Systematic Trading research and development, with a flavour of Trend Following
Au.Tra.Sy blog – Automated trading System header image 2

Were the Turtles just lucky?…

March 8th, 2010 · 23 Comments · Strategies, Trend Following

turtle-trader-dennis-eckhardt-pandiyan
 
Most people will have heard of the mythical Turle Traders, a group of novice traders set up and mentored by legendary “Prince of the Pit” Richard Dennis.

Dennis did so to set up an old argument with fellow trader Bill Eckhardt on whether trading could be taught or not (not unlike the story in classic movie Trading Places).

The experience was succesful in proving Dennis right (trading could be taught), with his turtle traders making him $100 million.

Why turtles?

The Turtles were nicknamed as such because of an analogy with how Richard Dennis expected to “grow” traders in the same way Singapore farms grow turtles. Dennis taught his students a mechanical Trend Following system and let them trade with his own capital. After being kept secret for more than a decade, the rules were revealed and floated on the internet for a while. Two enjoyable books have now been published on the topic (Complete Turtle Trader – featuring the actual turtle rules and The Way of the Turtle written by Curtis Faith, a former Turtle) if you are interested in learning more about it.

The turtle system

The Turtle system did not contain “magical” components. It was basically a combination of 2 different breakout systems with specific rules for money management, including position sizing, pyramiding, correlation limits and cutting down position size during drawdowns (a quick “Turtle trading rules” google search should yield some results for the exact rules).

System kaput?

Now that the rules have been made public, it is possible to backtest them and see how they would have performed on the recent markets. Such test result can be found on the Trading Blox forum.

click to zoom in

click to zoom in

It basically shows that the CAGR drops from 216% (!!) from 1970 to 1986 (when Dennis and Eckhardt were developing the system, and also when the students traded the system with real money) to barely double digits (10.5%) in the last 23 years (1986 to 2009), with a completely flat period from 1996 to 2009.

As a side note check the crazy amount that compounding generates at that rate of 216%!

One could argue that Dennis & co were just lucky to trade the system during what appears to be its golden period.

System overheating warning

During the Turtle experiment, Dennis came to the realisation that their position sizing rules were such that:

you have been trading as much as twice as big as we thought

Here is a snapshot of the memo that Dennis sent to all traders asking them to cut their position size in half.
Dennis-Memo
As Dennis said:

We must be living right

Another way of saying “We have been very lucky”?…

What does this mean?

Well, some say that the Turtle performance was a fluke – that the Turtles were actually the proverbial monkeys writing Hamlet (see the Infinite Monkey Theorem). I guess these people would be in the EMH camp (Efficient Market Hypothesis).

Some say that Trend Following is dying/dead and the Turtle system under-performance is an illustration of that. Problem is: these people seem to celebrate the demise of “simpleton” Trend Following strategy every so often (during major drawdown peiods), only for Trend Following to come back roaring again (think 2008).

Some might also say that market conditions are changing, and systems need to adapt to these changing conditions.

Although, some also say that this argument is specious, citing Bill Dunn as an example of a CTA claiming to use the same rules as when he started in the 70′s.

Reconciling it all

Instead of concluding this post quickly here – on what it does or does not means – I thought it might be more interesting to expand and spend more time on the possible interpretations above in a post of its own. Part 2 will take this discussion further.

Stay tuned, I will try to touch on whether or not Trend Following should, and can, adapt to changing market conditions – update: post here.

Related Posts with Thumbnails

Tags: ····

23 Comments so far ↓

  • whatever

    umm, firstly Dennis and Eckhardt ran the Turtle experiment to prove whether traders could be taught or whether they are born – the actual system they used to trade is completely irrelevant..

    Having said that, they obviously used a set of rules that they thought would be successful (or had proved to themselves was successful) for the type of market they were trading at the time.. Obviously if such an experiment were run today the system would have to be appropriate for the way markets behave today..

  • Suhendro

    This is very interesting discussion. I just finished read Covel’s Trend Following and The Complete Turtle Trader and found that it is possible to trading for a true living trhough trend following. Well, I’m happy to say that I will wait for another article you write and learn from it as well.

    Thanks!

  • Jez

    Hi “whatever” – thanks for popping by…
    Agree with the debate between Eckhardt and Dennis as the main point for the Turtle experiment. The angle I wanted to approach this was to look at the system rather than the experiment itself – and see how a Trend Following system performance can break down.
    I agree: I could have chosen any similar system to illustrate the fact that performance has “broken down” when comparing recent history to the 70′s (heck, even a random entry system used to perform much better than it does now – when applying a sound money management algorithm).
    Discussing the Turtle system makes it quite interesting because of the mytical apsect of it (ie real trading system that was traded by a Trend Following wizard – Rich Dennis – making him and his students milliions of dollars)…
    This is really only an introduction to an “endless” debate: “Do markets change and should/can your (TF) system adapt to these changes?”
    Cheers,
    Jez

  • Jez

    Thanks Suhendro
    These books by Covel are good for some insights on Trend Followers and great motivational tools… It gets a bit more “difficult” when you start delving in the details but hopefully this blog can help on that side (that’s the goal: look at trend following “under the hood”)

  • Sam

    I have not run into the trend following is dead crowd myself. Sounds to me they may be a little ignorant of how trading actually works.

    By nature trading models will go through periods of ups and downs, unless you’ve got a perfect arb. Given that trend following is long term draw down periods may also be longer than other short duration models.

    I’d be curious to know where you ran upon this crowd that believes if something is not working it is dead? Constructive criticism here so don’t take it personally, but perhaps you should distance yourself from those who make such silly claims about how markets work. It sounds to be to not only be an arrogant statement but also an uniformed statement. One cannot say if something is dead or not until it actually is. This is especially true of trading models.

    It should also be noted that I am not much of a trend follower so I am not biased to it one way or the other, but I do have a pretty firm understanding of speculative trading models.

  • Jez

    Sam – don’t worry. That crowd doesnt affects my thinking too much – maybe just for some healthy skepticism. Have you come across Vic Niederhoffer – he would be one of them.. I think basically all the traders that do not believe in trend following for its lack of “complexity” and just celebrate after years of bad performance – or even Trend Followers self-doubters.

  • Rajiv Vyas

    At the end of the day, there is some randomness (luck) in every skill you have. I can pretty confidently say that between three systems I follow, one would do very well over the next 5 years. What I can’t tell is which one. Now assume there are three hedge funds or CTAs, and each one adopts one of the three systems. One of them is likely to be lucky. Whether you are following a fundamental screen or a technical screen, there will always be a choice that you will have to make. And usually the choice you make will be influenced by your biases and preconditioning. Hence, the luck is always there. What we do by fundamental and technical screens is try and simplify/reduce our decision making and in doing that, we are trying to reduce noise/randomness and improve our chances of winning. In the above case, we now have a chance of delivering above average returns of 33% vs say 5%.

  • Lee

    My general feel of trend following systems is that the shorter-term ones have been losing their edge, but the longer-term ones are still holding their ground. It would be interesting if you could post up a chart comparing the turtle equity curve and an equity curve based on a simple 12 month momentum that AQR talks about in this paper.

    http://www.aqr.com/Research/UnderstandingManagedFutures.pdf

  • Jez

    Hi Lee, Thanks – this is a very valid point and in the follow-up article I will show how the Turtle system can be “fixed” by making it go longer term. This is a similar idea to a study I saw using random entries and “Trend Following” money management rules (ie let winners run and cut losses short). These sort of systems have broken down to, indicating that there might well be more “noise”, which needs to be filtered by looking at a longer term timeframe.
    Hopefully, this will be shown by the State of Trend Following report, looking at different timeframes (note that I am planning on including the AQR benchmark strategy to the State of TF report).

  • Jez

    @Rajiv,
    Sorry – misssed your comment.
    I completely agree that chance plays a role in everything (mostly) and the game is just trying to stack the odds in your favour as you explain…

    Unless you have a perfect crystal ball (let me know), trading in the markets you’ll always be subject to the random impacts of Lady Luck!

  • NordicTrader

    Hi. I have been trading self-made automated systems since early 2009. I trade only stocks, since I’m undercapitalized for FX and bonds. I have find vital diversification in different time frames and opposed strategies. I have recently started to re-think risk, and I have a question on that topic.

    I risk 2%/equity per trade. Expected drawdown on single stock is modest 10% of the capital. Nonetheless, I trade baskets of 30 stocks per system, and I have four different systems. So I actually trade 120 stocks with a potential 10%DD in all of them.

    Since the strategies are opposed (two mean-reversion vs. two trend-.following) and the timescales are different (two weekly vs. two daily) and some strategies have relatively rarely open positions, I have thought that my risk is acceptable. But is? Opinions, arguments?

    How do you calculate risk for strategy, which trades 30 stocks, spends 5% of time in the market, and expects a drawdown of 10%?

    Many thanks for the best posts in the blogosphere and all the best.

    -NordicTrader-

  • Jez

    Hi NordicTrader – thanks for the kind words!
    I think Risk and Money Management rules are one of the most important part of the system to keep you “alive”. And one thing you should always try to keep in mind is the eventual Black Swan that will show up and how you would survive it. Now, I am not sure I understand exactly how much you are risking per stock (2% or 10%), but assuming it is 2% you can see how having 120 positions open at once would risk 240% of your equity which is not acceptable (think Black Swan) – so I would suggest you might implement some safeguards such as a maximum portfolio heat (ie total risk on all open positions not more than X% of equity) and work within these limits (ie refuse new positions that would make you go over the thresholds or sell down existing positions to “make room” for the new ones. You might also want to look at things such as correlation although this is a double-edge sword: ie it can change quickly.

    In any case I really recommend that you play around with your rules in a simulation to try and understand how they affect your systems.

    Since I was discussing the Turtle rules, you could maybe get inspiration from them on Money/Risk management: they had different maximum risks based on sector/correlated markets and total positions
    -Jez-
    PS: OANDA might be a good platform for you to look at FX trading: they do not have position size minimums.

  • weirdenglishwriter

    No one could predict top and bottom, TF just follow trend to top and bottom with a trailing stop.
    Trend followers only make money when market in bullish and bearish run. In super bull run and super bear run, TF make super profit![ eg 2007,2008 and 2009.]
    But TF endure losses when market flat.
    Trend following will surely make money in very long term, when market will surely move up and down greatly.
    The only problem is how to cope with the emotional stress when traders lose money over the flat market which may last for years!

  • den

    The chart you use is logarithmic, not linear, so backtesting results seem flatter, than they really are. Trend following is not dead. Even if we use counter trading strategy for example mean reversal system we are looking for a coming trend. Though i think that now trading is more accelerated and trends are shorter than they used to be.

  • Jez Liberty

    Den – the point of the chart was not to establish whether trend following is dead or not… And if you read few more posts from the blog you’ll realise that I do believe TF is well alive!

    Ps: log charts are the best way to look at a track record, especially from a long-term perspective, when compounding returns. In a linear scale chart the beginning of the track record appears usually all flat because of the exponential effect of componding (a concept usually abused by people using “hockey stick” charts to dramatize growth of anything, from world population to pollution, etc.).

    The log scale chart removes this exponential effect, allowing for a better comparison of CAGR (which is simply proportional to the slope of the equity curve).
    a log chart

  • Vince

    Hi Jez, thanks very much for your dedication to trend following. At the risk of over simplifying, I really feel as though deciding what kind of trends you want to trade is crucial with regards to system development as they come in a variety of shapes and sizes. Odd as it may sound, It occupies a lot of my time. I have come to understand and appreciate the effects that different trend speeds and durations have on a trading model. A bit of thinking aloud really… Interested to hear your thoughts when you have the time.

  • Jez Liberty

    Vince – I agree with you, there is no “one-size-fits-all” for trends, which is why my thinkings are usually along the line of trading different systems for different trend timeframes in order to gain from the diversification that this can provide.

  • afamiii

    i) The experiment proved that trading can be taught.

    ii) The back testing you did provides evidence to support the theory that the more a system is widely known the poorer will be its results. No doubt this is why Richard Dennis kept it secret. As did Warren Buffet for the first half of his career.

    iii) The real test of a great trader is not that he can follow a system given to her. But that she can develop a system that works, use it whilst its working, realise when it stops working and stop using it and develop a new one that works. Or even better multiple systems that you can use at the same time. This is what Warren Buffet and Richard Dennis both did and a number of the Turtle graduates.

  • Trend Trader Jim

    “But that she can develop a system that works, use it whilst its working, realise when it stops working and stop using it and develop a new one that works.”

    Ummm… no. If you develop a relevant and robust trading system that actually works you don’t have to keep stopping and “develop a new one that works”. What you are describing is the beginner’s cycle of chasing the holy grail.

  • Jon

    There are a couple of things that I see are in error in this article:

    1) The system results that you provide in the article from tradingblox: Neither the original OP of the thread or the one who posted the graph made any claims that the algorithm tested followed the original turtle rules laid out in the pdf by Curtis Faith or any of the other turtle traders. And no code was posted, which wouldn’t have been a problem since the rules were already public anyway. Curtis Faith complete turtle rules pdf: http://www.metastocktools.com/downloads/turtlerules.pdf

    2) the internal memo from Richard Dennis to his followers; why was this even mentioned? The sentence immediately preceding “….We must be living right…” was “…The good news is that this has been true throughout the whole trading program–your profits were doubled, but at the cost of a doubling of the risk….”

    This along with the rest of the memo simply suggested that the position sizes currently in use were too large for market conditions, and that they should be reduced. It didn’t appear invalidate the underlying strategy in use.

    I also wanted to comment on what another comment, afamiii mentioned here #comment-2474 “”

    The article largely assumes that the rules were backtested exactly as it would have been traded mechanically. There is not enough evidence to prove the original rules were used. I’ve seen this done many times about the turtles or any other system that is being criticized. They will often show a bastard version of it and present the bastard as the original.

    But many vendors have come along using the turtle name and fame, including Michael Covey, to promote their improvements and turtle methods. In all of the vendors i’ve seen, their improvements are always in the form of a blackbox or some non-transparent or discretionary method. So you pay thousands of dollars for the ‘idea’ that you may be profitable some day.

  • Jez Liberty

    Jon,
    I wrote this article a while ago but looking back on it, it seems that the main points I put forward were:
    1- The performance of the Turtle trading system (and similar systems) had drastically reduced.
    2- The turtles were lucky to be trading at double the heat they should have been during a profitable period.

    Re: point 1, I’m sure you’ll agree that, whether a system is exactly as per the turtle system rules or simply similar, it will exhibit a similar performance curve as the one displayed in the post. Turtle trading is not as profitable as it used to be. Period. And I do not think anybody would argue that point.

    Re: point 2, the memo shows that – by their own admissions – the turtles were “trading twice as big as they thought” because they had “misconstrued the theoretical data”. The fact that this happened during a profitable trend following period inflated their returns (lucky outcome I’m sure you’ll admit) and even better did not trigger a blow-up that could have occurred, had the markets not been favorable (ask any CTA or trend follower how they’d feel right now if they realised they’d run their program at twice the heat by mistake). Avoiding a blow-up in this context: quite a lucky outcome as well.

    I believe the comment: “We must be living right” is actually a way of saying that they have been lucky: they made a mistake, took twice as much risk as they should have, yet did not get punished by the market but rather the opposite. I’m pretty sure a few people would qualify this as “lucky”, hence the relevance of the memo to the article.

  • Jon

    I have no idea whether the turtle system is still profitable today mainly because I do not trade it myself. But I do use mechanical trading systems extensively. To imply that something mechanical may or may not be as good performance wise, without looking at the fundamental inputs (the system itself in this case) would mean that we are relying on consensus in order to declare something as being valid or not. This is more like trading as a religion: http://edgesense.net/trading-as-religion-part-1-traders/

    In regards to the second point about being lucky, the risk may have been up to double what they originally planned for, but so would the drawdowns also as the memo stated. The memo stated that to reduce the risk, the position sizes needed to be scaled down 50%, not to rewrite the underlying system. It is unlikely that it was pure chance that lead to their success, especially considering that (according to the pdf I mentioned), they usually traded multiple liquid instruments at any given time.

    Do you recommend or propose a mechanical trading alternative that is not based on what you consider luck?

  • Jez Liberty

    Hey Jon,
    Thanks for that article, it’s an interesting read.

    Maybe the “tongue-in-cheek” title and tone of this post was not apparent to you, but I’m sure that if you check a few more articles on the blog, you’ll see that it is not representative of the overall ideas expressed here. Indeed I even track and try to reproduce performance from ex-Turtle traders… Not something a trader thinking that “it was pure chance that lead to their success” would do…
    To be sure: of course their system was sound and I do not believe luck is the main factor in their success (and maybe I did not express myself very well in the article or you’re taking the headline title a bit too literally, but this is not the point I was making here!…). Sure, I was arguing that there was some luck involved in the markets not “punishing” them for their risk-underestimation mistake. Also that they were lucky to able to trade in such a profitable period for short-term trend following (compared to trend followers trading now for example).

    Just a bit of attention-grabbing headline using a famous successful trading case to basically open up the discussion on the fact that trend following systems might have to change and adapt to conditions, as presented in the follow-up article: http://www.automated-trading-system.com/intricacies-of-market-and-trend-following-changes/ (which discussed a good article by Anthony Garner on tuning the Turtle system).

Leave a Comment