Do the original turtle trading rules still work in the modern trading world?
They certainly did in the 1980s!
In fact, the rules made millionaires out of normal folks with very little starting capital.
Since these rules were some of the very first ever used, it's important to take a closer look.
Let's see if they still work, how to improve them, and what the modern computer code looks like.
Who Were The Original Turtle Traders?
Conducting interviews of over 1000 candidates, they whittled down the list to just 10 lucky people.
They were called the "Turtle Traders".
Dennis wanted his new traders to only trade commodities such as gold.
Then he gave them exact trading rules and a chunk of his own money to trade with.
The results were amazing.
Several turtle traders made triple-digit returns in just a few years and some went on to start their own hedge funds.
Dennis’s experiment seemed to demonstrate that traders could be taught a relatively simple set of rules with little or no experience and become excellent traders.
*Fun fact, the movie "Trading Places" is loosely based on the story of these newly minted traders.
Original Turtle Trading Rules: What Were They?
As you know, I am a 100% rules-based trader.
Writing down exact rules lets you back-test them over past data to determine if they make or lose money.
But you can't do this by hand like most newbie traders do, because your human biases take over.
You end up cherry pick examples that look like they fit your model while ignoring all the negative examples.
Programing your rules into a computer is the solution to this problem.
The computer is a cold, hard number cruncher and will tell you if you're onto something, or if your trading ideas are a steaming pile of garbage.
So, what were the rules that Dennis shared with his traders?
Dennis's rules are called "breakout systems".
Rules that look to buy big moves higher above a past number of days and rules which sell after a breakdown in price over a certain number of days.
(Vice versa for shorting markets).
In truth, these are some of the simplest strategies known to traders.
Let's look at the original turtle trading rules:
/*Buy gold when this happens*/ If Close >= highest close in 20 days then Buy /*Sell gold when this happens*/ If Close <= lowest close in 10 days then Sell /*Always use a protective stop loss*/ If Close <= 2 ATR from buy price then Sell Stop
Simple is robust and hard to curve fit to past data.
Beware of trading rules that are overly complex as they will likely not work in the future.
Do The Original Turtle Trading Rules Still Work?
OK, so let's program these original turtle trading rules into a computer and test them on a commodity like Gold (using the futures contracts here @GC):
Meh, not very exciting.
While they are still profitable, you wouldn't want to trade them.
With only 38% profitable trades, these rules would drive you bananas.
You would feel as if the rules had stopped working.
It's only a few select trades that win big - you would be constantly taking small losses, waiting for those big trends to come along.
It would be demoralizing.
Tweaking The Original Turtle Trading Rules
But what if we tweak these rules some?
Markets do shift over time, which is why it's so important to occasionally reoptimize a strategy's parameters, keeping at least 25% out-of-sample data in reserve to test the new rules on.
This is called back-testing with out-of-sample data.
I've written about this back-testing process here, which I highly suggest you brush up on.
I typically reoptimize my strategies once a year, around the holidays when I have a little downtime.
So, let's take the original turtle trading rules and optimize them over the last 20 years.
Also, let's add a way to measure the trend and only trade in that direction (super important).
This is also a super crucial topic which I've discussed in this article on the importance of trend detection.
The Modern Turtle Trading Rules That Work
By adding in a trend filter and ever so slightly changing the number of days to buy breakouts of price and breakdowns in the price we can turn the original turtle trading rules into a surprisingly simple and profitable trading strategy.
Here are the new, modern rules that work best:
/*Buy gold when this happens*/ If moving_average_1 > moving_average_2 and moving_average_2 > moving_average_3 and Close >= highest close in 12 days then: Buy /*Sell gold when this happens*/ If Close <= lowest close in 5 days then: Sell /*Always use a protective stop loss*/ If Close <= 10 ATR from buy price then: Sell Stop
As you can see, I added a trend filter and only slightly changed the number of breakout days and breakdown days.
The trend filter is identical to the filter used in my QQQ trading strategy.
But let's examine these rules graphically and look at the trading statistics.
I used one subset of data as in-sample to make the system and held back the out-of-sample data to test the rules with.
The Modern Turtle Trading Rules
You can immediately see the difference and these new rules are quite tradable if you want to start trading Gold with them.
Any profit factor above 2 is fantastic and this strategy comes in at 2.8!
But you will still be taking lots of little losses as you trade it with only about 54% of your trades being profitable.
This can be hard for most traders.
Most new traders want a system that wins all the time - which when you think more about it, you know deep down is impossible.
Yet I see countless traders hop from one strategy or stock-picking service to another thinking that one is going to suddenly make them rich.
But as the strategy above shows, you must have patience.
Wealth creation is a process, not a one-time event.
Conclusion On The Original Turtle Trading Rules
As you can see, the original Turtle trading rules do still work.
And in fact, it's very impressive that they work at all in the modern world.
Markets can shift drastically over time, but these original rules show that some commodities, like gold, are still Turtle trading friendly.
Nonetheless, it's the slightly tweaked modern rules above that make for a truly tradeable strategy.