Windows-1252?Q?Re:_=5BRT=5D_Re:_=5BRT=5D_Fixed_Ratio_or_Fixed_Fract?=
Date: Wed, 1 Mar 2000 18:13:13 -0800
Dennis:
The risk of ruin calculation in Portfolio Management Formulas (Peter
Griffin's formula from The Theory of Blackjack) is designed to anticipate
the biggest string of average losses. Ralph Vince describes the concept in
great detail, but then dismisses it as useless in helping you actually
determine position size. By definition, trading at optimal f gives you the
optimum balance between return and risk of ruin. I suppose I see his point,
but personally I take comfort in calculating a low risk of ruin.
You're right that optimal f uses the largest loss to calculate position
size, but the nature of an anti-martingale (trade larger after a win and
smaller after a loss) and in particular a fixed fraction system manages the
effect of a string of losses.
The Monte Carlo simulation sounds kind of like Weibull analysis. Do you
have a spreadsheet application for the simulation? Can you point me to more
information on this method?
----- Original Message -----
From: "Dennis Holverstott" <dennis@xxxxxxxxxx>
To: "Glen Wallace" <gwallace@xxxxxxxx>
Cc: <realtraders@xxxxxxxxxxxxxxx>
Sent: Wednesday, March 01, 2000 2:50 PM
Subject: Re: [RT] Re: [RT] Fixed Ratio or Fixed Fractional??- Pros and Cons
of Each ...
> I tend to agree, but I wouldn't conclude that risking 10%
> necessarily translates into a near-certain risk of ruin. Risk of ruin is
> affected by the characteristics of your system, the markets you trade and
> the largest loss you will suffer.
I agree except I would make that the biggest STRING of losses you will
suffer. I haven't read Vince's books so correct me if I'm wrong but that
seems to be the big weakness of optimal f. It bases it's calcs on things
like the biggest loser, the average loser and the percentage of losers
but it doesn't consider what would happen if all the losers happened in
a row. So people try to give themselves a safety factor by trading a
some fraction of optimal f but Vince doesn't give you any guidance for
choosing that fraction. So, you might as well just hold your finger up
in the wind and guess. <joke> I've heard optimal-F nicknamed how F-ast
can you go broke. </joke>
> One thing it cannot anticipate is the 3
> Standard Deviation and 4 Standard Deviation losses (unless you have
> captured one in your system testing) that traders are exposed to at
> some point in their lifetime.
Yup. Niederhoffer, LTCM, and the list goes on. IMHO, a Monte Carlo
simulation is the best way to do that. Take your historical trades (100
or more, trading one contract), shuffle them into a random order, and
calculate the max drawdown. Repeat 10,000 or so times and you can get
some interesting stats like 50% chance of $ww drawdown, 10% chance of
$xx drawdown, 1% chance of $yy drawdown and 0.1% chance of $zz drawdown.
Then decide your risk comfort level (aggressive or chicken) and trade
one contract for each $jj in your account.
For example, with my own system, cutting my size in half cuts my risk of
ruin from 10% to 0.1%. Being basically a chicken, that's a trade-off I'm
more than willing to make.
Actually, that's the simple version. To do it right, you need to adjust
your historical trades so each one had an equal dollar risk. Then run
the Monte Carlo and you can come up with a safe percentage of your
account to risk on each trade. Don't know your risk before you enter
each trade? You should. Once you have those numbers, it's a piece of
cake to implement.
> Risk of ruin is certainly something you should calculate
> and be aware of, particularly if your money management system is telling
> you to risk big money.
Risk of ruin is the MOST important thing. That's where it all starts.
Without that, none of the rest of it matters.
--
Dennis
-- End --
From: "Glen Wallace" <gcwallace@xxxxxxxx>
To: <realtraders@xxxxxxxxxxxxxxx>
Subject: [RT] Re: Fixed Ratio or Fixed Fractional?? Pros and Cons of Each
...
Date: Wed, 1 Mar 2000 18:19:55 -0800
Gary:
Back testing over a long period of time over all kinds of market conditions
will tend to give you a more representative largest loss. Probably not the
career-ender loss that's in store for a person if they do this long enough,
but I suppose that's the risk ya' take :)
In my opinion, the key is to stay to the left of the optimal f peak (for
those not familiar with optimal f, picture the peak of a normal distribution
curve) and accept a lower return for the reduced risk. Not too far to the
left, mind you, because the reduction in return is not proportional to the
reduction in risk. Trading at less than optimal and how much less is up to
the individual and their own risk threshhold.
I would be interested in hearing more about your friend's Conservative f
technique. Also, be careful with substituting average loss for largest
loss. The effect of this is to reduce your equity-per-contract figure (f$)
which, in turn, might make your position sizes too big when you hit a string
of large or largest losses.
Regards.
----- Original Message -----
From: "Gary Fritz" <fritz@xxxxxxxx>
To: <realtraders@xxxxxxxxxxxxxxx>
Sent: Wednesday, March 01, 2000 2:45 PM
Subject: [RT] Re: Fixed Ratio or Fixed Fractional?? Pros and Cons of Each
> Before implementing a fixed ratio money management system, I would
> suggest you read Ralph Vince's Portfolio Management Formulas and
> his Mathematics of Money Management for a balanced viewpoint, not
> to mention a wake-up call regarding risk.
After reading Vince's book, I did a bit of experimentation with a
spreadsheet. The thing I didn't like about opt-f was that it
requires knowing the largest loss during the period -- which, of
course, you don't know ahead of time. It works great to decide what
you SHOULD have traded, after the fact, but it doesn't seem to me to
work so well for deciding what you should put on for your NEXT trade.
In reality, I think you'd have to use the "worst loss SO FAR," and
see what worked out best. Naturally, if at some point you take a
huge loss that wipes you out, that value of f will turn out to be
"too large" and won't come out as the optimal f value if you do an
opt-f analysis afterward. Which is nice in hindsight but wouldn't be
too comforting if you'd been trading near the previous opt f value
and got wiped out.
A friend of mine uses what he calls a "conservative f" which he
computes as AverageTrade / AverageWin. I forget his derivation of
this but he feels it's a safer approach than optimal f.
I decided to modify Vince's approach to use the AVERAGE loss SO FAR,
and use that to compute the f$ and betsize using that approach. The
formulas I used were:
AvgLoss = (average loss so far in the system's history)
f$ = -AvgLoss / f
BetSize = AccountValue / f$
Same as Vince, except I use average loss instead of largest loss.
Since the average loss is a bit more stable than the max loss, this
seemed to stabilize the results some. I usually ended up with an
"Average Loss f" value that was about 1/3 as large as the optimal f,
but the f$ value is only slightly smaller.
Interestingly enough, the TWR (Terminal Wealth Relative, the account
value at the end of the test) with my "Average Loss" f was MUCH
larger than the TWR for Vince's optimal f, which he claims is the
"largest possible TWR." (Although I used max loss SO FAR, and he may
be referring to the max loss during the whole test -- but that's not
too helpful for realtime trading.) E.g. in the test case I ran, my
"AvgLoss f" TWR was **25 times** larger than Vince's TWR.
Is this "AvgLoss f" any safer than optimal f? I think so. Certainly
trading *at* optimal f is lunacy, and you wouldn't want to do that.
I wouldn't want to trade at my "AvgLoss f" either. But you could
trade at 80% of the "AvgLoss f" (.142 instead of .178) and make as
much as you'd make trading at Vince's opt f, and I'd bet the non-
optimal "AvgLoss f" value is a LOT safer than trading at Vince's opt
f. So for less risk you get equivalent return. It seems to be a bit
more resistant to large losses, but I haven't studied that very much.
Is this really better than Vince's opt f? I have no idea. :-) I
don't have the math background nor the inclination to pursue it to
prove if it's a safer approach. But it sure *seems* to provide
better return for equivalent risk.
If anybody would like to pursue it, I can send you my spreadsheet.
It's about 420kb compressed, so it's too big to post to RT. It's not
really documented, but if you would like to take the above writeup
and a few additional pointers and figure out the spreadsheet on your
own, let me know and you're welcome to a copy.
Gary
-- End --
(Sun Internet Mail Server sims.3.5.1999.09.16.21.57.p8)
with SMTP id <0FQS0072P0FUH9@xxxxxxxxxxxxxxxxxxxxxx> for fritz@xxxxxxxx;
Wed,
1 Mar 2000 21:59:58 -0600 (CST)
Date: Wed, 01 Mar 2000 21:59:48 -0600
From: Clyde Lee <clydelee@xxxxxxx>
Subject: Re: [RT] Re: Fixed Ratio or Fixed Fractional?? Pros and Cons of
Each...
To: Fritz Gary Fritz <fritz@xxxxxxxx>, realtraders@xxxxxxxxxxxxxxx
Organization: SYTECH Corporation
boundary="----=_NextPart_000_00AF_01BF83C9.762BA460"
This is a multi-part message in MIME format.
------=_NextPart_000_00AF_01BF83C9.762BA460
------=_NextPart_001_00B0_01BF83C9.762BA460
First, let me set the scene!
The input data is the OEX cash index from some time in
1996 to now taken on a 30 minute interval. For purposes
of this study I assigned a value of $300 for each OEX
full point.
The system was a modification of the old Aberration system
using an adaptive smoothed input series of closes (thanks
to an idea from Mark Brown of a while back) and +/-2 standard
deviations for entry long and short with exit at the average
line.
This is what I used for "optimal f" computation. Note that
this averaged the total loss of up to three sequential losses
rather than just the simple single loss. This to get=20
around an earlier criticism about not fully adjusting for
a string of losses. However, I am inclined to think that a
moving average of losses would work better than either=20
average approach.
If PositionProfit(1)<0 then begin
AvgLoss =3D AvgLoss*NumbLoss-PositionProfit(1);
NumbLoss=3D NumbLoss+1;
If PositionProfit(2)<0 then=20
AvgLoss =3D AvgLoss-PositionProfit(1)
Else If PositionProfit(3)<0 then=20
AvgLoss =3D AvgLoss-PositionProfit(1)
Else If PositionProfit(4)<0 then=20
AvgLoss =3D AvgLoss-PositionProfit(1);
AvgLoss =3D AvgLoss/NumbLoss;
End;
if avgloss>0 then Ncont =3D =
Round(RiskPerc*(InitCash+TotalProfit)/avgloss,0);
And the following is what I used for Fixed Percent .
Value1 =3D Stddev(Close, StdDevLn)*2*BigPointValue ;
If Value1>0 then Ncont =3D =
round(RiskPerc*(InitCash+TotalProfit)/Value1, 0);
Here are some statistics for these two methods over
the range of approximately the same DOLLAR risks.
Percentage risks are higher for Average Loss case since
the Average Loss is considerably higher than 2 times StdDev
through most of the period. The attached image shows each
loss over this period. In order the date, the amount of loss,
the "average" loss (as defined above), and 2*StdDev.
I don't know about you but I'm sure sticking with this
fixed percentage of account divided by 2 times Standard
Deviation for determining bet size.
Clyde
FIXED PERCENTAGE OF ACCOUNT DIVIDED BY AVG LOSS
=20
%RISK NetPrft L:NetPrft S:NetPrft ROA =20
11.00 56219.00 118965.00 -62746.00 166.32
13.00 62739.00 127912.00 -65173.00 171.73
15.00 54839.00 123659.00 -68820.00 131.52
17.00 50520.00 125055.00 -74535.00 112.24
19.00 70540.00 151839.00 -81299.00 147.19
21.00 80118.00 163475.00 -83357.00 173.16
23.00 65716.00 162291.00 -96575.00 125.69
25.00 65639.00 159729.00 -94090.00 116.97
27.00 65191.00 164590.00 -99399.00 106.38
29.00 53265.00 162720.00 -109455.00 82.59
31.00 71063.00 177374.00 -106311.00 118.09
%RISK MaxDD BigWTrd BigLTrd
11.00 -33801.00 32275.00 -16766.00
13.00 -36533.00 32275.00 -16766.00
15.00 -41696.00 38730.00 -16766.00
17.00 -45010.00 38730.00 -19551.00
19.00 -47924.00 45185.00 -19551.00
21.00 -46267.00 45185.00 -25149.00
23.00 -52284.00 45185.00 -25149.00
25.00 -56115.00 45185.00 -25149.00
27.00 -61280.00 51640.00 -25149.00
29.00 -64494.00 51640.00 -26068.00
31.00 -60179.00 51640.00 -26068.00
=20
FIXED PERCENTAGE OF ACCOUNT DIVIDED BY STD DEV=20
=20
%RISK NetPrft L:NetPrft S:NetPrft ROA =20
1.00 26418.00 53929.00 -27511.00 181.14
2.00 30361.00 60230.00 -29869.00 174.10
3.00 62974.00 102988.00 -40014.00 236.12
4.00 99460.00 145056.00 -45596.00 337.93
5.00 115231.00 180029.00 -64798.00 249.73
6.00 122285.00 213547.00 -91262.00 215.03
7.00 167405.00 279071.00 -111666.00 247.38
8.00 169136.00 307220.00 -138084.00 195.62
9.00 163406.00 322314.00 -158908.00 160.74
10.00 191363.00 384644.00 -193281.00 170.63
11.00 176102.00 395815.00 -219713.00 133.62
=20
=20
%RISK MaxDD BigWTrd BigLTrd
1.00 -14584.00 13772.00 -8383.00
2.00 -17439.00 13772.00 -8383.00
3.00 -26670.00 19628.00 -10814.00
4.00 -29432.00 27544.00 -16221.00
5.00 -46142.00 38455.00 -27035.00
6.00 -56868.00 41316.00 -32442.00
7.00 -67671.00 55088.00 -48663.00
8.00 -86461.00 61528.00 -54070.00
9.00 -101659.00 68860.00 -59477.00
10.00 -112149.00 82632.00 -81105.00
11.00 -131797.00 84601.00 -86512.00
=20
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
Clyde Lee Chairman/CEO (Home of SwingMachine)
SYTECH Corporation email: <clydelee@xxxxxxx>=20
7910 Westglen, Suite 105 Work: (713) 783-9540
Houston, TX 77063 Fax: (713) 783-1092 =20
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
----- Original Message -----=20
From: "Gary Fritz" <fritz@xxxxxxxx>
To: <realtraders@xxxxxxxxxxxxxxx>
Sent: Wednesday, March 01, 2000 17:36
Subject: [RT] Re: Fixed Ratio or Fixed Fractional?? Pros and Cons of =
Each...
> > Just for kicks, why would not WORST CASE DRAWDOWN
> > to date be something to base the computation on rather
> > than average loss or total loss in a string.
>=20
> Might work. Let us know what you find!
>=20
> > That's what the Monte Carlo does. It assumes there is no
>=20
------=_NextPart_001_00B0_01BF83C9.762BA460
<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.0 Transitional//EN">
<HTML><HEAD>
<META http-equiv=3DContent-Type content=3D"text/html; =
charset=3Diso-8859-1">
<META content=3D"MSHTML 5.50.3825.1300" name=3DGENERATOR>
<STYLE></STYLE>
</HEAD>
<BODY bgColor=3D#ffffff>
<DIV><FONT face=3D"Courier New" size=3D2>First, let me set the =
scene!</FONT></DIV>
<DIV><FONT face=3D"Courier New" size=3D2></FONT> </DIV>
<DIV><FONT face=3D"Courier New" size=3D2>The input data is the OEX cash =
index from=20
some time in</FONT></DIV>
<DIV><FONT face=3D"Courier New" size=3D2>1996 to now taken on a 30 =
minute=20
interval. For purposes</FONT></DIV>
<DIV><FONT face=3D"Courier New" size=3D2>of this study I assigned a =
value of $300=20
for each OEX</FONT></DIV>
<DIV><FONT face=3D"Courier New" size=3D2>full point.</FONT></DIV>
<DIV><FONT face=3D"Courier New" size=3D2></FONT> </DIV>
<DIV><FONT face=3D"Courier New" size=3D2>The system was a modification =
of the old=20
Aberration system</FONT></DIV>
<DIV><FONT face=3D"Courier New" size=3D2>using an adaptive smoothed =
input series of=20
closes (thanks</FONT></DIV>
<DIV><FONT face=3D"Courier New" size=3D2>to an idea from Mark Brown of a =
while back)=20
and +/-2 standard</FONT></DIV>
<DIV><FONT face=3D"Courier New" size=3D2>deviations for entry long and =
short with=20
exit at the average</FONT></DIV>
<DIV><FONT face=3D"Courier New" size=3D2>line.</FONT></DIV>
<DIV><FONT face=3D"Courier New" size=3D2></FONT> </DIV>
<DIV><FONT face=3D"Courier New" size=3D2></FONT> </DIV>
<DIV><FONT face=3D"Courier New" size=3D2>This is what I used for =
"optimal f"=20
computation. Note that</FONT></DIV>
<DIV><FONT face=3D"Courier New" size=3D2>this averaged the total loss of =
up to three=20
sequential losses</FONT></DIV>
<DIV><FONT face=3D"Courier New" size=3D2>rather than just the simple =
single=20
loss. This to get </FONT></DIV>
<DIV><FONT face=3D"Courier New" size=3D2>around an earlier criticism =
about not fully=20
adjusting for</FONT></DIV>
<DIV><FONT face=3D"Courier New" size=3D2>a string of losses. =
However, I am=20
inclined to think that a</FONT></DIV>
<DIV><FONT face=3D"Courier New" size=3D2>moving average of losses would =
work better=20
than either </FONT></DIV>
<DIV><FONT face=3D"Courier New" size=3D2>average approach.</FONT></DIV>
<DIV><FONT face=3D"Courier New" size=3D2></FONT> </DIV>
<DIV><FONT face=3D"Courier New" size=3D2> If =
PositionProfit(1)<0 then=20
begin<BR> AvgLoss =3D=20
AvgLoss*NumbLoss-PositionProfit(1);<BR> NumbLoss=3D=20
NumbLoss+1;<BR> If PositionProfit(2)<0 then=20
<BR> AvgLoss =3D=20
AvgLoss-PositionProfit(1)<BR> Else If =
PositionProfit(3)<0=20
then <BR> AvgLoss =3D=20
AvgLoss-PositionProfit(1)<BR> Else If =
PositionProfit(4)<0=20
then <BR> AvgLoss =3D=20
AvgLoss-PositionProfit(1);<BR> AvgLoss =3D=20
AvgLoss/NumbLoss;<BR> End;<BR> if avgloss>0 then Ncont =
=3D=20
Round(RiskPerc*(InitCash+TotalProfit)/avgloss,0);</FONT></DIV>
<DIV><FONT face=3D"Courier New" size=3D2></FONT> </DIV>
<DIV><FONT face=3D"Courier New" size=3D2></FONT> </DIV>
<DIV><FONT face=3D"Courier New" size=3D2>And the following is what I =
used for Fixed=20
Percent .</FONT></DIV>
<DIV><FONT face=3D"Courier New" size=3D2></FONT> </DIV>
<DIV><FONT face=3D"Courier New" size=3D2> Value1 =3D Stddev(Close, =
StdDevLn)*2*BigPointValue ;<BR> If Value1>0 then Ncont =
=3D=20
round(RiskPerc*(InitCash+TotalProfit)/Value1, 0);</FONT></DIV>
<DIV><FONT face=3D"Courier New" size=3D2></FONT> </DIV>
<DIV><FONT face=3D"Courier New" size=3D2>Here are some statistics for =
these two=20
methods over</FONT></DIV>
<DIV><FONT face=3D"Courier New" size=3D2>the range of approximately the =
same DOLLAR=20
risks.</FONT></DIV>
<DIV><FONT face=3D"Courier New" size=3D2></FONT> </DIV>
<DIV><FONT face=3D"Courier New" size=3D2>Percentage risks are higher for =
Average=20
Loss case since</FONT></DIV>
<DIV><FONT face=3D"Courier New" size=3D2>the Average Loss is =
considerably higher=20
than 2 times StdDev</FONT></DIV>
<DIV><FONT face=3D"Courier New" size=3D2>through most of the =
period. The=20
attached image shows each</FONT></DIV>
<DIV><FONT face=3D"Courier New" size=3D2>loss over this period. In =
order the=20
date, the amount of loss,</FONT></DIV>
<DIV><FONT face=3D"Courier New" size=3D2>the "average" loss (as defined =
above), and=20
2*StdDev.</FONT></DIV>
<DIV><FONT face=3D"Courier New" size=3D2></FONT> </DIV>
<DIV><FONT face=3D"Courier New" size=3D2>I don't know about you but I'm =
sure=20
sticking with this</FONT></DIV>
<DIV><FONT face=3D"Courier New" size=3D2>fixed percentage of account =
divided by 2=20
times Standard</FONT></DIV>
<DIV><FONT face=3D"Courier New" size=3D2>Deviation for determining bet=20
size.</FONT></DIV>
<DIV><FONT face=3D"Courier New" size=3D2></FONT> </DIV>
<DIV><FONT face=3D"Courier New" size=3D2>Clyde</FONT></DIV>
<DIV><FONT face=3D"Courier New" size=3D2></FONT> </DIV>
<DIV><FONT face=3D"Courier New" size=3D2> FIXED PERCENTAGE OF =
ACCOUNT DIVIDED=20
BY AVG LOSS<BR> <BR> %RISK NetPrft =
L:NetPrft =20
S:NetPrft ROA <BR> 11.00 =
56219.00 =20
118965.00 -62746.00 166.32<BR> 13.00 =
62739.00 =20
127912.00 -65173.00 171.73<BR> 15.00 =
54839.00 =20
123659.00 -68820.00 131.52<BR> 17.00 =
50520.00 =20
125055.00 -74535.00 112.24<BR> 19.00 =
70540.00 =20
151839.00 -81299.00 147.19<BR> 21.00 =
80118.00 =20
163475.00 -83357.00 173.16<BR> 23.00 =
65716.00 =20
162291.00 -96575.00 125.69<BR> 25.00 =
65639.00 =20
159729.00 -94090.00 116.97<BR> 27.00 =
65191.00 =20
164590.00 -99399.00 106.38<BR> 29.00 =
53265.00 =20
162720.00 -109455.00 82.59<BR> 31.00 =
71063.00 =20
177374.00 -106311.00 118.09</FONT></DIV>
<DIV> </DIV>
<DIV><FONT face=3D"Courier New" =
size=3D2> %RISK =20
MaxDD BigWTrd =
BigLTrd<BR> 11.00 =20
-33801.00 32275.00 -16766.00<BR> 13.00 =20
-36533.00 32275.00 -16766.00<BR> 15.00 =20
-41696.00 38730.00 -16766.00<BR> 17.00 =20
-45010.00 38730.00 -19551.00<BR> 19.00 =20
-47924.00 45185.00 -19551.00<BR> 21.00 =20
-46267.00 45185.00 -25149.00<BR> 23.00 =20
-52284.00 45185.00 -25149.00<BR> 25.00 =20
-56115.00 45185.00 -25149.00<BR> 27.00 =20
-61280.00 51640.00 -25149.00<BR> 29.00 =20
-64494.00 51640.00 -26068.00<BR> 31.00 =20
-60179.00 51640.00 -26068.00<BR> </FONT></DIV>
<DIV> </DIV>
<DIV><FONT face=3D"Courier New" size=3D2></FONT> </DIV>
<DIV> </DIV>
<DIV><FONT face=3D"Courier New" size=3D2><BR>FIXED PERCENTAGE OF ACCOUNT =
DIVIDED BY=20
STD DEV <BR> <BR> %RISK NetPrft =
L:NetPrft =20
S:NetPrft ROA <BR> 1.00 =20
26418.00 53929.00 -27511.00 181.14<BR> =20
2.00 30361.00 60230.00 -29869.00 =20
174.10<BR> 3.00 62974.00 102988.00 =
-40014.00 =20
236.12<BR> 4.00 99460.00 145056.00 =
-45596.00 =20
337.93<BR> 5.00 115231.00 180029.00 =
-64798.00 =20
249.73<BR> 6.00 122285.00 213547.00 =
-91262.00 =20
215.03<BR> 7.00 167405.00 279071.00 -111666.00 =20
247.38<BR> 8.00 169136.00 307220.00 -138084.00 =20
195.62<BR> 9.00 163406.00 322314.00 -158908.00 =20
160.74<BR> 10.00 191363.00 384644.00 -193281.00 =20
170.63<BR> 11.00 176102.00 395815.00 -219713.00 =20
133.62<BR> <BR> <BR> %RISK =20
MaxDD BigWTrd BigLTrd<BR> =
1.00 =20
-14584.00 13772.00 -8383.00<BR> 2.00 =20
-17439.00 13772.00 -8383.00<BR> 3.00 =20
-26670.00 19628.00 -10814.00<BR> 4.00 =20
-29432.00 27544.00 -16221.00<BR> 5.00 =20
-46142.00 38455.00 -27035.00<BR> 6.00 =20
-56868.00 41316.00 -32442.00<BR> 7.00 =20
-67671.00 55088.00 -48663.00<BR> 8.00 =20
-86461.00 61528.00 -54070.00<BR> 9.00 =
-101659.00 =20
68860.00 -59477.00<BR> 10.00 -112149.00 82632.00=20
-81105.00<BR> 11.00 -131797.00 84601.00 =
-86512.00</FONT></DIV>
<DIV> </DIV>
<DIV><FONT face=3D"Courier New" size=3D2></FONT> </DIV>
<DIV><FONT face=3D"Courier New" size=3D2></FONT> </DIV>
<DIV><FONT face=3D"Courier New" size=3D2><BR></FONT> </DIV>
<DIV><FONT face=3D"Courier New" size=3D2>- - - - - - - - - - - - - - - - =
- - - - - -=20
- - - - - - - -<BR>Clyde Lee =20
Chairman/CEO (Home of=20
SwingMachine)<BR>SYTECH=20
Corporation &n=
bsp; =20
email: <</FONT><A href=3D"mailto:clydelee@xxxxxxx";><FONT=20
face=3D"Courier New" size=3D2>clydelee@xxxxxxx</FONT></A><FONT =
face=3D"Courier New"=20
size=3D2>> <BR>7910 Westglen, Suite =
105 =20
Work: (713) 783-9540<BR>Houston, TX =20
77063 =20
Fax: (713) 783-1092 <BR>- - - =
- - - -=20
- - - - - - - - - - - - - - - - - - - - - - -<BR></FONT></DIV>
<DIV><FONT face=3D"Courier New" size=3D2>----- Original Message ----- =
</FONT>
<DIV><FONT face=3D"Courier New" size=3D2>From: "Gary Fritz" =
<</FONT><A=20
href=3D"mailto:fritz@xxxxxxxx";><FONT face=3D"Courier New"=20
size=3D2>fritz@xxxxxxxx</FONT></A><FONT face=3D"Courier New"=20
size=3D2>></FONT></DIV>
<DIV><FONT face=3D"Courier New" size=3D2>To: <</FONT><A=20
href=3D"mailto:realtraders@xxxxxxxxxxxxxxx";><FONT face=3D"Courier New"=20
size=3D2>realtraders@xxxxxxxxxxxxxxx</FONT></A><FONT face=3D"Courier =
New"=20
size=3D2>></FONT></DIV>
<DIV><FONT face=3D"Courier New" size=3D2>Sent: Wednesday, March 01, 2000 =
17:36</FONT></DIV>
<DIV><FONT face=3D"Courier New" size=3D2>Subject: [RT] Re: Fixed Ratio =
or Fixed=20
Fractional?? Pros and Cons of Each...</FONT></DIV></DIV>
<DIV><FONT face=3D"Courier New"><BR><FONT =
size=3D2></FONT></FONT></DIV><FONT=20
face=3D"Courier New" size=3D2>> > Just for kicks, why would not =
WORST CASE=20
DRAWDOWN<BR>> > to date be something to base the computation on=20
rather<BR>> > than average loss or total loss in a string.<BR>> =
<BR>> Might work. Let us know what you find!<BR>> <BR>> =
>=20
That's what the Monte Carlo does. It assumes there is no<BR>>=20
</FONT></BODY></HTML>
------=_NextPart_001_00B0_01BF83C9.762BA460--
------=_NextPart_000_00AF_01BF83C9.762BA460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------=_NextPart_000_00AF_01BF83C9.762BA460--
-- End --
From: "Scot Billington" <scot.billington@xxxxxxxxxxxxx>
To: "<realtraders@xxxxxxxxxxxxxxx>"
Subject: [RT] Re: Fixed Ratio or Fixed Fractional?? A different look
Date: Thu, 2 Mar 2000 08:49:35 -0600
We use optimal f in a little different manner. During most winning trades
your risk increases intratrade as the market moves in your direction more
quickly than you move up your stop. If you have a profit objective, you
have more risk as you near the objective etc.
We take a very small % initial risk which allows us roughly 100 consecutive
losses before we hit a 50% drawdown. As the trade moves in our favor,
creating more risk, we accept that extra risk since it is not our 'core
captial' (total equity not at risk in other positions) at risk. We allow
the intra trade risk to increase up to 80% of our approximated optimal f.
If a market moves enough in our direction to reach optimal f, meaning the
risk has expanded multiple times, we reduce the risk by reducing our
position size. This is done by either covering contracts or selling options
vs. them. This in essence pins our risk to the optimal f for the duration
of the trade.
I find this strategy very useful. It minimizes initial risk and therefore
risk of ruin. It also allows you to take more risk when you are winning on
a trade by trade and portfolio basis. Large winning trade may operate right
at the optimal number. It has improved our risk/return ratio on the trades
on which it was used by over 7-1.
sb
----- Original Message -----
From: "Glen Wallace" <gcwallace@xxxxxxxx>
To: <realtraders@xxxxxxxxxxxxxxx>
Sent: Wednesday, March 01, 2000 8:19 PM
Subject: [RT] Re: Fixed Ratio or Fixed Fractional?? Pros and Cons of Each
..
> Gary:
>
> Back testing over a long period of time over all kinds of market
conditions
> will tend to give you a more representative largest loss. Probably not
the
> career-ender loss that's in store for a person if they do this long
enough,
> but I suppose that's the risk ya' take :)
>
> In my opinion, the key is to stay to the left of the optimal f peak (for
> those not familiar with optimal f, picture the peak of a normal
distribution
> curve) and accept a lower return for the reduced risk. Not too far to the
> left, mind you, because the reduction in return is not proportional to the
> reduction in risk. Trading at less than optimal and how much less is up
to
> the individual and their own risk threshhold.
>
> I would be interested in hearing more about your friend's Conservative f
> technique. Also, be careful with substituting average loss for largest
> loss. The effect of this is to reduce your equity-per-contract figure
(f$)
> which, in turn, might make your position sizes too big when you hit a
string
> of large or largest losses.
>
> Regards.
>
>
-- End --
|