Skip to content

Parsimony

April 27, 2015

parsimony

Note: James Picerno of  The Capital Spectator recently did an interesting piece evaluating the Self-Similarity Metric and provides some R code which is valuable for many of our readers. 

The principle of parsimony relates to being frugal with resources such as money or the use of computing time. It is closely tied to the principles of simplicity, elegance and efficiency.  It also complements the philosophical theory of Occam’s Razor which states that the simplest explanation with the fewest assumptions is often closest to the truth.  Whether doing statistical modelling or building trading systems, it would be wise to respect the power of this principle. Parsimonious models or trading systems are often robust, while overly complex models with too many assumptions are not. The difficulty is in telling the difference- which is not obvious even to a talented and experienced developer.  The ability to distinguish between parsimony and excess complexity is virtually invisible to almost everyone else.

The backtest is the problem and great distractor in the quest for parsimony. It is like a picture of a beautiful woman that is scantily clothed beside a paragraph of important text– no one is interested in the fine print. A beautiful backtest is admittedly just as satisfying to look at (perhaps even more so for quants!) and can blind us from the details that got us to the end point. And while we all appreciate some good “chart porn”, there are some important questions to consider: What universe did we select and why? Why did we omit certain assets or choose certain parameters over others? Why did we choose one indicator or variable over another-and how do we know it is superior? Why do we trade at a certain rebalancing frequency versus another and is this relevant to the model? Most importantly is: Can I create a trading system with similar results with far fewer assumptions and with less computational power? That should be your goal- to achieve the maximum results with the least number of assumptions and resource usage.

For example, I am well aware than the Minimum Correlation Algorithm does not mathematically optimize the correlation matrix or find the most diversified portfolio. The Minimum Variance Algorithm does not minimize variance either relative to a true MVP solution. But they both use an intuitive and simple method that meets or often exceeds the results of the more complex solutions with less resources, and hence can be considered parsimonious. They are also less dependent on estimates for optimization inputs. Such  systems are more likely to work in the uncertain and messy world that we actually live in. Cooking is a hobby of mine, and more recently I have strived to achieve the most with the least, and ensuring that all of my marginal choices of ingredients or differences in traditional technique are actually adding value. There is no point sounding fancy by adding exotic ingredients or using fancy techniques if they don’t change the taste for the better. These give the illusion of expertise to the unsophisticated, but to top chefs judging these dishes on FoodTV they only serve to highlight their deficiencies as cooks. My advice is to work with things that you can understand or intuitively grasp and be very careful when trying newer and more complex methodologies. Master what you can with the tools you have at your disposal instead of reaching for latest and greatest new toy. This may sound strange coming from a blog that was built around offering new ideas and concepts-  but rest assured this is some of the best advice you will ever receive.

All of the questions  I posed above relating to trading systems are quite material, and many cannot be answered quantitatively. Unfortunately for the quantitatively inclined, the principles of good logic often get lost while decoding proofs, cleaning data, or debugging computer code. Furthermore, the elegance of complex math is like comfort food for those that are highly intelligent and it is easy to forget that the assumptions of these models are a far cry from describing reality. Even for the more experienced developers that are aware of these problems, they may arrive at the wrong approach to system development. The solution is not to avoid making ANY decisions or assumptions (although relying less on specific parameters or universes is desirable for example), but rather to make sensible choices with few assumptions.  Another alternative is to build a methodology that directly makes choices quantitatively to create a parsimonious model. Both methods have their strengths and weaknesses. 

At the end of the day, there is no point making something more complicated than it needs to be unless the benefits are material. The same is true for the length of time/complexity of the run for the computer program that runs the trading. My brother is a professional hiker and has traversed extreme mountain terrain. Unlike most amateurs, he does not pack everything under the sun that might be useful for his trip. Instead he focuses only on the essentials and on minimizing weight. More importantly, he focuses on planning for what can go wrong and makes his choice of gear and specific hiking route accordingly. The black and white realities of survival bring these questions to the forefront. In contrast the more comfortable and forgiving world of offices and computers make trading system decisions seem almost like a video game. Rest assured, it is not…..

 

 

Advertisements
9 Comments leave one →
  1. GeraldM permalink
    April 27, 2015 5:36 am

    David,
    This is a wonderful article and a good reminder for quants who are getting just a little to cute with their algorithms (highly focused universe, cherry picked time frames, techniques that only work on a specific historical regimen, questionable assumptions, etc.). The backrest is not reality anymore than the future is the past. Yes, it can be helpful in forming an opinion, but too many purveyors of backtested systems sell them as doctrine. As you said, it can be a distractor. I don’t know what specifically inspired you to write this opinion piece but I’m glad it did. Thanks!

    • david varadi permalink*
      May 1, 2015 1:38 am

      thank you Gerald, i wanted to emphasize that it is good to explore a lot of ideas, but ultimately a final model must use the ones that have the most impact and are synergistic.
      best
      david

  2. Bob permalink
    April 27, 2015 4:22 pm

    Good stuff but you are not answering the questions about the discrepancies in your self-similarity metric and the Picerno’s results do not check. Please if you can answer questions by Omar in the second post. This will help a lot.

    • david varadi permalink*
      May 1, 2015 1:40 am

      Bob, thank you, not sure what you are referring to since several people have reconciled the self-similarity results. i believe the questions in the second post were addressed– the guest author simply transformed the indicator to center around zero and reversed direction.
      best
      david

  3. Ryan Turner permalink
    April 30, 2015 11:19 am

    I agree, but find it somewhat ironic that you have about 50 variations of momentum described on this blog.

    • david varadi permalink*
      May 1, 2015 1:42 am

      hi Ryan, i did mention that point on my post—in order to emphasize that while it is good to explore new ideas (and small details/variations can make a difference) you need to ensure you aren’t extracting ideas that don’t provide marginal value or create extra layers of complexity in a final model or trading system.

      best
      david

  4. July 30, 2015 11:23 pm

    Hi David,

    To extend the hiking metaphor, I would suggest a few tests on non US markets… What if the US market (SPY) some day soon behaved like Taiwan (EWT) or Japan (EWJ)? Would the various momentum models (backpack & gear variations) developed to date still ensure a successful TAA trip? Or would using the models to date be taking mountain camping gear to the bayou?

    And if the models don’t cross over to other national markets, is there a “training” period that would ensure the right camping gear gets packed; i.e., if we train the models in the bayou will we end up with different parameters that in fact work in the bayou?

    So, when hiking, is planning the variations and optimizing on one trail important, or is it more important to understand what is needed to prepare for a different style of trail?

    I’m curious how the models work as is on different markets, and if training periods would make a difference.

    Your thoughts?
    Carl

  5. david varadi permalink*
    August 2, 2015 3:48 am

    hi Carl, your question is a good one– first I already did studies on Japan which is important and momentum would have definitely improved results–however absolute results are always a function of how buy and hold does, so in absolute returns they were obviously not great. but there is no question that certain conditions are more or less beneficial to certain types of strategies. furthermore there are periods where all strategies will underperform even if they are robust. in a perfect world one can adapt or try to adapt to the hike, but that carries with it the new problem of demonstrating the ability to adapt online. ultimately it is both important to optimize for the trail and also know what is robust across trails even if you are not prepared.

    best
    david

Trackbacks

  1. Quantocracy's Daily Wrap for 04/27/2015 | Quantocracy

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: