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

100th post and 1-year Anniversary

September 13th, 2010 · 9 Comments · Blog

anniversary-cake

Last week marked the first anniversary of the blog, and coincidentally the 100th post as well. Last September, the blog was born (does that make it a birthday or an anniversary?…) and I have not really looked back since.

The first thing I want to do is to thank all of you, readers. I have had genuinely interesting conversations with some of you, whether on or off the blog – and even met some of you in real-life. This has definitely played a role in pushing some of my thinkings further, or simply giving me new ideas/concepts to work on. Writing these articles/posts regularly is also an excellent driver for my research.

In that respect I consider the blogging experience as successful and I hope that it has been mutual, and brought you interesting points of view for your own trading.

I cannot fail to mention all the other authors in this niche of the blogosphere, which have also been an important part of this idea generation/exchange process. The ones I read regularly are on the blogroll but I want to single out my main sources of inspiration for starting this blog:

If you’re reading this: a special thanks guys!

Trend Following and Motor Racing

I find it interesting to draw parallels between trading and real-life, especially with things I am interested in (as in Surfing and Trend Following).

And to “celebrate” the blog birthday I treated myself to some track time (I feel passionate about trading, but I agree with Mike on not wasting a good life trading — except I don’t do MMA fights, I prefer having fun with a bit of racing).

And I could see such parallel between some racing line theory and some marketing itterature from BlueTrend, that I was reading not long ago… Bear with me:

This is not really a motor racing blog so I’ll keep the explanation short; but in effect most beginner drivers tend to go into a turn (too) early because it feels safer: It gives an impression of more space and time. However, there is a clear benefit in turning in later – as you can see from the diagram below. The exit line is straighter and allows to get a better “drive” out of the turn. Another very important aspect to make this work is to be very smooth and progressive with the throttle (for bike stability, holding the line, etc.).

Racing line: get in later to drive in faster

Racing line: get in later to drive in faster

One of the main marketing points in BlueTrend’s brochure, describes their strategy as more progressive (ie gradually building positions), which allows them to capture only the bulk of the trend, by getting on the trend later (and avoid some early whipsaw action).

trend-belly

I find the concept very similar to wanting to turn in earlier on the track, and keeping a smooth, progressive throttle.

This is merely an extract from BlueTrend marketing brochure, but is a concept worth investigating. My gut feel is that the main benefit would lie in producing a smoother equity curve, rather than an improved raw performance. Note that this is not a new concept: the “Turtles” were also adding/pyramiding to their positions.

With regards to the exits, BlueTrend try to anticipate reversals and lighten up their positions ahead of them, not much “cornering” parallel I could find there.

And to conclude a few pics of yours truly, trying to turn in late…

2WW_9956 2WW_9431 RPS_8251
Related Posts with Thumbnails

Tags: ·

9 Comments so far ↓

  • Troy S

    Congratulations, and thanks to you too! Your posts are always interesting and educational, and I only wish I was half as knowledgeable as you so I could contribute more :) Keep it up!

    I learned about apex turning from playing video games, but never thought about it the context of trend following. All you quant bloggers seem to have such fun lives outside of trading!

    Small clarification, BlueTrend is a fund of BlueCrest Capital Management?

  • cordura21

    Happy anniversary Jezz. It’s been fun to read your blog, it reminds me of Bill Rempel’s and Gummy’s site, although I hope yours won’t disappear. I like your humble approach to this issues, which is itself a fat tail event within people who has these interests. Keep it up.
    Cheers, Cord

  • Jez Liberty

    Thanks guys for the nice words!

    @Troy: yep, BlueTrend is the (huge – $10B) Trend Following fund of BlueCrest Capital – good short article about them here
    ps: I started learning about racing lines with video games too – not allowed anything with an engine when I was a kid. trying to catch up a bit now ;-)

  • QuantingDutchman

    Congrats Jezz!

    Keep up the good work, always enjoying and learning from your posts.

    Michel

  • Pumpernickel

    Good luck searching for “scaling in” (also called “pyramiding”) rules that improve performance compared to full-positionsize-entries. The reason there are so few publicly known scaling-in systems, is also the reason why you cited the Turtle system as an example of scaling-in even though the Turtle Experiment happened 25 years ago. The truth is stark and ugly: it’s damned difficult to find positive-alpha when flailing about in the scaling-in sandbox. Not impossible, mind you; merely damned difficult. The mind of man has cracked the nut, with Blue Trend as a witness. But it’s not the sort of very well known, dilute wisdom that you find in SSRN papers or $65 books on Amazon.com.

    Scaling out (exiting a position piecemeal), by contrast, IS very well known, and can be found in speeches by Chesapeake Capital managers, books galore, freebie systems on your beloved Blox Forum, and countless other places. Because it’s relatively easy to find scaling out rules that handily beat all-at-once exits, such rules HAVE been found by dimwits, halfwits, and nitwits (among others), and these exalted persons have publicized their rulesets hither and yon. One needs but look.

    The downside of scaling-in or scaling-out is seldom mentioned. You can’t scale-in (or scale-out) N times, unless all your position sizes are at least N contracts or greater. It takes a big account (or an undiversified account) to scale-in or to scale-out. Even $5M managed accounts have a hard time doing so, if they trade >75 market portfolios. Eww.

  • Jez Liberty

    Hey Pumpernickel – thanks for these insights.
    To be honest, I am wondering how much of this is “marketing spiel” – i.e. with such large AUM, they must surely be running systems on multiple timeframes simultaneously, which would actually give the effect of scaling in when a large trends occur (ie if you have simple 20, 40, 60 and 100-day Donchian systems, these systems would gradually get on board the trend one after another). That timeframe diversification would smoothen the equity curve. They might simply market this as their proprietary “trend-only capture” algorithm…
    Of course account size is a constraint (as per your comment – but I assume BlueCrest are sorted on that end) – which is why I want to start looking as ETFs as an alternative to futures (bearing in mind they have downsides of their own)…

  • Michael (MarketSci)

    Hello Jez – humbled by the mention – thank you sir. Here’s to a rocking second year for Au.Tra.Sy!

    P.S. loving the racing pics…sweet. I used to race dirtbikes in my teenage years. Never got into the crotch rockets though. Something about hitting concrete at mach 2 that didn’t appeal to me =) Good on you sir.

    michael

  • Pumpernickel

    Remember that Assets Under Management (“AUM”) are either deployed in commodity pools (“funds”) or Managed Accounts. Institutional investors typically insist upon managed accounts, especially in the post-Madoff era, as a safety measure. Usually the minimum account size for a managed account is in the range $1M to $5M. I invite you to comb through the prospectii of your Trend Following Wizards and extract the minimum account size of each.

    The trading program is applied to each managed account, individually. The fact that there are two thousand customers, each with a $5M account (thus $10B AUM), is irrelevant. Each $5M account is traded individually, so the total number of systems and positions and scalings-in must suit a $5M account. Not a $10B account.

    Key point: the trading program is designed to trade the minimum account size.

    Commodity pools (“funds”) are often marketed to smaller customers such as HNWIs, who don’t have the means or the desire to open a $3M managed account. Minimum account size in pools is in the range $0.05M to $0.25M. If you gather together enough small-fry to make a $200M pool (which is a single trading account), then yes, you can trade more systems and use fine grained scaling. BUT you can’t market this trading program to institutional investors, because institutions demand managed accounts and they insist on “giving you a tryout” with a $5M test account for the first 18 months.

  • Jez Liberty

    very good point Pumpernickel, and the figures are indeed in line with what I usually see on the “wizards” prospectii – but I was under the impression that BlueTrend trades as a fund, and not as a CTA with managed accounts…

    Actually, I looked again at some of their prospectus and I believe this is the case (ie BlueTrend Fund Limited/LP for offshore/US). Also from previous communication with them, they do NOT market or give access to the BlueTrend fund to any individual, only institutions (their individuals offering is the listed AllBlue fund – which is not a true Trend Following fund but a mix of several funds BlueCrest sell). So I believe BlueTrend might trade as one large account (and with such large amounts running to $10B they surely must diversify with multiple systems/timeframes – if only to limit their impact on the market).

Leave a Comment