Do markets change? Trend Following.

Discussion in 'Trading' started by AFJ Garner, Oct 15, 2009.

  1. I recently came across a statement on the web along the following lines:

    “Occasionally, someone trying to promote something or start a debate will argue that trend following rules must always change due to changing market conditions. This is nonsense. It is a specious argument.”

    Some well known trend following CTAs have indeed been quoted as saying that they still trade the same system as when they started out in business 20 or even 30 years ago.

    Well, yes and no. Let us take an example. Imagine that a CTA trades an ultra simple channel breakout reversal system: he buys when an instrument breaches an X day high to the upside, he reverses and goes short when the instrument penetrates an X day low on the downside. His position size for each trade is Y percent of total equity, based on the risk represented by the width of the channel.

    Now imagine that over the years, as markets change, the CTA increases the value of X in an attempt to avoid the increasing whipsaws and choppiness he sees in the market. He may also decide to increase the value of Y, perhaps proportionately, in order to keep a similar exposure to the market after the widening out of his stops. Let us also imagine that over the years, our CTA adds a bell and perhaps a whistle or two – perhaps he introduces sector and or overall risk limits, perhaps he introduces a way to reduce the risk from highly profitable trades, locking in some profit before the trend inevitable begins to reverse.

    Is the CTA still trading the same system at the end of the period as he was trading when he started out in his trend following career? Well yes, he is still trading a channel breakout reversal system. And no, the profile of his system is very, very different indeed from the way it started out.

    To emphasize the point, look at the many CTA disclosure documents out there. You are unlikely to find many which do not extol the benefits of continuing research and development. Of course they change their systems in the light of changing markets. Not to do so would be commercial suicide.
  2. (1) Here is an excerpt from the website of Abraham Trading Company:
    Research and Modifications
    Abraham Trading Company attributes its exceptional long-term performance to superior research methods.
    At Abraham Trading Company we believe we have simply come up with better techniques. Over the years the markets have become choppier, and many traders’ returns have suffered. We believe that continuous research is the only viable way to consistently come up with better ideas and better ways to profit in any market condition.

    (2) Here is an excerpt from the disclosure document of Tactical Investment management Corporation:
    The System is the result of over 27 years of ongoing research,
    evolving very slowly over the years as new discoveries have been implemented.

    (3) Here is an excerpt from the website of Campbell & Company:
    The firm’s Intellectual Property is protected and preserved by an independent research team. Enhancements to existing models must be vetted and approved through the same process as any new model being introduced. Campbell’s edge lies in its ability to evolve with changing market dynamics and conditions.

    (4) Here is an excerpt from the JW Henry disclosure document relating to the Global Diversified Portfolio:
    Beginning in October 2007, the Global Diversified Portfolio was modified as the result of extensive research. The modification involved the introduction of features designed to reduce the number of new trades that may be impacted by whipsaw markets, in which trends reverse before developing to the point where they generate profits for the program. Based on the research it has conducted on the effects of the modification had it been in place historically, JWH expects that this modification will reduce the number of false trading signals and preserve the ability of the program to trade profitably the trends that emerge in the marketplace.

    (5) Here is an excerpt form the Millburn Ridgefield disclosure document:
    Millburn Ridgefield Corporation is engaged in an ongoing research effort to improve its trading methods and to apply its quantitative analytic expertise to new financial products. Successful systematic futures trading depends primarily on two factors: development and selection of the trading systems used in each market, and allocation of portfolio risk among the markets available for trading.

    (6) Here is an excerpt from the disclosure document of the Dreiss Research Corporation:
    The trading strategies have been and will be enhanced or revised from time to time.
  3. Alexis


    Might become an interesting thread. i'll jump in if there is some debate
  4. bighog

    bighog Guest

    Think of trend following like the wheel. It still goes round and round.

    Gravity still works also. :)
  5. It seems like a lot of these people have something to learn since they have not fallen in line with the very authoritative and explite opinion first expressed.

    Below is today's market and it is evident that doing the X and Y thing seems to work.. The saucer ended and trending began. Count the break outs of higher highs. Or where they BO's or just a continuing trend that became more volatile as the day continued.


    Look at those five internal items (the "two bar yellows: in Audobon language) not in the index of the very authoritive author of your quote. Oh, volume is also on the chart too,; that is not in the index either. I know this is bush beating; but bushes are there and they get beaten.

    What if the X and Y strategy were enhanced a little more. Lets emphasize the anti whipsaw concept mentioned. See page 230 for the introduction of whipsaw. Note well on 231, the same authors quote:

    "Trendfollowers need thos home runs to pay for their whipsaw losses."

    I know the author has whipsaw losses. He needed an anti whipsaw gift so I gave it to him and asked him to not commercialize the gift. Here is the anti whipsaw gift: B2B 2R 2B. That is how the day started. After that the gift was R2R 2B 2R. This took 17 bars and the X and Y person had one trade from bar 15 to bar 17. But he did have a great trade from bar 61 to 69 inclusive.

    The anti whip saw given gift does not stop giving. as companies 1 through 6 could see if the author gave the gift he got to them. Not commerciallizing it forgoes that possibility. Retail traders could use it and pass it forward, however. That is what retail traders are doing.

    Make a score card and see how many B2B 2R 2B's there are in the day. Do the same for R2R 2B 2R. there are six of each; you know this by counting 12 moves and dividing by two. Or you count one side and know that the other side is equal. Or you draw RTLs and count 12 BO's and dived by two. Or you color the trendlines and count the balck ones and count the red ones and get 6 of each.

    That would be 12 linked trades making money on each. Say you wanted to trade the LTL and the RTL to make the money offered between them. If so then you just do 36 linked trades. This involves annotating. Annotating is not listed in the index either.

    Suppose you want to step it up a little and trade Outside Bars. (OB's) That is 6 more trades. Outside bars are not mentioned in the book's index either.

    Notice that they overlap beginning on the FTT and ending overlap on the former RTL. Failure to Traverse (FTT) and Right Trend Line (RTL) are not in the index either. The overlap is the B2 and R2 of the B2B and R2R, respectively.

    Lets take up the final opportunity to put the icing on anti whip saw. It name is Volatility Expansion (VE). VE is fun because it is like hitting a homerun. VE is not in the index either. What do you know about what happens @ VE? You know it is NOT an FTT. you know if it is after a point 3 (the middle term of the gift, 2R or 2B) that you need a new point 3 AND you get to ACCELERATE the RTL. Accelerate means make it steeper.

    One or more VE's in a row make it possible to throw in two more profitable trades per VE moment. Add the VE's the the chart's annotation.

    The gift given to the author is taking on some nice helpful aspects. Imagine people doing these things with respect to trendfollowing. The author of trendfollowing forgot to put any of these things in his book except whipsaws are where trendfollowers lose money.

    On page 96 the yopic of drawdowns is fleshed out with two examples 25% and 52%. Ask yourself:" if you drew RTL's and LTL's could you lose a percentage by trendfollowing within those lines. Why isn't there a listing in the index for trendlines in the book?

    I asked some of these questions of John Pierce and Mark Levine on 03NOV06.

    Final note. the author thinks trend followers do not get in at the bottom nor get out at the top. As shown in figure 10.4 on page 231, they do everything too late to make the market's offer. Don't look up "anticipation" in the index, it is not there either. What if a "trendfollower" knew volume leads price during the end effects of trends. Don't look up end effects in the book; it is not there either.
  6. .
  7. The difference lies in the fundamental concept one is trying to capture (via some quantitative model) versus the numerical details/representation of that model.

    Qualitatively, one can define a trend as a breakout with continuation. Quantitatively, the problem can take all sorts of forms and have a variety of answers.

    This may sound like circular logic, and, in a way it is. But, look at it this way, using an MA cross to capture trends doesn't work. Why? Because of the quantitative representation, i.e. the MA cross, not because trend following doesn't work. The fundamental concept is valid but the attempt at quantifying it is poor.

    The best systems are those which stem from a sound fundamental concept (the real "system" per se) and are properly combined with a valid execution model. The execution model has to change over time as markets do continually evolve, however, the fundamental concept remains the same.
  8. Here is an annotated chart to show the 12 patterns and the 12 trades or the 36 trades or the VE trades and OB trades. The scale is a 2 point scale and the bars are 5 minutes.

    for the original example change the X to FTT reversals or point reversals and use a Y that deal practically with doing adding of contracts from profits only. That is, it is reasonable that a person has the "right" to trade his profits; he earned the "right" by what is called "performance".

    The OP's Q is about performance in markets and does the strategy change with time? It is like in economics as Krugman says (06SEP09) there are freshwater and saltwater views. The superior view goes beyond either and it is based on skills and knowledge.

    CAPM was silly. Taking the market's offer under varying conditions can be done by Principle or by technological mechanics. Principle wins as may be seen.

    For those that log, you can see how the logging sheet can be filled in in advance for the day.

  9. I was asked to write an article on this topic by Active Traders magazine and have done so. Hopefully, this may do something to dispel some of the uninformed comments posted on the net.
  10. LOL.

    It won't stop the uninformed comments here.

    But fwiw, I too agree with the "trade the principle but adjust to the markets shifts" theory.
    #10     Oct 28, 2009