The turtle sizing rules are flawed

Discussions about Money Management and Risk Control.
Post Reply
Trading Leech
Senior Member
Senior Member
Posts: 48
Joined: Mon Nov 29, 2010 2:07 pm

The turtle sizing rules are flawed

Post by Trading Leech » Wed Jul 13, 2011 2:13 pm

The turtle rules (as seen in e.g. Michael Covel's "The Complete TurtleTrader") for telling how many contracts to buy are these:
  1. You have $1,000,000 account, and you risk 2% of that on each trade, i.e. you risk $2,000 on each trade.
  2. You calculate N (the ATR) for the particular contract, and you want to risk 2N, i.e. your stop-loss is 2N.
  3. Therefore, the number of contracts you buy is $2,000/(2N).
However, this doesn't make any sense to me whatsoever. To demonstrate why, consider this extreme case scenario: A stock currently priced at $1,000,000 per share with N = ATR = $1.

According to the rules above, I should be buying 1,000 shares. And since the share price is $1,000,000, I should be buying $1 billion worth of shares!

This is ridiculous, because I may not even have $1 billion in my trading account to begin with. As such, the rules are just ridiculous since they don't properly take into account how much money I have in my account (my trading capital). Sure, the trading capital is looked at when calculating the risk of each trade, but not for the trade as a whole.

How can this issue be addressed? In other words: Do there exist any good sizing rules that actually take my whole capital into account, so my whole capital is "optimally" allocated?

Tim Arnold
Site Admin
Site Admin
Posts: 8500
Joined: Tue Apr 06, 2004 1:41 pm
Location: Boston, MA
Contact:

Post by Tim Arnold » Wed Jul 13, 2011 2:30 pm

Turtles traded futures, not stocks.

Trading Leech
Senior Member
Senior Member
Posts: 48
Joined: Mon Nov 29, 2010 2:07 pm

Post by Trading Leech » Wed Jul 13, 2011 2:58 pm

Tim Arnold wrote:Turtles traded futures, not stocks.
I know, but I just used stocks above for "simplicity" (because I'm more familiar with stocks). But it doesn't matter. My question holds no matter what the contract type is.

mojojojo
Roundtable Knight
Roundtable Knight
Posts: 131
Joined: Mon Oct 31, 2005 2:07 pm

Post by mojojojo » Wed Jul 13, 2011 3:35 pm

Trading Leech wrote:
Tim Arnold wrote:Turtles traded futures, not stocks.
I know, but I just used stocks above for "simplicity" (because I'm more familiar with stocks). But it doesn't matter. My question holds no matter what the contract type is.
I don't trade futures but I'm pretty sure that it does matter. Futures have leverage built in where as stocks do not.

Example. The gold future contract is for 100 oz so at say $1500 /oz, they contract has a value of $150,000 and I believe the initial margin is only around $6,000.

Try to buy 150K worth of stock with 6K worth of collateral.

AFJ Garner
Roundtable Knight
Roundtable Knight
Posts: 2040
Joined: Fri Apr 25, 2003 3:33 pm
Location: London
Contact:

Post by AFJ Garner » Wed Jul 13, 2011 4:33 pm

There has been a great deal of discussion of this topic on the forum. In particular somewhere or other I set out worked examples showing why you can't effectively position size on a fixed fractional basis using cash stocks. Looked at in another way, you could not benefit from price movements in 3 month interest rates without very significant leverage.

The use of leverage is at least twofold: to increase returns and to provide an effective method to size positions. Margin facilities on stocks don't begin to cover what you need.

riScalh
Full Member
Full Member
Posts: 13
Joined: Thu Apr 17, 2003 4:36 am

Re: The turtle sizing rules are flawed

Post by riScalh » Wed Jul 13, 2011 5:24 pm

Trading Leech wrote:
However, this doesn't make any sense to me whatsoever. To demonstrate why, consider this extreme case scenario: A stock currently priced at $1,000,000 per share with N = ATR = $1.
Your example is rather theoretical one because ATR = 1$ with price $1000 000 has never been seen in real prices fluctuations. It is 0.0001% in relation to the price. Assuming that daily standard deviation on avg market is approx. 1%, one can calculate 5min standard deviation as 1%/square root of 288 which is equal to 0.0589%. In your example it determines 5min ATR as $589.
It means that you can buy 3 shares trading on 5min bars risk. On higher time frames number of shares allowed for trading should decrease.

Your mathematical reservation - IMHO - is correct, but as these conditions do not exist in actual market environment it has no practical implications to trading.

Trading Leech
Senior Member
Senior Member
Posts: 48
Joined: Mon Nov 29, 2010 2:07 pm

Re: The turtle sizing rules are flawed

Post by Trading Leech » Wed Jul 13, 2011 7:43 pm

riScalh wrote:
Trading Leech wrote:
However, this doesn't make any sense to me whatsoever. To demonstrate why, consider this extreme case scenario: A stock currently priced at $1,000,000 per share with N = ATR = $1.
Your example is rather theoretical one because ATR = 1$ with price $1000 000 has never been seen in real prices fluctuations. It is 0.0001% in relation to the price. Assuming that daily standard deviation on avg market is approx. 1%, one can calculate 5min standard deviation as 1%/square root of 288 which is equal to 0.0589%. In your example it determines 5min ATR as $589.
It means that you can buy 3 shares trading on 5min bars risk. On higher time frames number of shares allowed for trading should decrease.

Your mathematical reservation - IMHO - is correct, but as these conditions do not exist in actual market environment it has no practical implications to trading.
Thanks for this input. You're right. In the meantime, I also recalled the important truth that a big part of trading is often just staying completely in cash while sitting on your hands and doing nothing -- if there are no good trends to follow.

So I suppose it does make sense after all. It's all about the risk. It's not about "optimally" using all of your money, because the game is not about using all of your money. Sometimes it's wise to just have cash.

Anyway, this was the only place I was going with my question. It appears that I was misunderstood by some. I wasn't talking about leverage, etc. but instead asking about utilizing one's whole trading capital vs. only looking at the risks of individual trades. Apologizes for my bad explanations.

Moto moto
Roundtable Knight
Roundtable Knight
Posts: 427
Joined: Mon Jun 01, 2009 4:12 am
Location: once again in the UK

Post by Moto moto » Wed Jul 13, 2011 10:50 pm

when looking at "utilizing one's whole trading capital vs. only looking at the risks of individual trades. "
you need also to take it to the next level for filtering instruments based on liquidity, a maximum position size per the portfolio size, correlated instruments and the like. these limitations might restrict your returns but they will also likely keep you in business.

sgbeamer
Contributing Member
Contributing Member
Posts: 5
Joined: Tue Mar 08, 2011 10:13 pm
Location: Houston, TX
Contact:

Leverage

Post by sgbeamer » Sat Jul 16, 2011 8:16 am

Leech,

The turtle sizing rules are for leveraged trading which stocks are not. The concept is to take the same amount of risk on each trade no matter what market because you don't know which one will be the big one. If you trade stock options you can use the turtle rules to size how many option positions you buy (leverage). For straight stock trading you really need a couple of rules

a risk limit ($ risked = entry price - stop price) AND
a maximum % of total equity to invest in that one stock (10%) - as in don't bet everything on one horse[/list]

Post Reply