Timid Equity / Bold Equity - Position Sizing Method

Discussions about Money Management and Risk Control.
Post Reply
TK
Roundtable Knight
Roundtable Knight
Posts: 167
Joined: Tue Apr 15, 2003 5:45 pm

Timid Equity / Bold Equity - Position Sizing Method

Post by TK »

Over two years ago, Mark Johnson presented Van Tharp’s position sizing technique that he chose to call Timid Equity / Bold Equity. For the description of the method, please follow the following link:

http://groups.yahoo.com/group/aaft_ta/message/953

Mark wrote:
I ran Aberration on the 22 market portfolio, establishing
a baseline by using a Fixed Fractional betsize approach
which risked 2.5% of equity on every trade. From 1/1/1980
till 3/2/2001 this FF approach gave a compound annual growth
rate ("CAGR") of +70.46 percent per year, for over 20 years.
Each dollar of initial equity on 1/1/1980 had grown to
almost $67,000 by 3/2/2001.

Then I ran it again on the same portfolio for the same
time period, but using the TIMID + BOLD equity idea from
Van Tharp. CAGR increased to +96.84 percent per year.
Each dollar of initial equity had grown to $1.34 million
by 3/2/2001.

The neat thing is, that drawdown *OF* *TIMID* *EQUITY*
didn't increase.

I found the results very attractive and decided to test the idea myself. The results that I got were not that appealing. Luckily, Mark was also kind enough to provide raw equity data so that those interested could analyse them and ponder the results. I would like to share my comments/concerns:

What I don’t understand is why I should compare FF approach drawdown of TOTAL EQUITY with Timid&Bold approach drawdown of TIMID EQUITY (both of which are about 42%). While I understand that in the Timid&Bold approach we are concerned with the drawdown of timid equity, I would not have any problem with being concerned with the drawdown of timid equity only also in case of the FF approach, if only it could offer higher reward. I analysed Mark's data and I found the following drawdown figures (see the attached spreadsheet):

FIXED FRACTIONAL APPROACH:
Max %DD of Total Equity: -41.32%
Max %DD of Timid Equity: -30.14%

TIMID & BOLD APPROACH:
Max %DD of Total Equity: -57.91%
Max %DD of Timid Equity: -42.59%


Now, I think that on the basis of those figures, you cannot claim that Timid & Bold approach allows you to increase profits WITHOUT increasing the drawdown. If think a valid comparison could only be made, if drawdowns of *timid equities* OR *total equities* in the two position sizing approaches were the same. Then, however, it could turn out that when the drawdowns of total equity of both approaches are equal, the drawdowns of timid equity are also similar and the profit is much higher in the case of the good old FF approach.

This is what I have found in my tests, although not on a regular basis. There were some periods / portfolios / systems on which FF approach worked better, and there were others that benefited more from the Timid & Bold approach. The only thing that I would like to stress here is that you cannot expect the Timid & Bold approach to improve your MAR ratio without redefining the drawdown concept. But then you can redefine it as well in the FF approach and get "better" results.

Of course, I do not mean to downgrade or criticize Mark Johnson’s excellent input on this or other forums. In fact, it is perfectly possible that I have misunderstood the whole concept or Mark’s conclusions. If this is the case, please explain what I am missing.
Attachments
timidbold TK.rar
(765.19 KiB) Downloaded 1425 times
steady_jake
Full Member
Full Member
Posts: 22
Joined: Fri Jun 27, 2003 8:05 pm
Location: Warsaw, Poland
Contact:

Post by steady_jake »

Tomek,
Now, I think that on the basis of those figures, you cannot claim that Timid & Bold approach allows you to increase profits WITHOUT increasing the drawdown.

The logic of playing market's money as presented in Van's report is not to improve your reward to risk characteristics, but to go for the maximum return while limiting the risk on your starting equity only. You may not confuse the risk defined by "I want to limit my risk on my initial capital to be lost" with the risk defined by the maximum drawdown further along the equity curve. These are two different things.

This particular money management model is a good example of something tailored to a particular objective. While your risk on your initial equity is limited, the risk on market's money is greater. The overall drawdown may put you easily back into your own equity, but with this particular approach you hope that you will have a good start right away and you will not have a drawdown early. How early it will come is an additional probability factor in this model and is clearly connected with the goal one is trying to achieve with it. This uncertainty needs to be taken into consideration in overall evaluation. If you want steady returns and you are not necessarily willing to take an additional risk of having no returns at all in return for a probability of having high returns faster, then probably you will be well off with a simple % Risk or % Volatility model.

Hope this clarification helps.

Regards,

Jakub
TK
Roundtable Knight
Roundtable Knight
Posts: 167
Joined: Tue Apr 15, 2003 5:45 pm

Post by TK »

Jakub,

I understand the logic behind that model and I do differentiate between the two kinds of risks / drawdowns. However, if you use %Risk model and persuade yourself that from that time on the only drawdown to look at will be the drawdown of your initial / timid equity that you want to protect, you may find that on some occasions the %Risk model generates better returns while protecting your initial / timid equity equally well as the timid / bold equity model. And if so, what is the value of the timid / bold equity model?

Have you run any tests and looked at the numbers?

Cheers,
Tomasz
steady_jake
Full Member
Full Member
Posts: 22
Joined: Fri Jun 27, 2003 8:05 pm
Location: Warsaw, Poland
Contact:

Post by steady_jake »

Tomasz,
... you may find that on some occasions the %Risk model generates better returns while protecting your initial / timid equity equally well as the timid / bold equity model. And if so, what is the value of the timid / bold equity model?
Let me look into this and get back to you in a couple of days.

Jakub
Kiwi
Roundtable Knight
Roundtable Knight
Posts: 513
Joined: Wed Apr 16, 2003 1:18 am
Location: Nowhere near

Post by Kiwi »

A thought.

To me the benefit here seems to be that you are making it psychologically easier to take a high risk (because you say to yourself that you are only risking more on money you havent got yet :-) ).

I looked at this for a short term system I have (200+ trades per year) and it works well. On the long term systems (5 trades*10 commodities) it didnt work as well because it often took longer to get to the point where "bold rules" cut in.

What would it look like if you used a longer time period - so you run it out for 2 years or 5 years. Effectively you compare the equity curve of the bold equity model with the underlying timid equity curve. You might well find that after year one your drawdown on timid equity was zero - how would that feel?
Pretty relaxed?

Just a thought.

John
sdzh
Contributor
Contributor
Posts: 2
Joined: Wed Jan 26, 2005 9:17 am
Location: The Steppes of Central Asia

Post by sdzh »

Interesting thread, but the link http://groups.yahoo.com/group/aaft_ta/message/953 doesn't appear to work any more.
Can anyone assist with providing the text ?? :D
TK
Roundtable Knight
Roundtable Knight
Posts: 167
Joined: Tue Apr 15, 2003 5:45 pm

Post by TK »

Mark Johnson wrote:Van Tharp sells a book for $80 called "Special Report on
Money Management". One of its ideas (pp. 53-59) is
something I will call Timid Equity / Bold Equity.
I programmed it up and tested it out.

First, a description of the betsizing idea:

You mentally divide your total account equity into a
TIMID part and a BOLD part. You make small bets with
your TIMID equity and large bets with your BOLD equity.
You will accept huge drawdowns of your BOLD equity, but
you can tolerate only small drawdowns of your TIMID equity.
From time to time, you sweep all the BOLD equity into
your TIMID equity column.

I tested the idea in the following way:

1. I used the Aberration(80, 2.0) futures trading system
on a portfolio of 22 markets. I deducted $150.00 per contract
for commissions and slippage.

2. I bet 2.5% of TIMID equity, *plus* 7.5% of BOLD equity,
on every trade.

3. I reclaimed BOLD equity and turned it into TIMID equity,
every time the total account value doubled.

Here is my computer program code for the betsizing part
of the backtesting software, written for "Trading Recipes"
( http://www.tradingrecipes.com ):

STARTUPCASH = 100000
' PUT TIMID EQUITY INTO MEMORY[3]
IF (MEMORY[3] < 100000) THEN MEMORY[3] = 100000
IF (EQUITY > (2.0 * MEMORY[3])) THEN MEMORY[3] = EQUITY
' PUT BOLD EQUITY INTO MEMORY[4]
MEMORY[4] = EQUITY - MEMORY[3]
IF (MEMORY[4] < 0) THEN MEMORY[4] = 0
' CALCULATE BETSIZE
MEMORY[5] = ((0.025 * MEMORY[3]) + (0.075 * MEMORY[4])) / SYSTEM
NEWCONTRACTS = MEMORY[5]


I fooled around with the vendor system called "Aberration"
when backtesting this idea. They sell it in the classified
ads of _Futures_ magazine among other places. You can look
it up on a web search engine if you like. [Question: will I
help you obtain a pirated copy of Aberration? Answer: no]

I ran Aberration on the 22 market portfolio, establishing
a baseline by using a Fixed Fractional betsize approach
which risked 2.5% of equity on every trade. From 1/1/1980
till 3/2/2001 this FF approach gave a compound annual growth
rate ("CAGR") of +70.46 percent per year, for over 20 years.
Each dollar of initial equity on 1/1/1980 had grown to
almost $67,000 by 3/2/2001.

Then I ran it again on the same portfolio for the same
time period, but using the TIMID + BOLD equity idea from
Van Tharp. CAGR increased to +96.84 percent per year.
Each dollar of initial equity had grown to $1.34 million
by 3/2/2001.

The neat thing is, that drawdown *OF* *TIMID* *EQUITY*
didn't increase. I am making the raw equity curves available
to you so that you may verify [or refute] this result if
you wish. Any question you may wish to ask, can be
answered by doing your own analyses of the equity
curves. Jump in, jack around, calculate some stats
and have some fun!

Equity curves and the raw data that generated them
may be freely downloaded from
http://www.mjohnson.com/trecipes/timidbold.xls

As you'll see from the equity curves, the Van Tharp
method generates quite a noticeable improvement

I hope you enjoyed reading this
sdzh
Contributor
Contributor
Posts: 2
Joined: Wed Jan 26, 2005 9:17 am
Location: The Steppes of Central Asia

Post by sdzh »

Wow, that's a swift response TK
Many thanks 8)
AFJ Garner
Roundtable Knight
Roundtable Knight
Posts: 2071
Joined: Fri Apr 25, 2003 3:33 pm
Location: London
Contact:

Timid/Bold Equity

Post by AFJ Garner »

As Steady Jake said, the start date is vital. So it is for any system. If you are lucky enough to start at a favourable moment for your system (assuming it is a half reasonable LTTF with a sensible bet size and portfolio), your build up of capital will be rapid - especially risking 7.5% of Bold Equity!

Even without the facility in Veritrader for an automated start date analysis, it is still useful and easy to run a batch of tests for a given system/portfolio at, say, successive 6 month intervals to give you some idea of the effect.

Somewhere or other I have a list of the 22 markets MJ said he traded at that time - presumably the same ones he used for this test. Interesting to see how the scheme fared at and around the past three years.

MJ does not state that he used any filter or profit taking for this test - so presumably 80/2 are the only parameters one needs to use.

In which case I suspect the max DD may be substantially higher if one includes the last 3 years than the 40% MJ has stated as the max DD he feels he can stomach.
Post Reply